ABAQUS

ABAQUS is a finite element analysis program owned and supported by SIMULIA, the Dassault Systèmes brand for Realistic Simulation.

Important Note: From March 14, 2018, Abaqus is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

Versions

The available programs are ABAQUS/CAE, ABAQUS/Standard and ABAQUS/Explicit. The versions currently available at OSC are:

VERSION	RUBY	OWENS	NOTES
6.13	X
6.14	X*	X
2016		X*	Version scheme has been changed
2017		X

*: Default Version

You can use module spider abaqus to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

OSC's ABAQUS license can only be used for educational, institutional, instructional, and/or research purposes. Only users who are faculty, research staff or students at the following institutions are permitted to utilized OSC's license:

The Ohio State University
University of Toledo
University of Cincinnati
Miami University of Ohio
Wright State University
University of Akron

Users from additional degree granting academic institutions may request to be added to this list per a cost by contacting OSC Help.

The use of ABAQUS for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

(link sends e-mail)

Access for Commerical Users

Contac OSC Help for getting access to SOFTWARE if you are a commerical user.

Usage

Token Usage

ABAQUS software usage is monitored though a token-based license manager. This means very time you run a ABAQUS job, tokens are checked out from our pool for your tasks usage. To ensure your job is only started when its required ABAQUS tokens are available it is important to include a software flag within your job script's PBS directives. A minimum of 5 tokens are required per a job, so a 1 node, 1 processor ABAQUS job would need the following PBS software flag: #PBS -l software=abaqus+5 . Jobs requiring more cores will need to request more tokens as calculated with the formula: M = int(5 x N^0.422) , where N is the total number of cores. For common requests, you can refer to the following table:

Cores (nodes x ppn each):	1	2	3	4	6	8	12	16	24	32	48
Tokens needed:	5	6	7	8	10	12	14	16	19	21	25

Usage on Ruby

Set-up on Ruby

To load the default version of ABAQUS module, use module load abaqus . To select a particular software version, use module load abaqus/version . For example, use module load abaqus/6.13 to load ABAQUS version 6.13 on Ruby.

Using ABAQUS

Example input data files are available with the ABAQUS release. The abaqus fetch utility is used to extract these input files for use. For example, to fetch input file for one of the sample problems including 4 input files, type:

abaqus fetch job=knee_bolster

abaqus fetch job=knee_bolster_ef1

abaqus fetch job=knee_bolster_ef2

abaqus fetch job=knee_bolster_ef3

Also, use the abaqus help utility is to list all the abaqus execution procedures.

Batch Usage on Ruby

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your ABAQUS analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Ruby, one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=1:00:00

which gives you 20 cores ( -l nodes=1:ppn=20 ) with 1 hour ( -l walltime=1:00:00 ). You may adjust the wall time per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice.

Assume that you have fetched the four input files above into your work directory (where you submit your job, represented by $PBS_O_WORKDIR ). Below is the example batch script ( job.txt ) for a serial run:

#PBS -N knee
#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=20
#PBS -l software=abaqus+17
#PBS -j oe
#
# The following lines set up the ABAQUS environment
#
module load abaqus
#
# Move to the directory where the job was submitted
#
cd $PBS_O_WORKDIR
cp *.inp $TMPDIR
cd $TMPDIR
#
# Run ABAQUS
#
abaqus job=knee_bolster cpus=20 interactive
#
# Now, copy data (or move) back once the simulation has completed
#
cp * $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt .

NOTE:

Make sure to copy all the files needed (input files, restart files, user subroutines, python scripts etc.) from your work directory ( $PBS_O_WORKDIR ) to $TMPDIR , and copy your results back at the end of your script. Running your job on $TMPDIR ensures maximum efficiency.
The keyword interactive is required in the execution line abaqus job=knee_bolster interactive for the following reason: If left off, ABAQUS will background the simulation process. Backgrounding a process in the OSC environment will place it outside of the batch job and it will receive the default 1 hour of CPU time and corresponding default memory limits. The keyword interactive in this case simply tells ABAQUS not to return until the simulation has completed.
The name of the input file is sometimes omitted in the execution line, which may work fine if you've copied only the input files for one specific model. Although, it is better practice to designate the main input file explicitly by adding input=<my_input_file_name>.inp to the execution line so that it looks like abaqus job=knee_bolster input=<my_input_file_name>.inp interactive .
Ruby has full node charging policy. Define nodes=1:ppn=20 for a serial run.

Non-interactive Batch Job (Parallel Run)

Below is the example batch script ( job.txt ) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=20
#PBS -N my_abaqus_job
#PBS -l software=abaqus+23
#PBS -j oe
#
# The following lines set up the ABAQUS environment
#
module load abaqus
#
# Move to the directory where the job was submitted
#
cd $PBS_O_WORKDIR
cp *.inp $TMPDIR/
cd $TMPDIR
#
# Run ABAQUS, note that in this case we have provided the names of the input files explicitly
#
abaqus job=test input=<my_input_file_name1>.inp cpus=40 interactive
#
# Now, move data back once the simulation has completed
#
mv * $PBS_O_WORKDIR

NOTE:

Define nodes>1 for a parallel run.
Specify cpus=<n> in the execution line, where n=nodes*ppn.
Everything else is similar to the serial script above.

Usage on Owens

Set-up on Owens

To load the default version of ABAQUS module, use module load abaqus . To select a particular software version, use module load abaqus/version . For example, use module load abaqus/2016 to load ABAQUS version 2016 on Owens.

Using ABAQUS

Example input data files are available with the ABAQUS release. The abaqus fetch utility is used to extract these input files for use. For example, to fetch input file for one of the sample problems including 4 input files, type:

abaqus fetch job=knee_bolster

abaqus fetch job=knee_bolster_ef1

abaqus fetch job=knee_bolster_ef2

abaqus fetch job=knee_bolster_ef3

Also, use the abaqus help utility is to list all the abaqus execution procedures.

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your ABAQUS analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Owens, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores ( -l nodes=1:ppn=28 ) with 1 hour ( -l walltime=1:00:00 ). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice.

Assume that you have fetched the four input files above into your work directory (where you submit your job, represented by $PBS_O_WORKDIR ). Below is the example batch script ( job.txt ) for a serial run:

#PBS -N knee
#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=1
#PBS -l software=abaqus+5
#PBS -j oe
#
# The following lines set up the ABAQUS environment
#
module load abaqus
#
# Move to the directory where the job was submitted
#
cd $PBS_O_WORKDIR
cp *.inp $TMPDIR
cd $TMPDIR
#
# Run ABAQUS
#
abaqus job=knee_bolster interactive
#
# Now, copy data (or move) back once the simulation has completed
#
cp * $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt

NOTE:

Make sure to copy all the files needed (input files, restart files, user subroutines, python scripts etc.) from your work directory ( $PBS_O_WORKDIR ) to $TMPDIR , and copy your results back at the end of your script. Running your job on $TMPDIR ensures maximum efficiency.
The keyword interactive is required in the execution line abaqus job=knee_bolster interactive for the following reason: If left off, ABAQUS will background the simulation process. Backgrounding a process in the OSC environment will place it outside of the batch job and it will receive the default 1 hour of CPU time and corresponding default memory limits. The keyword interactive in this case simply tells ABAQUS not to return until the simulation has completed.
The name of the input file is sometimes omitted in the execution line, which may work fine if you've copied only the input files for one specific model. Although, it is better practice to designate the main input file explicitly by adding input=<my_input_file_name>.inp to the execution line so that it looks like abaqus job=knee_bolster input=<my_input_file_name>.inp interactive .
Define nodes=1 (1<=ppn<=28 for Owens) for a serial run.
If ppn>1, add cpus=<n> to the execution line, where n=ppn. This is how it should look like: abaqus job=test input=<my_input_file_name1>.inp cpus=<n> interactive .

Non-interactive Batch Job (Parallel Run)

Note: abaqus will not run correctly in parallel with input files in $TMPDIR! Use the scratch file system.

Below is the example batch script ( job.txt ) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=28:pfsdir
#PBS -N my_abaqus_job
#PBS -l software=abaqus+27
#PBS -j oe
#
# The following lines set up the ABAQUS environment
#
module load abaqus
#
# Move to the directory where the job was submitted
#
cd $PBS_O_WORKDIR
#
# Cope input files to /fs/scratch and run Abaqus there
#
cp *.inp $PFSDIR
cd $PFSDIR
#
# Run ABAQUS, note that in this case we have provided the names of the input files explicitly
#
abaqus job=test input=<my_input_file_name1>.inp cpus=56 interactive
#
# Now, move data back once the simulation has completed
#
mv * $PBS_O_WORKDIR

NOTE:

Define nodes>1 for a parallel run. 1<=ppn<=28 for Owens. Please note that if you use a partial node (using less than 28 cores), there will be performance drop. We are currently investigating this issue (see Known issue).
Specify cpus=<n> in the execution line, where n=nodes*ppn.
Everything else is similar to the serial script above.
Usage of user-defined material (UMAT) script in Fortran is limited on Owens as following:
1. abaqus 2017: correctly running on single and multi-nodes
2. abaqus 6.14 and 2016: correctly running on a single node

Please read Known issue for more detail.

ABINIT

ABINIT is a package whose main program allows one to find the total energy, charge density and electronic structure of systems made of electrons and nuclei (molecules and periodic solids) within Density Functional Theory (DFT), using pseudopotentials and a planewave or wavelet basis. ABINIT also includes options to optimize the geometry according to the DFT forces and stresses, or to perform molecular dynamics simulations using these forces, or to generate dynamical matrices, Born effective charges, and dielectric tensors, based on Density-Functional Perturbation Theory, and many more properties.

Availability and Restrictions

Versions

The following versions of ABINIT are available on OSC clusters:

Version	Oakley
7.10.5	X*

* Current default version

You can use module spider abinit to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

ABINIT is available to all OSC users without restriction.

Usage on Oakley

Set-up

To configure your environment for use of ABINIT on Oakley, run the following command: module load abinit. The default version will be loaded. To select a particular ABINIT version, use module load abinit/version. For example, use module load abinit/7.10.5 to load ABINIT 7.10.5 on Oakley.

AMBER

The Assisted Model Building with Energy Refinement (AMBER) package contains many molecular simulation programs targeted at biomolecular systems. A wide variety of modelling techniques are available. It generally scales well on modest numbers of processors, and the GPU enabled CUDA programs are very efficient.

Availability and Restrictions

Versions

AMBER is available on Oakley, Ruby, and Owens Clusters. The following versions are currently available at OSC (S means serial executables, P means parallel, C means CUDA, i.e., GPU enabled, and M means MIC, i.e., Xeon Phi enabled):

Version	Oakley	ruby	Owens	notes
11	SPC			Default version on Oakley prior to 09/15/2015
12	SPC
14	SPC	SPC		Default version on Oakley and Ruby prior to 02/21/2017
16	SPC*	SPCM*	SPC*

* Current default version

You can use module spider amber to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

OSC's Amber is available to not-for-profit OSC users; simply contact OSC Help to request the appropriate form for access.

Access for Commerical Users

For-profit OSC users must obtain their own Amber license.

Usage

Usage on Oakley

Set-up

To load the default version of AMBER module, use module load amber. To select a particular software version, use module load amber/version. For example, use module load amber/14 to load AMBER version 14.

Using AMBER

A serial Amber program in a short duration run can be executed interactively on the command line, e.g.:

tleap

Parallel Amber programs must be run in a batch environment with mpiexec, e.g.:

mpiexec pmemd.MPI

The CUDA Amber programs, which must be run in a batch environment, are recommended due to their computational performance; e.g.:

pmemd.cuda

For supported features of the GPU accelerated version of PMEMD see Amber GPU Support.

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your AMBER simulation to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and Amber input files are available here:

~srb/workshops/compchem/amber/

Below is the example batch script (job.txt) for a serial run:

# AMBER Example Batch Script for the Basic Tutorial in the Amber manual
#PBS -N 6pti
#PBS -l nodes=1:ppn=1
#PBS -l walltime=0:20:00
module load amber
# Use TMPDIR for best performance.
cd $TMPDIR
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cp -p $PBS_O_WORKDIR/6pti.prmtop .
cp -p $PBS_O_WORKDIR/6pti.prmcrd .
# Running minimization for BPTI
cat << eof > min.in
# 200 steps of minimization, generalized Born solvent model
&cntrl
maxcyc=200, imin=1, cut=12.0, igb=1, ntb=0, ntpr=10,
/
eof
sander -i min.in -o 6pti.min1.out -p 6pti.prmtop -c 6pti.prmcrd -r 6pti.min1.xyz
cp -p min.in 6pti.min1.out 6pti.min1.xyz $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt.

Usage on Ruby

Set-up

To load the default version of AMBER module, use module load amber. To select a particular software version, use module load amber/version. For example, use module load amber/14 to load AMBER version 14.

Using AMBER

A serial Amber program in a short duration run can be executed interactively on the command line, e.g.:

tleap

Parallel Amber programs must be run in a batch environment with mpiexec, e.g.:

mpiexec pmemd.MPI

The CUDA Amber programs, which must be run in a batch environment, are recommended due to their computational performance; e.g.:

pmemd.cuda

For supported features of the GPU accelerated version of PMEMD see Amber GPU Support.

The MIC, i.e., Xeon Phi enabled, Amber programs must be run in a batch environment. The Xeon Phi coprocessors on Ruby were removed from service. Please contact OSC Help if you have any questions or want to get access to such resources.

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your AMBER simulation to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=1:00:00

which gives you 20 cores (-l nodes=1:ppn=20) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and Amber input files are available here:

~srb/workshops/compchem/amber/

Below is the example batch script (job.txt) for a serial run:

# AMBER Example Batch Script for the Basic Tutorial in the Amber manual
#PBS -N 6pti
#PBS -l nodes=1:ppn=20
#PBS -l walltime=0:20:00
module load amber
# Use TMPDIR for best performance.
cd $TMPDIR
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cp -p $PBS_O_WORKDIR/6pti.prmtop .
cp -p $PBS_O_WORKDIR/6pti.prmcrd .
# Running minimization for BPTI
cat << eof > min.in
# 200 steps of minimization, generalized Born solvent model
&cntrl
maxcyc=200, imin=1, cut=12.0, igb=1, ntb=0, ntpr=10,
/
eof
sander -i min.in -o 6pti.min1.out -p 6pti.prmtop -c 6pti.prmcrd -r 6pti.min1.xyz
cp -p min.in 6pti.min1.out 6pti.min1.xyz $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt.

Usage on Owens

Set-up

To load the default version of AMBER module, use module load amber. To select a particular software version, use module load amber/version. For example, use module load amber/16 to load AMBER version 16.

Using AMBER

A serial Amber program in a short duration run can be executed interactively on the command line, e.g.:

tleap

Parallel Amber programs must be run in a batch environment with mpiexec, e.g.:

mpiexec pmemd.MPI

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your AMBER simulation to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and Amber input files are available here:

~srb/workshops/compchem/amber/

Below is the example batch script (job.txt) for a serial run:

# AMBER Example Batch Script for the Basic Tutorial in the Amber manual
#PBS -N 6pti
#PBS -l nodes=1:ppn=28
#PBS -l walltime=0:20:00
module load amber
# Use TMPDIR for best performance.
cd $TMPDIR
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cp -p $PBS_O_WORKDIR/6pti.prmtop .
cp -p $PBS_O_WORKDIR/6pti.prmcrd .
# Running minimization for BPTI
cat << eof > min.in
# 200 steps of minimization, generalized Born solvent model
&cntrl
maxcyc=200, imin=1, cut=12.0, igb=1, ntb=0, ntpr=10,
/
eof
sander -i min.in -o 6pti.min1.out -p 6pti.prmtop -c 6pti.prmcrd -r 6pti.min1.xyz
cp -p min.in 6pti.min1.out 6pti.min1.xyz $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt.

Troubleshooting

In general, the scientific method should be applied to usage problems. Users should check all inputs and examine all outputs for the first signs of trouble. When one cannot find issues with ones inputs, it is often helpful to ask fellow humans, especially labmates, to review the inputs and outputs. Reproducibility of molecular dynamics simulations is subject to many caveats. See page 24 of the Amber16 manual for a discussion.

ANSYS

ANSYS offers a comprehensive software suite that spans the entire range of physics, providing access to virtually any field of engineering simulation that a design process requires. Supports are provided by ANSYS, Inc.

Important Note: From March 14, 2018, Ansys is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

Versions

VERSION	OWENS
17.2	X*
18.1	X

* Current default version

OSC has Academic Multiphysics Campus Solution license from Ansys. The license includes most of all the features that Ansys provides. See "Academic Multiphyscis Campus Solution Products" in this table for all available products at OSC.

Access for Academic Users

OSC has an "Academic Research " license for ANSYS. This allows for academic use of the software by Ohio faculty and students, with some restrictions. To view current ANSYS node restrictions, please see ANSYS's Terms of Use.

Use of ANSYS products at OSC for academic purposes requires validation. Please contact OSC Help for further instruction.

Access for Commerical Users

Contact OSC Help for getting access to ANSYS if you are a commercial user.

Usage

For more information on how to use each ANSYS product at OSC systems, refer to its documentation page provided at the end of this page.

Note

Due to the way our Fluent and ANSYS modules are configured, simultaneously loading multiple of either module will cause a cryptic error. The most common case of this happening is when multiple of a user's jobs are started at the same time and all load the module at once. In order for this error to manifest, the modules have to be loaded at precisely the same time; a rare occurrence, but a probable occurrence over the long term.

If you encounter this error you are not at fault. Please resubmit the failed job(s).

If you frequently submit large amounts of Fluent or ANSYS jobs, we recommend you stagger your job submit times to lower the chances of two jobs starting at the same time, and hence loading the module at the same time. Another solution is to establish job dependencies between jobs, so jobs will only start one after another. To do this, you would add the PBS directive:

#PBS -W after:jobid

To jobs you want to only start after another job has started. You would replace jobid with the job ID of the job to wait for. If you have additional questions, please contact OSC Help.

ANSYS Mechanical

ANSYS Mechanical is a finite element analysis (FEA) tool that enables you to analyze complex product architectures and solve difficult mechanical problems. You can use ANSYS Mechanical to simulate real world behavior of components and sub-systems, and customize it to test design variations quickly and accurately.

Important Note: From March 14, 2018, Ansys is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

ANSYS Mechanical is available on the Oakley Cluster. The versions currently available at OSC are:

Version	owens	Notes
17.2	X*
18.1	X

*: Current default version

You can use module spider ansys to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

Use of ANSYS for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Access for Commerical Users

Contact OSC Help for getting access to ANSYS if you are a commerical user.

Usage

Usage on Owens

Set-up on Owens

To load the default version of ANSYS module, use module load ansys . To select a particular software version, use module load ansys/version . For example, use module load ansys/17.2 to load ANSYS version 17.2 on Owens.

Using ANSYS Mechanical

Following a successful loading of the ANSYS module, you can access the ANSYS Mechanical commands and utility programs located in your execution path:

ansys <switch options> <file>

The ANSYS Mechanical command takes a number of Unix-style switches and parameters.

The -j Switch

The command accepts a -j switch. It specifies the "job id," which determines the naming of output files. The default is the name of the input file.

The -d Switch

The command accepts a -d switch. It specifies the device type. The value can be X11, x11, X11C, x11c, or 3D.

The -m Switch

The command accepts a -m switch. It specifies the amount of working storage obtained from the system. The units are megawords.

The memory requirement for the entire execution will be approximately 5300000 words more than the -m specification. This is calculated for you if you use ansnqs to construct an NQS request.

The -b [nolist] Switch

The command accepts a -b switch. It specifies that no user input is expected (batch execution).

The -s [noread] Switch

The command accepts a -s switch. By default, the start-up file is read during an interactive session and not read during batch execution. These defaults may be changed with the -s command line argument. The noread option of the -s argument specifies that the start-up file is not to be read, even during an interactive session. Conversely, the -s argument with the -b batch argument forces the reading of the start-up file during batch execution.

The -g [off] Switch

The command accepts a -g switch. It specifies that the ANSYS graphical user interface started automatically.

ANSYS Mechanical parameters

ANSYS Mechanical parameters may be assigned values on the command. The parameter must be at least two characters long and must be a legal parameter name. The ANSYS Mechanical parameter that is to be assigned a value should be given on the command line with a preceding dash (-), a space immediately after, and the value immediately after the space:

module load ansys
ansys -pval1 -10.2 -EEE .1e6
sets pval1 to -10.2 and EEE to 100000

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your ANSYS Mechanical analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

Interactive mode is similar to running ANSYS Mechanical on a desktop machine in that the graphical user interface will be sent from OSC and displayed on the local machine. Interactive jobs are run on compute nodes of the cluster, by turning on X11 forwarding. The intention is that users can run ANSYS Mechanical interactively for the purpose of building their model and preparing their input file. Once developed this input file can then be run in no-interactive batch mode.

To run interactive ANSYS Mechanical, a batch job need to be submitted from the login node, to request necessary compute resources, with X11 forwarding. For example, the following command requests one whole node with 28 cores ( -l nodes=1:ppn=28 ), for a walltime of 1 hour ( -l walltime=1:00:00 ), with one ANSYS license:

qsub -I -X -l nodes=1:ppn=28 -l walltime=1:00:00 -l software=ansys+1

You may adjust the numbers per your need. This job will queue until resources becomes available. Once the job is started, you're automatically logged in on the compute node; and you can launch ANSYS Mechanical and start the graphic interface with the following commands:

module load ansys
ansys -g

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. For a given model, prepare the input file with ANSYS Mechanical commands (named ansys.in for example) for the batch run. Below is the example batch script ( job.txt ) for a serial run:

#PBS -N ansys_test  
#PBS -l walltime=30:00:00  
#PBS -l nodes=1:ppn=1
#PBS -l software=ansys+1  
#PBS -j oe
cd $TMPDIR  
cp $PBS_O_WORKDIR/ansys.in .    
module load ansys  
ansys < ansys.in   
cp <output files> $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt .

Non-interactive Batch Job (Parallel Run)

To take advantage of the powerful compute resources at OSC, you may choose to run distributed ANSYS Mechanical for large problems. Multiple nodes and cores can be requested to accelerate the solution time. Note that you'll need to change your batch script slightly for distributed runs.

Starting from September 15, 2015, a job using HPC tokens (with "ansyspar" flag) should be submitted to Owens clusters due to scheduler issue.

For distributed ANSYS Mechanical jobs using one node (nodes=1), the number of processors needs to be specified in the command line with options '-dis -np':

#PBS -N ansys_test 
#PBS -l walltime=3:00:00 
#PBS -l nodes=1:ppn=28
#PBS -l software=ansys+1%ansyspar+24
...
ansys -dis -np 12 < ansys.in  
...

Notice that in the script above, the ansys parallel license is requested as well as ansys license in the format of

#PBS -l software=ansys+1%ansyspar+n

where n=m-4, with m being the total cpus called for this job. This line is necessary when the total cpus called is greater than 4 (m>4), which applies for the parallel example below as well.

For distributed jobs requesting multiple nodes, you need to specify the number of processors for each node in the command line. This information can be obtained from $PBS_NODEFILE. The following shows changes in the batch script if 2 nodes on Owens are requested for a parallel ANSYS Mechanical job:

#PBS -N ansys_test 
#PBS -l walltime=3:00:00 
#PBS -l nodes=2:ppn=28
#PBS -l software=ansys+1%ansyspar+52
...
export MPI_WORKDIR=$PWD
machines=`uniq -c ${PBS_NODEFILE} | awk '{print $2 ":" $1}' | paste -s -d ':'`
ansys -dis -machines $machines < ansys.in  
...
pbsdcp -g '<output files>' $PBS_O_WORKDIR

The 'pbsdcp -g' command in the last line in the script above makes sure that all result files generated by different compute nodes are copied back to the work directory.

CFX

ANSYS CFX (called CFX hereafter) is a computational fluid dynamics (CFD) program for modeling fluid flow and heat transfer in a variety of applications.

Important Note: From March 14, 2018, Ansys is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

CFX is available on the Oakley Cluster. The versions currently available at OSC are:

VERSION	owens	notes
17.2	X*
18.1	X

*: Current default version

You can use module spider fluent to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

Use of ANSYS products for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Currently, there are in total 50 ANSYS CFD base license tokens and 256 HPC tokens for academic users. These base tokens are shared by available ANSYS CFD related projusts (see "Academic Research -> ANSYS Academic Research CFD" in this table for details). These HPC tokens are shared with all ANSYS products we have at OSC. A base license token will allow CFX to use up to 4 cores without any additional tokens. If you want to use more than 4 cores, you will need an additional "HPC" token per core. For instance, a serial CFX job with 1 core will need 1 base license token while a parallel CFX job with 12 cores will need 1 base license token and 8 HPC tokens.

Access for Commerical Users

Contact OSC Help for getting access to CFX if you are a commerical user.

Usage

Usage on Owens

Set-up on Owens

To load the default version, use module load ansys . To select a particular software version, use module load ansys/version . For example, use module load ansys/17.2 to load CFX version 17.2 on Owens.

Batch Usage on owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

Interactive mode is similar to running CFX on a desktop machine in that the graphical user interface will be sent from OSC and displayed on the local machine. Interactive jobs are run on compute nodes of the cluster, by turning on X11 forwarding. The intention is that users can run CFX interactively for the purpose of building their model and preparing their input file. Once developed this input file can then be run in no-interactive batch mode.

To run interactive CFX GUI, a batch job need to be submitted from the login node, to request necessary compute resources, with X11 forwarding. Please follwoing the steps below to use CFX GUI interactivly:

Ensure that your SSH client software has X11 forwarding enabled
Connect to Owens system
Request an interactive job. The command below will request one whole node with 28 cores ( -l nodes=1:ppn=28 ), for a walltime of one hour ( -l walltime=1:00:00 ), with one ANSYS CFD license (modify as per your own needs):
```
qsub -I -X -l nodes=1:ppn=28 -l walltime=1:00:00 -l software=ansys+1
```
Once the interactive job has started, run the following commands to setup and start the CFX GUI:
```
module load ansys
cfx5 
```

Non-interactive Batch Job (Serial Run Using 1 Base Token)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice.

Below is the example batch script ( job.txt ) for a serial run with an input file test.def ) on Glenn:

#PBS -N serialjob_cfx
#PBS -l walltime=1:00:00
#PBS -l software=ansys+1
#PBS -l nodes=1:ppn=1
#PBS -j oe
#PBS -S /bin/bash
#Set up CFX environment.
module load ansys
#'cd' directly to your working directory
cd $PBS_O_WORKDIR
#Copy CFX files like .def to $TMPDIR and move there to execute the program
cp test.def $TMPDIR/
cd $TMPDIR
#Run CFX in serial with test.def as input file
cfx5solve -batch -def test.def 
#Finally, copy files back to your home directory
cp  * $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt

Non-interactive Batch Job (Parallel Execution using HPC token)

CFX can be run in parallel, but it is very important that you read the documentation in the CFX Manual on the details of how this works.

In addition to requesting the base license token ( -l software=ansys+1 ), you need to request copies of the ansyspar license, i.e., HPC tokens ( -l software=ansys+1%ansyspar+[n] ), where [n] is equal to the number of cores you requested minus 4.

Parallel jobs have to be submitted on Owens via the batch system. An example of the batch script follows:

#PBS -N paralleljob_cfx
#PBS -l walltime=10:00:00
#PBS -l nodes=2:ppn=28
#PBS -l software=ansys+1%ansyspar+52
#PBS -j oe
#PBS -S /bin/bash
#Set up CFX environment.
module load ansys
#'cd' directly to your working directory
cd $PBS_O_WORKDIR
#Copy CFX files like .def to $TMPDIR and move there to execute the program
cp test.def $TMPDIR/
cd $TMPDIR
#Convert PBS_NODEFILE information into format for CFX host list
nodes=`cat $PBS_NODEFILE`
nodes=`echo $nodes | sed -e 's/ /,/g'`
#Run CFX in parallel with new.def as input file
#if multiple nodes
cfx5solve -batch -def test.def  -par-dist $nodes -start-method "Platform MPI Distributed Parallel"
#if one node
#cfx5solve -batch -def test.def -par-dist $nodes -start-method "Platform MPI Local Parallel"
#Finally, copy files back to your home directory
cp  * $PBS_O_WORKDIR

FLUENT

ANSYS FLUENT (called FLUENT hereafter) is a state-of-the-art computer program for modeling fluid flow and heat transfer in complex geometries.

Important Note: From March 14, 2018, Ansys is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

FLUENT is available on the Oakley Cluster. The versions currently available at OSC are:

Version	owens	NOTEs
17.2	X*
18.1	X

*: Current default version

You can use module spider ansys for Owens to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

Use of ANSYS products for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Currently, there are in total 50 ANSYS CFD base license tokens and 256 HPC tokens for academic users. These base tokens are shared by available ANSYS CFD related projects (see "Academic Research -> ANSYS Academic Research CFD" in this table for details). These HPC tokens are shared with all ANSYS products we have at OSC. A base license token will allow FLUENT to use up to 4 cores without any additional tokens. If you want to use more than 4 cores, you will need an additional "HPC" token per core. For instance, a serial FLUENT job with 1 core will need 1 base license token while a parallel FLUENT job with 12 cores will need 1 base license token and 8 HPC tokens.

Access for Commerical Users

Contact OSC Help for getting access to FLUENT if you are a commercial user.

Usage

Usage on Owens

Set-up on Owens

To load the default version of FLUENT module, use module load ansys . To select a particular software version, use module load ansys/version . For example, use module load ansys/17.2 to load FLUENT version 17.2 on Owens.

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a Linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your FLUENT analysis to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

Interactive mode is similar to running FLUENT on a desktop machine in that the graphical user interface will be sent from OSC and displayed on the local machine. Interactive jobs are run on compute nodes of the cluster, by turning on X11 forwarding. The intention is that users can run FLUENT interactively for the purpose of building their model and preparing their input file. Once developed this input file can then be run in non-interactive batch mode.

To run interactive FLUENT GUI, a batch job need to be submitted from the login node, to request necessary compute resources, with X11 forwarding. Please following the steps below to use FLUENT GUI interactively:

Ensure that your SSH client software has X11 forwarding enabled
Connect to Owens system
Request an interactive job. The command below will request one whole node with 28 cores ( -l nodes=1:ppn=28 ), for a walltime of one hour ( -l walltime=1:00:00 ), with one FLUENT license (modify as per your own needs):
```
qsub -I -X -l nodes=1:ppn=28 -l walltime=1:00:00 -l software=ansys+1
```
Once the interactive job has started, run the following commands to setup and start the FLUENT GUI:
```
module load ansys
fluent 
```

Non-interactive Batch Job (Serial Run Using 1 Base Token)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice.

Below is the example batch script ( job.txt ) for a serial run with an input file run.input ) on Owens:

#PBS -N serial_fluent
#PBS -l walltime=5:00:00 
#PBS -l nodes=1:ppn=1
#PBS -l software=ansys+1
#PBS -j oe
#
# The following lines set up the FLUENT environment
#
module load ansys
#
# Move to the directory where the job was submitted from
# You could also 'cd' directly to your working directory
cd $PBS_O_WORKDIR
#
# Copy files to $TMPDIR and move there to execute the program
#
cp test_input_file.cas test_input_file.dat run.input $TMPDIR
cd $TMPDIR
#
# Run fluent
fluent 3d -g < run.input  
#
# Where the file 'run.input' contains the commands you would normally
# type in at the Fluent command prompt.
# Finally, copy files back to your home directory
cp *   $PBS_O_WORKDIR

As an example, your run.input file might contain:

file/read-case-date test_input_file.cas 
solve/iterate 100
file/write-case-data test_result.cas
file/confirm-overwrite yes    
exit  
yes

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt

Non-interactive Batch Job (Parallel Execution using HPC token)

FLUENT can be run in parallel, but it is very important that you read the documentation in the FLUENT Manual on the details of how this works.

In addition to requesting the FLUENT base license token ( -l software=ansys+1 ), you need to request copies of the ansyspar license, i.e., HPC tokens ( -l software=ansys+1%ansyspar+[n] ), where [n] is equal to the number of cores you requested minus 4.

Parallel jobs have to be submitted to Owens via the batch system. An example of the batch script follows:

#PBS -N parallel_fluent   
#PBS -l walltime=1:00:00   
#PBS -l nodes=2:ppn=28
#PBS -j oe
#PBS -l software=ansys+1%ansyspar+52
#PBS -S /bin/bash
set echo on   
hostname   
#   
# The following lines set up the FLUENT environment   
#   
module load ansys
#   
# Move to the directory where the job was submitted from and   
# create the config file for socket communication library   
#   
cd $PBS_O_WORKDIR   
#   
# Create list of nodes to launch job on   
rm -f pnodes   
cat  $PBS_NODEFILE | sort > pnodes   
export ncpus=`cat pnodes | wc -l`   
#   
#   Run fluent   
fluent 3d -t$ncpus -pinfiniband.ofed -cnf=pnodes -g < run.input

Workbench Platform

ANSYS Workbench platform is the backbone for delivering a comprehensive and integrated simulation system to users. See ANSYS Workbench platform for more information.

Important Note: From March 14, 2018, Ansys is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

ANSYS Workbench is available on Oakley Cluster. The versions currently available at OSC are:

Version		owens
17.2	SF	X*
	CFD	X*
18.1	SF	X
	CFD	X

*: Current default version

Note:

SF: Structural-Fluid dynamics related applications. See "Academic Research -> ANSYS Academic Research Mechanical and CFD" in this table for all available products
CFD: CFD related applications. See "Academic Research -> ANSYS Academic Research CFD" in this table for all available products

You can use module spider ansys to view available modules for a given machine if you want to use structural-fluid dynamics related applications or module spider fluent to view available modules for a given machine if you want to use CFD related applications. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

Use of ANSYS products for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Access for Commerical Users

Contact OSC Help for getting access to ANSYS if you are a commerical user.

Usage

Usage on Owens

Set-up for Structural-Fluid dynamics related applications

To load the default version , use module load ansys . To select a particular software version, use module load ansys/version . For example, use module load ansys/17.2 to load version 17.2 on Owens. After the module is loaded, use the following command to open Workbench GUI:

runwb2

Set-up for CFD related applications

To load the default version , use module load ansys . To select a particular software version, use module load ansys/version . For example, use module load ansys/17.2 to load version 17.2 on Owens. After the module is loaded, use the following command to open Workbench GUI:

runwb2

ARM HPC tools

ARM HPC tools analyze how HPC software runs. It consists of three applications, ARM DDT, ARM Performance Reports and ARM MAP:

ARM DDT: graphical debugger for HPC applications.
ARM MAP: HPC application profiler with easy-to-use GUI environment.
ARM Performance Reports: simple tool to generate a single-page HTML or plain text report that presents overall performance characteristics of HPC applications.

NOTE: Because ARM has aquired Allinea, all Allinea module files have been renamed accordingly. Allinea modules are still available and have same functionality as new ARM modules.

Availability & Restrictions

Versions

The following versions of ARM HPC tools are available on OSC clusters:

version	oakley	Ruby	Owens
5.0.1		X
5.1	X	X
6.0	X	X
6.0.1	X	X
6.0.6			X
6.1.1		X
7.0	X*	X*	X*
7.1			X

* Current default version

You can use module spider arm to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

ARM DDT, MAP and Performance Reports are available to all OSC users.

Usage

ARM DDT

ARM DDT is a debugger for HPC software that automatically alerts users of memory bugs and divergent behavior. For more features and benefits, visit ARM HPC tools and libraries - DDT.

For usage instructions and more iformation, read ARM DDT.

ARM MAP

ARM MAP produces a detailed profile of HPC software. Unlike ARM Performance Reports, you must have the source code to run ARM MAP because its analysis details the software line-by-line. For more features and benefits, visit ARM HPC tools and libraries - MAP.

For usage instructions and more information, read ARM MAP.

ARM Performance Reports

ARM Performance Reports analyzes and documents information on CPU, MPI, I/O, and Memory performance characteristics of HPC software, even third party code, to aid understanding about the overall performance. Although it should not be used all the time, ARM Performance Reports is recommended to OSC users as a viable option to analyze how an HPC application runs. View an example report to navigate the format of a typical report. For more example reports, features and benefits, visit ARM HPC tools and libraries - Performance Reports.

For usage instructions and more information, read ARM Performance Reports.

Troubleshooting

Using ARM software with MVAPICH2

This note from ARM's Getting Started Guide applies to both perf-report and MAP:

Some MPIs, most notably MVAPICH, are not yet supported by ARM's Express Launch mode
(in which you can just put “perf-report” in front of an existing mpirun/mpiexec line). These can
still be measured using the Compatibility Launch mode.

Instead of this Express Launch command:

perf-report mpiexec <mpi args> <program> <program args> # BAD

Use the compatibility launch version instead:

perf-report -n <num procs> --mpiargs="<mpi args>" <program> <program args>

ARM Performance Reports

ARM Performance Reports is a simple tool used to generate a single-page HTML or plain text report that presents the overall performance characteristics of HPC applications. It supports pthreads, OpenMP, or MPI code on CPU, GPU, and MIC based architectures.

Availability and Restrictions

Versions

The versions currently available at OSC are:

VERSION	OAKLEY	RUBY	OWENS
5.0.1		X
5.1	X	X
6.0	X	X
6.0.1	X	X
6.0.6			X
6.1.1		X
7.0	X*	X*	X*
7.1			X

* Current default version

You can use module spider arm-pr to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

ARM Performance Reports is available to all OSC users. We have 64 seats with 64 HPC tokens. Users can monitor the license status here.

Usage

Set-up

To load the module for the ARM Performance Reports default version, use module load arm-pr. To select a particular software version, use module load arm-pr/version. For example, use module load arm-pr/6.0 to load ARM Performance Reports version 6.0, provided the version is available on the OSC cluster in use.

Using ARM Performance Reports

You can use your regular executables to generate performance reports. The program can be used to analyze third-party code as well as code you develop yourself. Performance reports are normally generated in a batch job.

To generate a performance report for an MPI program:

module load arm-pr
perf-report -np <num procs> --mpiargs="<mpi args>" <program> <program args>

where <num procs> is the number of MPI processes to use, <mpi args> represents arguments to be passed to mpiexec (other than -n or -np), <program> is the executable to be run and <program args> represents arguments passed to your program.

For example, if you normally run your program with mpiexec -n 12 wave_c, you would use

perf-report -np 12 wave_c

To generate a performance report for a non-MPI program:

module load arm-pr
perf-report --no-mpi <program> <program args>

The performance report is created in both html and plain text formats. The file names are based on the executable name, number of processes, date and time, for example, wave_c_12p_2016-02-05_12-46.html. To open the report in html format use

firefox wave_c_12p_2016-02-05_12-46.html

For more details, download the ARM Performance Reports User Guide.

Performance Reports with GPU

ARM Performance Reports can be used for CUDA codes. If you have an executable compiled with the CUDA library, you can launch ARM Performance Reports with

perf-report {executable}

For more information, please read the section 6.10 of the ARM Performance Reports User Guide.

ARM MAP

ARM MAP is a full scale profiler for HPC programs. We recommend using ARM MAP after reviewing reports from ARM Performance Reports. MAP supports pthreads, OpenMP, and MPI software on CPU, GPU, and MIC based architectures.

Availability & Restrictions

Versions

The ARM MAP versions currently available at OSC are:

VERSION	OAKLEY	RUBY	OWENS
5.0.1		X
5.1	X	X
6.0	X	X
6.0.1	X	X
6.0.6			X
6.1.1		X
7.0	X*	X*	X*
7.1			X

* Current default version

You can use module spider arm-map to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

ARM MAP is available to all OSC users. We have 64 seats with 80 HPC tokens. Users can monitor the ARM License Server Status.

Usage

Set-up

To load the default version of the ARM MAP module, use module load arm-map. To select a particular software version, use module load arm-map/version. For example, use module load arm-map/6.0 to load ARM MAP version 6.0, provided the version is available on the cluster in use.

Using ARM MAP

Profiling HPC software with ARM MAP typically involves three steps:

1. Prepare the executable for profiling.

Regular executables can be profiled with ARM MAP, but source code line detail will not be available. You need executables with debugging information to view source code line detail: re-compile your code with a -g option added among the other appropriate compiler options. For example:

mpicc wave.c -o wave -g -O3

This executable built with the debug flag can be used for ARM Performance Reports as well.

Note: The -g flag turns off all optimizations by default. For profiling your code you should use the same optimizations as your regular executable, so explicitly include the -On flag, where n is your normal level of optimization, typically -O2 or -O3, as well as any other compiler optimization options.

2. Run your code to produce the profile data file (.map file).

Profiles are normally generated in a batch job. To generate a MAP profile for an MPI program:

module load arm-map
map --profile -np <num proc> --mpiargs="<mpi args>" <program> <program args>

where <num procs> is the number of MPI processes to use, <mpi args> represents arguments to be passed to mpiexec (other than -n or -np), <program> is the executable to be run and <program args> represents arguments passed to your program.

For example, if you normally run your program with mpiexec -n 12 wave_c, you would use

map --profile -np 12 wave_c

To profile a non-MPI program:

module load arm-map
map --profile --no-mpi <program> <program args>

The profile data is saved in a .map file in your current directory.

As a result of this step, a .map file that is the profile data file is created in your current directory. The file name is based on the executable name, number of processes, date and time, for example, wave_c_12p_2016-02-05_12-46.map.

For more details on using ARM MAP, refer to the ARM Forge User Guide.

3. Analyze the profile data file using either the ARM local client or the MAP GUI.

You can open the profile data file using a client running on your local desktop computer. For client installation and usage instructions, please refer to the section: Client Download and Setup. This option typically offers the best performance.

Alternatively, you can run MAP in interactive mode, which launches the graphical user interface (GUI). For example:

map wave_c_12p_2016-02-05_12-46.map

For the GUI application, one should use an OnDemand VDI (Virtual Desktop Interface) or have X11 forwarding enabled (see Setting up X Windows). Note that X11 forwarding can be distractingly slow for interactive applications.

MAP with GPU

ARM MAP can be used for CUDA codes. If you have an executable compiled with the CUDA library, you can launch ARM MAP with

map {executable}

For more information, please read the Chapter 15 of the ARM Forge User Guide.

Client Download and Setup

1. Download the client.

To download the client, go to the ARM website and choose the appropriate ARM Forge remote client download for Windows, Mac, or Linux. For Windows and Mac, just double click on the downloaded file and allow the installer to run. For Linux, extract the tar file using the command tar -xf file_name and run the installer in the extracted file directory with ./installer. Please contact OSC Help, if you have any issues on downloading the client.

Note: ARM was recently bought out by Arm. The ARM Forge client can be downloaded from https://developer.arm.com/products/software-development-tools/hpc/downloads/download-arm-forge/older-versions-of-remote-client-for-arm-forge.

2. Configure the client.

After installation, you can configure the client as follows:

Open the client program. For Windows or Mac, just click the desktop icon or navigate to the application through its file path. For Linux use the command {arm-forge-path}/bin/map.
Once the program is launched, select ARM MAP in the left column.
In the Remote Launch drop down menu, select "Configure...".
Click Add to create a new profile for your login.
In the Host Name section, type your ssh connection. For example: "username@ruby.osc.edu".
For Remote Installation Directory, type /usr/local/arm/forge-{version}, specifying the ARM Forge version number that created the data profile file you are attempting to view. For example, /usr/local/arm/forge-7.0 for ARM Forge version 7.0.
You can test your login information by clicking Test Remote Launch. It will ask your password. Use the same password for the cluster login.
Close the Configure window. You will see a new option under the Remote Launch drop down menu for the host name you entered. Select your profile and login with your password.
If the login was successful, then you should see License Serial:XXX in the bottom left corner of the window.

This login configuration is needed only for the first time of use. In subsequent times, you can just select your profile.

3. Open the profile data file.

After login, click on LOAD PROFILE DATA FILE. This opens a file browser of your home directory on the OSC cluster you logged onto. Go to the directory that contains the .map file and select it. This will open the file and allow you to navigate the source code line-by-line and investigate the performance characteristics.

A license is not required to simply open the client, so it is possible to skip 2. Configure the client, if you download the profile data file to your desktop. You can then open it by just selecting LOAD PROFILE DATA FILE and navigating through a file browser on your local system.

Note that the client is ARM Forge, a client that contains ARM MAP and ARM DDT. ARM DDT is a debugger, and OSC has license only for ARM MAP. If you need a debugger, you can use Totalview instead.

Arm DDT

Arm DDT is a graphical debugger for HPC applications. It supports pthreads, OpenMP, or MPI code on CPU, GPU, and MIC based architectures.

Availability & Restrictions

Versions

The Arm DDT versions currently available at OSC are:

VERSION	OAKLEY	RUBY	OWENS
5.0.1		X
5.1	X	X
6.0	X	X
6.0.1	X	X
6.0.6			X
6.1.1		X
7.0	X*	X*	X*
7.1			X

* Current default version

You can use module spider arm-ddt to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Arm DDT is available to all OSC users. We have 64 seats with 80 HPC tokens. Users can monitor the Arm License Server Status.

Usage

Set-up

To load the module for the Arm DDT default version, use module load arm-ddt. To select a particular software version, use module load arm-ddt/version. For example, use module load arm-ddt/7.0 to load Arm DDT version 7.0, provided the version is available on the OSC cluster in use.

Using Arm DDT

DDT debugs executables to generate DDT reports. The program can be used to debug third-party code as well as code you develop yourself. DDT reports are normally generated in a batch job.

To generate a DDT report for an MPI program:

module load arm-ddt
ddt --offline -np <num procs> --mpiargs="<mpi args>" <program> <program args>

where <num procs> is the number of MPI processes to use, <mpi args> represents arguments to be passed to mpiexec (other than -n or -np), <program> is the executable to be run and <program args> represents arguments passed to your program.

For example, if you normally run your program with mpiexec -n 12 wave_c, you would use

ddt --offline -np 12 wave_c

To debug a non-MPI program:

module load arm-ddt
ddt --offline --no-mpi <program> <program args>

The DDT report is created in html format. The file names are based on the executable name, number of processes, date and time, for example, wave_c_12p_2016-02-05_12-46.html. To open the report use

firefox wave_c_12p_2016-02-05_12-46.html

Using the Arm DDT GUI

To debug with the DDT GUI remove the --offline option. For example, to debug the MPI program above, use

ddt -np 12 wave_c

For a non-MPI program:

ddt --no-mpi <program> <program args>

This will open the DDT GUI, enabling interactive debugging options.

For the GUI application, one should use an OnDemand VDI (Virtual Desktop Interface) or have X11 forwarding enabled (see Setting up X Windows). Note that X11 forwarding can be distractingly slow for interactive applications.

For more details, download the Arm Performance Reports User Guide.

DDT with GPU

DDT can be used for CUDA codes. If you have an executable compiled with the CUDA library, you can launch Arm Performance Reports with

ddt {executable}

For more information, please read the chapter 15 of the Arm Forge User Guide.

Altair HyperWorks

HyperWorks is a high-performance, comprehensive toolbox of CAE software for engineering design and simulation.

Availability & Restrictions

Versions

The following version of Altair Hyperworks can be found for the following environments:

Version	Oakley	OWENS	Statewide
10			X
11	X		X
12	X		X
13	X*	X*	X
14	X		X
2017.1	X	X	X

* Current Default Version

You can use module spider hyperworks to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

HyperWorks is available to all academic clients. Please contact OSC Help to request the appropriate form for access.

Usage

Using HyperWorks through OSC installation

To use HyperWorks on the OSC clusters, first ensure that X11 forwarding is enabled as the HyperWorks workbench is a graphical application. Then, load the hyperworks module:

module load hyperworks

The HyperWorks GUI can be launched then with the following command:

hw

State-wide access for HyperWorks

For information on downloading and installing a local copy through the state-wide license, follow the steps below.

NOTE: To run Altair HyperWorks, your computer must have access to the internet. The software contacts the license server at OSC to check out a license when it starts and periodically during execution. The amount of data transferred is small, so network connections over modems are acceptable.

Usage of HyperWorks on a local machine using the statewide license will vary from installation to installation.

Go to http://www.altairhyperworks.com/
Click on "Login" in the upper right hand corner of the page.
If you have already registered with the Altair web site, enter the e-mail address that you registered with and your password and skip to step #5.
If you have not registered yet, click the link that says "Click here to register now". You will be prompted for some contact information and an e-mail address which will be your unique identifier.
- IMPORTANT: The e-mail address you give must be from your academic institution. Under the statewide license agreement, registration from Ohio universities is allowed on the Altair web site. Trying to log in with a yahoo or hotmail e-mail account will not work. If you enter your university e-mail and the system will not register you, please contact OSChelp at oschelp@osc.edu.
Once you have logged in, click on "SUPPORT" and then "SOFTWARE DOWNLOADS"
In addition to downloading the software, download the "Installation Guide and Release Notes" for instructions on how to install the software.
- IMPORTANT: If you have any questions or problems, please contact OSChelp at oschelp@osc.edu, rather than HyperWorks support. The software agreements outlines that problems should first be sent to OSC. If the OSC support line cannot answer or resolve the question, they have the ability to raise the problem to Altair support.
Please contact OSC Help for further instruction and license server information. In order to be added to the allowed list for the state-wide software access, we will need your IP address/range of machine that will be running this software.
You need to set an environment variable (ALTAIR_LICENSE_PATH) on your local machine to point at our license server (7790@license6.osc.edu). See this link for instructions if necessary.

BLAS

The BLAS (Basic Linear Algebra Subprograms) are routines that provide standard building blocks for performing basic vector and matrix operations.

Availability and Restrictions

Access

A highly optimized implementation of the BLAS is available on all OSC clusters as part of the Intel Math Kernel Library (MKL). We recommend that you use MKL rather than building the BLAS for yourself. MKL is available to all OSC users.

Usage

See OSC's MKL software page for usage information. Note that there is no library named libblas.a or libblas.so. The flag "-lblas" on your link line will not work. You should modify your makefile or build script to link to the MKL libraries instead.

BLAST

The BLAST programs are widely used tools for searching DNA and protein databases for sequence similarity to identify homologs to a query sequence. While often referred to as just "BLAST", this can really be thought of as a set of programs: blastp, blastn, blastx, tblastn, and tblastx.

Availability & Restrictions

Versions

The following versions of BLAST are available on OSC systems:

Version	Oakley	Owens
2.2.24+	X
2.2.25+	X
2.2.26	X
2.2.31+	X
2.4.0+		X*
2.6.0+	X*

* Current Default Version

You can use module spider blast to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

If you need to use blastx, you will need to load one of the C++ implimenations modules of blast (any version with a "+").

Access

BLAST is available without restriction to all OSC users.

Usage

Set-up

To load BLAST, type the following into the command line:

module load blast

Then create a resource file .ncbirc, and put it under your home directory.

Using BLAST

The five flavors of BLAST mentioned above perform the following tasks:

blastp: compares an amino acid query sequence against a protein sequence database

blastn: compares a nucleotide query sequence against a nucleotide sequence database

blastx: compares the six-frame conceptual translation products of a nucleotide query sequence (both strands) against a protein sequence database

tblastn: compares a protein query sequence against a nucleotide sequence database dynamically translated in all six reading frames (both strands).

tblastx: compares the six-frame translations of a nucleotide query sequence against the six-frame translations of a nucleotide sequence database. (Due to the nature of tblastx, gapped alignments are not available with this option)

We provide local access to nt and refseq_protein databases. You can access the database using BLASTDB environment variable. If you need other databases, please send a request email to OSC Help .

Batch Usage

A sample batch script is below:

#PBS -l nodes=1:ppn=1
#PBS -l walltime=10:00
#PBS -N Blast
#PBS -S /bin/bash
#PBS -j oe
module load blast
set -x
cd $PBS_O_WORKDIR
mkdir $PBS_JOBID
cp 100.fasta $TMPDIR
cd $TMPDIR
/usr/bin/time blastn -db nt -query 100.fasta -out test.out
cp test.out $PBS_O_WORKDIR/$PBS_JOBID

NCBI BLAST Database

The NCBI BLAST database is updated on Oct 30th 2017 and is available at /fs/project/pub_data/blastDB-Oct2017/db

BLAT

BLAT is a sequence analysis tool which performs rapid mRNA/DNA and cross-species protein alignments. BLAT is more accurate and 500 times faster than popular existing tools for mRNA/DNA alignments and 50 times faster for protein alignments at sensitivity settings typically used when comparing vertebrate sequences.

BLAT is not BLAST. DNA BLAT works by keeping an index of the entire genome (but not the genome itself) in memory. Since the index takes up a bit less than a gigabyte of RAM, BLAT can deliver high performance on a reasonably priced Linux box. The index is used to find areas of probable homology, which are then loaded into memory for a detailed alignment. Protein BLAT works in a similar manner, except with 4-mers rather than 11-mers. The protein index takes a little more than 2 gigabytes.

Availability & Restrictions

Versions

The following versions of BLAT are available at OSC:

Version	Oakley
34	X*

* Current Default Version

You can use module spider blat to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

BLAT is available without restriction to all OSC users.

Usage

Set-up

To initalize the Oakley system prior to using BLAT, run the following commands:

module load blat

Using BLAT

The main programs in the blat suite are:

gfServer – a server that maintains an index of the genome in memory and uses the index to quickly find regions with high levels of sequence similarity to a query sequence.

gfClient – a program that queries gfServer over the network, and then does a detailed alignment of the query sequence with regions found by gfServer.

blat –combines client and server into a single program, first building the index, then using the index, and then exiting.

webBlat – a web based version of gfClient that presents the alignments in an interactive fashion. (not included on OSC server)

Building an index of the genome typically takes 10 or 15 minutes. Typically for interactive applications one uses gfServer to build a whole genome index. At that point gfClient or webBlat can align a single query within few seconds. If one is aligning a lot of sequences in a batch mode then blat can be more efficient, particularly if run on a cluster of computers. Each blat run is typically done against a single chromosome, but with a large number of query sequences.

Other programs in the blat suite are:

pslSort – combines and sorts the output of multiple blat runs.  (The blat default output format is .psl).

pslReps – selects the best alignments for a particular query sequence, using a ‘near best in genome’ approach.

pslPretty – converts alignments from the psl format, which is tab-delimited format and does not include the bases themselves, to a more readable alignment format.

faToTwoBit – convert Fasta format sequence files to a dense randomly-accessable  .2bit format that gfClient can use.

twoBitToFa – convert from the .2bit format back to fasta

faToNib – convert from Fasta to a somewhat less dense randomly accessible format that predates .2bit.  Note each .nib file can only contain a single sequence.

nibFrag – convert portions of a nib file back to fasta.

The command line options of each of the programs is described below. Similar summaries of usage are printed when a command is run with no arguments.

Batch Usage

A sample batch script is as below:

#PBS -N blat
#PBS -j oe
#PBS -l nodes=1:ppn=1
#PBS -S /bin/bash

module load blat 

cd $PBS_O_WORKDIR
cp $BLAT_TEST_DIR/gfServer-tests/input/creaGeno.2bit .
cp $BLAT_TEST_DIR/gfServer-tests/input/crea.mrna .

blat -stepSize=5 -repMatch=2253 -minScore=0 -minIdentity=0 creaGeno.2bit crea.mrna output.psl

BWA

BWA is a software package for mapping low-divergent sequences against a large reference genome, such as the human genome. It consists of three algorithms: BWA-backtrack, BWA-SW and BWA-MEM.

Availability and Restrictions

Versions

The following versions of BWA are available on OSC clusters:

Version	Oakley	owens
0.7.13 0.7.17	X*	X* X

Version

Oakley

owens

0.7.13

0.7.17

X*

X

* Current default version

You can use module spider bwa to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

BWA is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of BWA, run the following command: module load bwa. The default version will be loaded. To select a particular BWA version, use module load bwa/version. For example, use module load bwa/0.7.13 to load BWA 0.7.13.

Usage on Owens

Set-up

To configure your environment for use of BWA, run the following command: module load bwa. The default version will be loaded. To select a particular BWA version, use module load bwa/version. For example, use module load bwa/0.7.13 to load BWA 0.7.13.

BamTools

BamTools provides both a programmer's API and an end-user's toolkit for handling BAM files.

Availability and Restrictions

Versions

The following versions of BamTools are available on OSC clusters:

Version	Oakley	Owens
2.2.2	X	X*
2.4.0	X*

* Current default version

You can use module spider bamtools to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

BamTools is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of BamTools, run the following command: module load bamtools. The default version will be loaded. To select a particular BamTools version, use module load bamtools/version. For example, use module load bamtools/2.2.2 to load BamTools 2.2.2.

Usage on Owens

Set-up

To configure your environment for use of BamTools, run the following command: module load bamtools. The default version will be loaded. To select a particular BamTools version, use module load bamtools/version. For example, use module load bamtools/2.2.2 to load BamTools 2.2.2.

BioPerl

BioPerl offers a set of perl modules which can be used for sequence manipulation. Knowledge of PERL programming is required.

Availability & Restrictions

Versions

The following versions of BioPerl are available:

Version	Oakley
1.6.1	X*

* Current Default Version

You can use module spider bioperl to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

BioPerl is available without restriction to all OSC users.

Usage

Using BioPerl

This is an example of how to use bioperl and access the sequence package on the Oakley Cluster.

use lib '/usr/local/bioperl/1.6.1/lib/perl5/'
use Bio::Seq;

Boost

Boost is a set of C++ libraries that provide helpful data structures and numerous support functions in a wide range of aspects of programming, such as, image processing, gpu programming, concurrent programming, along with many algorithms. Boost is portable and performs well on a wide variety of platforms.

Availability & Restrictions

Versions

The following version of Boost are available on OSC systems:

Version	Oakley	Ruby	owens	Notes
1.41.0	System Install	System Install		No Module Needed
1.53.0			System Install	No Module Needed
1.56.0	X(GI)	X(GI)
1.58.0	X(G)
1.60.0	X(G)
1.63.0	X(GI)*	X(GI)*	X(GI)*
1.64.0	X(GI)	X(GI)	X(GI)

* Current default version; G = available with gnu; I = available with intel

You can use module spider boost to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Boost is available without restriction to all OSC Users.

Usage

Usage on Oakley

Set-up

Initalizing the system for use of the Boost library is independent of the compiler you are using. To load the boost module run the following command:

module load boost

Building With Boost

The following environment variables are setup when the Boost library is loaded:

Variable	Use
`$BOOST_CFLAGS`	Use during your compilation step for C++ programs.
`$BOOST_LIBS`	Use during your link step.

Below is a set of example commands used to build and run a file called example2.cpp. First copy the example2.cpp and jayne.txt with the following commands.

cp /usr/local/src/boost/boost-1.56.0/test.osc/example2.cpp .
cp /usr/local/src/boost/boost-1.56.0/test.osc/jayne.txt .
module load boost
g++ $BOOST_CFLAGS example2.cpp -o boostTest $BOOST_LIBS -lboost_regex
./boostTest < jayne.txt

Usage on Ruby

Set-up

Initalizing the system for use of the Boost library is independent of the compiler you are using. To load the boost module run the following command:

module load boost

Building With Boost

The following environment variables are setup when the Boost library is loaded:

VARIABLE	USE
`$BOOST_CFLAGS`	Use during your compilation step for C++ programs.
`$BOOST_LIBS`	Use during your link step.

Below is a set of example commands used to build and run a file called example2.cpp. First copy the example2.cpp and jayne.txt from Oakley into your home directory with the following commands:

cp /usr/local/src/boost/boost-1_56_0/test.osc/example2.cpp ~
cp /usr/local/src/boost/boost-1_56_0/test.osc/jayne.txt ~

Then compile and test the program with the folllowing commands:

module load boost
g++ $BOOST_CFLAGS example2.cpp -o boostTest $BOOST_LIBS -lboost_regex
./boostTest < jayne.txt

Usage on Owens

Set-up

Initalizing the system for use of the Boost library is independent of the compiler you are using. To load the boost module run the following command:

module load boost

Building With Boost

The following environment variables are setup when the Boost library is loaded:

VARIABLE	USE
`$BOOST_CFLAGS`	Use during your compilation step for C++ programs.
`$BOOST_LIBS`	Use during your link step.

Below is a set of example commands used to build and run a file called example2.cpp. First copy the example2.cpp and jayne.txt from Oakley into your home directory with the following commands:

cp /usr/local/src/boost/boost-1_56_0/test.osc/example2.cpp ~
cp /usr/local/src/boost/boost-1_56_0/test.osc/jayne.txt ~

Then compile and test the program with the folllowing commands:

g++ example2.cpp -o boostTest -lboost_regex
./boostTest < jayne.txt

Bowtie1

Bowtie1 is an ultrafast, memory-efficient short read aligner. It aligns short DNA sequences (reads) to the human genome at a rate of over 25 million 35-bp reads per hour. Bowtie indexes the genome with a Burrows-Wheeler index to keep its memory footprint small: typically about 2.2 GB for the human genome (2.9 GB for paired-end).

Availability and Restrictions

Versions

The following versions of Bowtie1 are available on OSC clusters:

Version	Oakley	Owens
1.1.2	X*	X*

* Current default version

You can use module spider bowtie1 to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Bowtie1 is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of Bowtie1, run the following command: module load bowtie1. The default version will be loaded. To select a particular Bowtie1 version, use module load bowtie/version. For example, use module load bowtie1/1.1.2 to load Bowtie1 1.1.2.

Usage on Owens

Set-up

To configure your environment for use of Bowtie1, run the following command: module load bowtie1. The default version will be loaded. To select a particular Bowtie1 version, use module load bowtie/version. For example, use module load bowtie1/1.1.2 to load Bowtie1 1.1.2.

Bowtie2

Bowtie2 is an ultrafast and memory-efficient tool for aligning sequencing reads to long reference sequences. It is particularly good at aligning reads of about 50 up to 100s or 1,000s of characters, and particularly good at aligning to relatively long (e.g. mammalian) genomes. Bowtie 2 indexes the genome with an FM Index to keep its memory footprint small: for the human genome, its memory footprint is typically around 3.2 GB. Bowtie 2 supports gapped, local, and paired-end alignment modes.

Please note that bowtie (and tophat) cannot run in parallel, that is, on multiple nodes. Submitting multi-node jobs will only waste resources. In addition you must explicitly include the '-p' option to use multiple threads on a single node.

Availability and Restrictions

Versions

The following versions of Bowtie2 are available on OSC clusters:

Version	Oakley	Owens
2.2.8	X*
2.2.9		X*

* Current default version

You can use module spider bowtie2 to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Bowtie2 is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of Bowtie2, run the following command: module load bowtie2. The default version will be loaded. To select a particular Bowtie2 version, use module load bowtie2/version. For example, use module load bowtie2/2.2.8 to load Bowtie2 2.2.8.

Usage on Owens

Set-up

To configure your environment for use of Bowtie2, run the following command: module load bowtie2. The default version will be loaded. To select a particular Bowtie2 version, use module load bowtie2/version. For example, use module load bowtie2/2.2.9 to load Bowtie2 2.2.9.

CDO

CDO (Climate Data Operators) is a collection of command line operator to manipulate and analyse climate and NWP model data. It is open source and released under the terms of the GNU General Public License v2 (GPL).

Availability and Restrictions

Versions

CDO is available on Oakley Cluster. The versions currently available at OSC are:

version	oakley
1.5.4	X
1.6.8	X*
1.7.0	X

* Current default Version

You can use module spider cdo to view available modules on Oakley. Feel free to contact OSC Help if you need other versions for your work.

Access

CDO is available for use by all OSC users.

Usage

Usage on Oakley

Set-up on Oakley

To load the default version of CDO module, use module load cdo. To select a particular software version, use module load cdo/version. For example, use module load cdo/1.5.4 to load CDO version 1.5.4 on Oakley.

Batch Usauge on Oakley

Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Oakley and Scheduling Policies and Limits for more info.

Interactive Batch Session

For an interactive batch session on Oakley, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Once your job is ready, you can start to use CDO. For example, use the following command to display general information about a data file named infile:

cdo info infile

Use cdo -h [operator] for operator descriptions.

COMSOL

COMSOL Multiphysics (formerly FEMLAB) is a finite element analysis and solver software package for various physics and engineering applications, especially coupled phenomena, or multiphysics. owned and supported by COMSOL, Inc.

Important Note: From March 14, 2018, Comsol is unavailable from Oakley for academic users. If you have any questions, please contact OSC Help.

Availability and Restrictions

Versions

COMSOL is available on the Oakley and Ruby clusters. The versions currently available at OSC are:

Version	Ruby	Owens
52	X*
52a		X*
53a		X

* Current default version

You can use module spider comsol to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

COMSOL is for academic use, available to all Ohio State University users. OSC does not provide COMSOL licenses for academic use to students and faculty outside of Ohio State University due to licensing restrictions. If you or your institution have a network COMSOL license server, you may be able to use it on OSC. For connections to your license server from OSC, please read this document. If you need further help, please contact OSC Help.

To use COMSOL you will have to be added to the license server. Please contact OSC Help to be added.

Access for Commerical Users

Contact OSC Help for getting access to COMSOL if you are a commercial user.

Usage

Usage on Ruby

Set-up

To load the default version of COMSOL module, use module load comsol. To select a particular software version, use module load comsol/version. For example, use module load comsol/52 to load COMSOL version 5.2.

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=1:00:00 -l software=comsolscript

which gives you 20 cores (-l nodes=1:ppn=20) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

Assume that you have had a comsol script file mycomsol.m in your working direcory ($PBS_O_WORKDIR). Below is the example batch script (job.txt) for a serial run:

#PBS -N COMSOL
#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=1
#PBS -j oe
#PBS -l software=comsolscript
#
# The following lines set up the COMSOL environment
#
module load comsol
#
# Move to the directory where the job was submitted
#
cd $PBS_O_WORKDIR
cp *.m $TMPDIR
cd $TMPDIR
#
# Run COMSOL
#
comsol batch mycomsol
#
# Now, copy data (or move) back once the simulation has completed
#
cp * $PBS_O_WORKDIR

Non-interactive Batch Job (Parallel Run for COMSOL 4.3 and Later)

As of version 4.3, it is not necessary to start up MPD before launching a COMSOL job. Below is the example batch script (job.txt) for a parallel run using COMSOL 4.3 or later versions:

#PBS -l walltime=01:00:00
#PBS -l nodes=2:ppn=20
#PBS -N COMSOL
#PBS -j oe
#PBS -r n
#PBS -l software=comsolscript
cd ${PBS_O_WORKDIR}
module load comsol
echo "--- Copy Input Files to TMPDIR and Change Disk to TMPDIR"
cp input_cluster.mph $TMPDIR
cd $TMPDIR
np=20
echo "--- Running on ${np} processes (cores) on the following nodes:"
cat $PBS_NODEFILE | uniq > hostfile
echo "--- COMSOL run"
comsol -nn 2 -np ${np} batch -f hostfile -mpirsh rsh -inputfile input_cluster.mph -outputfile output_cluster.mph
echo "--- Copy files back"
cp output_cluster.mph output_cluster.mph.status ${PBS_O_WORKDIR}
echo "---Job finished at: 'date'"
echo "---------------------------------------------"

Note:

Set nodes to 2 and ppn to 20 (-l nodes=2:ppn=20). You can change the values per your need.
Copy files from your directory to $TMPDIR.
Provide the name of the input file and output file.

Usage on Owens

Set-up

To load the default version of COMSOL module, use module load comsol. To select a particular software version, use module load comsol/version. For example, use module load comsol/52a to load COMSOL version 5.2a.

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00 -l software=comsolscript

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

Assume that you have had a comsol script file mycomsol.m in your working direcory ($PBS_O_WORKDIR). Below is the example batch script (job.txt) for a serial run:

#PBS -N COMSOL
#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=1
#PBS -j oe
#PBS -l software=comsolscript
#
# The following lines set up the COMSOL environment
#
module load comsol
#
# Move to the directory where the job was submitted
#
cd $PBS_O_WORKDIR
cp *.m $TMPDIR
cd $TMPDIR
#
# Run COMSOL
#
comsol batch mycomsol
#
# Now, copy data (or move) back once the simulation has completed
#
cp * $PBS_O_WORKDIR

Non-interactive Batch Job (Parallel Run for COMSOL 4.3 and Later)

As of version 4.3, it is not necessary to start up MPD before launching a COMSOL job. Below is the example batch script (job.txt) for a parallel run using COMSOL 4.3 or later versions:

#PBS -l walltime=01:00:00
#PBS -l nodes=2:ppn=28
#PBS -N COMSOL
#PBS -j oe
#PBS -r n
#PBS -l software=comsolscript
cd ${PBS_O_WORKDIR}
module load comsol
echo "--- Copy Input Files to TMPDIR and Change Disk to TMPDIR"
cp input_cluster.mph $TMPDIR
cd $TMPDIR
np=28
echo "--- Running on ${np} processes (cores) on the following nodes:"
cat $PBS_NODEFILE | uniq > hostfile
echo "--- COMSOL run"
comsol -nn 2 -np ${np} batch -f hostfile -mpirsh ssh -inputfile input_cluster.mph -outputfile output_cluster.mph
echo "--- Copy files back"
cp output_cluster.mph output_cluster.mph.status ${PBS_O_WORKDIR}
echo "---Job finished at: 'date'"
echo "---------------------------------------------"

Note:

Set nodes to 2 and ppn to 28 (-l nodes=2:ppn=28). You can change the values per your need.
Use "-mpirsh ssh" option for multi-node jobs
Copy files from your directory to $TMPDIR.
Provide the name of the input file and output file.

Interactive Parallel COMSOL Job

This documentation is to discuss how to set up an interactive parallel COMSOL job at OSC. The following example demonstrates the process of using COMSOL version 5.1 on Oakley. Depending on the version of COMSOL and cluster you work on, there mighe be some differences from the example. Feel free to contact OSC Help if you have any questions.

Launch COMSOL GUI application following the instructions on this page. Get the information on the node(s) allocated to your job and save it in the file named hostfile using the following command:

cat $PBS_NODEFILE | uniq > hostfile

Make sure the hostfile is located in the same directory where you COMSOL input file is put

Open COMSOL GUI application. To enable the cluster compuitng feature, click the show button and select Advanced Study Options, as shown in the picture below:

Advanced study

In the Model Builder, right-click a Study node and select Cluster Computing, as shown in the picture below:

cluster computing

In the Cluster Computing node's setting window, select General from the Cluster type list. Provide the name of Host file as hostfile. Browse to the directory where your COMSOL input file is located, as shown in the picture below:

setting

Save all the settings. Then you should be able to run an interactive parallel COMSOL job at OSC

Fields of Science:

Aerospace Engineering

Materials

Mechanical Engineering

CUDA

CUDA™ (Compute Unified Device Architecture) is a parallel computing platform and programming model developed by Nvidia that enables dramatic increases in computing performance by harnessing the power of the graphics processing unit (GPU).

Availability and Restrictions

Versions

CUDA is available on the clusters supporting GPUs. The versions currently available at OSC are:

Version	Oakley	Ruby	Owens	notes
4.1.28	X
5.0.35	X	X		Default version on Oakley prior to 09/15/2015
5.5	X
5.5.22		X
6.0.37	X	X
6.5.14	X*	X
7.0.28	X	X
7.5.18	X	X
8.0.44 8.0.61 9.1.85		X* X X	X* X X	GPU drivers prevent the use of CUDA 8 on Oakley

* Current default version

You can use module spider cuda to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

CUDA is available for use by all OSC users.

Usage

Usage on Oakley

Set-up on Oakley

To load the default version of CUDA module, use module load cuda. To select a particular software version, use module load cuda/version. For example, use module load cuda/7.0.28 to load CUDA version 7.0.28 on Oakley.

GPU Computing SDK

The NVIDIA GPU Computing SDK provides hundreds of code samples and covers a wide range of applications/techniques to help you get started on the path of writing software with CUDA C/C++ or DirectCompute. On Oakley, the SDK binaries are located in $CUDA_HOME/bin/linux/release ($CUDA_HOME is an environment variable set when you load the module).

Programming in CUDA

Please visit the following link to learn programming in CUDA, http://developer.nvidia.com/cuda-education-training. The link also contains tutorials on Optimizing CUDA codes to obtain greater SpeedUp. One can also refer to the following webpage for some more CUDA optimization techniques, http://www.cs.berkeley.edu/~volkov/

Compiling CUDA Code

One can type module show cuda-version-number to view the list of environment variables.
To compile a CUDA C code contained in a file, let say mycudaApp.cu , the following could be done after loading the appropriate CUDA module: nvcc -o mycudaApp mycudaApp.cu. This will create an executable by name mycudaApp.

Important: The devices are configured in exclusive mode. This means that 'cudaSetDevice' should NOT be used if requesting one GPU resource. Once the first call to CUDA is executed, the system will figure out which device it is using. If both cards per node is in use by a single application, please use 'cudaSetDevice'.

Debugging CUDA code

cuda-gdb can be used to debug CUDA codes. module load cuda will make it available to you. For more information on how to use the CUDA-GDB please visit http://developer.nvidia.com/cuda-gdb.

Detecting memory access errors

CUDA-MEMCHECK could be used for detecting the source and cause of memory access errors in your program. For more information on how to use CUDA-MEMCHECK please visit http://developer.nvidia.com/cuda-memcheck.

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a Linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=1:gpus=1 -l walltime=00:20:00

which requests one core (-l nodes=1:ppn=1), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Please note that on Oakley, you can request any mix of ppn and gpus you need; please see the Batch Limit Rules and Job Scripts page in our batch guide for more information.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=01:00:00
#PBS -l nodes=1:ppn=1:gpus=1
#PBS -N compute
#PBS -j oe
module load cuda
cd $HOME/cuda
cp mycudaApp $TMPDIR
cd $TMPDIR
./mycudaApp

Usage on Ruby

Set-up on Ruby

To load the default version of CUDA module, use module load cuda. To select a particular software version, use module load cuda/version. For example, use module load cuda/7.0.28 to load CUDA version 7.0.28 on Ruby.

GPU Computing SDK

The NVIDIA GPU Computing SDK provides hundreds of code samples and covers a wide range of applications/techniques to help you get started on the path of writing software with CUDA C/C++ or DirectCompute.

Programming in CUDA

Please visit the following link to learn programming in CUDA, http://developer.nvidia.com/cuda-education-training. The link also contains tutorials on Optimizing CUDA codes to obtain greater SpeedUp. One can also refer to the following webpage for some more CUDA optimization techniques, http://www.cs.berkeley.edu/~volkov/

Compiling CUDA Code

One can type module show cuda-version-number to view the list of environment variables.
To compile a cuda code contained in a file, let say mycudaApp.cu, the following could be done after loading the appropriate CUDA module: nvcc -o mycudaApp mycudaApp.cu. This will create an executable by name mycudaApp.

Important: The devices are configured in exclusive mode. This means that 'cudaSetDevice' should NOT be used if requesting one GPU resource. Once the first call to CUDA is executed, the system will figure out which device it is using. If both cards per node is in use by a single application, please use 'cudaSetDevice'.

Debugging CUDA code

cuda-gdb can be used to debug CUDA codes. module load cuda will make it available to you. For more information on how to use the CUDA-GDB please visit http://developer.nvidia.com/cuda-gdb.

Detecting memory access errors

CUDA-MEMCHECK could be used for detecting the source and cause of memory access errors in your program. For more information on how to use CUDA-MEMCHECK please visit http://developer.nvidia.com/cuda-memcheck.

Setting the GPU compute mode on Ruby

The GPUs on Ruby can be set to different compute modes as listed here.

They can be set by adding the following to the GPU specification:

-l nodes=1:ppn=20:gpus=1:default
-l nodes=1:ppn=20:gpus=1:exclusive
-l nodes=1:ppn=20:gpus=1:exclusive_process

Note, the prohibited mode is not an option.

Batch Usage on Ruby

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=20:gpus=1 -l walltime=00:20:00

which requests one whole node with 20 cores (-l nodes=1:ppn=20), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=01:00:00
#PBS -l nodes=1:ppn=20:gpus=1
#PBS -N compute
#PBS -j oe
module load cuda
cd $HOME/cuda
cp mycudaApp $TMPDIR
cd $TMPDIR
./mycudaApp

Caffe

Caffe is "a fast open framework for deep learning."

From their README:

Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by the Berkeley Vision and Learning Center (BVLC) and by community contributors. Yangqing Jia created  the project during his PhD at UC Berkeley. Caffe is released under the BSD 2-Clause license.

Caffe also includes interfaces for both Python and Matlab, which have been built but have not been tested.

Availability and Restrictions

Versions

The following versions of Caffe are available on OSC clusters:

version	Ruby	Owens
1.0.0-rc3	X*	X*

* Current Default Version

You can use module spider caffe to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

The current versions of Caffe on Ruby require the use of python/2.7.8, cuda/7.5.18, boost/1.56.0, opencv/3.1.0 on Ruby for GPU calculations.

The current version of Caffe on Owens requires cuda/8.0.44 for GPU calculations.

Access

Caffe is available to all OSC users without restriction.

Usage

Usage on Ruby or Owens

Setup on Ruby or Owens

To configure the Ruby or Owens cluster for the use of Caffe, use the following commands:

module load caffe

Batch Usage on Ruby or Owens

Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Ruby, and Scheduling Policies and Limits for more info. In particular, Caffe should be run on a GPU-enabled compute node.

An Example of Using Caffe with MNIST Training Data on Ruby

Below is an example batch script (job.txt) for using Caffe, see this resource for a detailed explanation http://caffe.berkeleyvision.org/gathered/examples/mnist.html

#PBS -N Caffe
#PBS -l nodes=1:ppn=20:gpus=1:default
#PBS -l walltime=30:00
#PBS -j oe
#PBS -S /bin/bash 
cd $PBS_O_WORKDIR
. /etc/profile.d/lmod.sh
# Load the modules for Caffe
ml caffe
# Migrate to job temp directory and copy folders over
cd $TMPDIR
cp -r $CAFFE_HOME/{examples,build,data} .
# Download, create, train
./data/mnist/get_mnist.sh
./examples/mnist/create_mnist.sh
./examples/mnist/train_lenet.sh
# Serialize log files
echo; echo 'LOG 1'
cat convert_mnist_data.bin.$(hostname)*
echo; echo 'LOG 2'
cat caffe.INFO
echo; echo 'LOG 3'
cat convert_mnist_data.bin.INFO
echo; echo 'LOG 4'
cat caffe.$(hostname).*
cp examples/mnist/lenet_iter_10000.* $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

An Example of Using Caffe with MNIST Training Data on Owens

Below is an example batch script (job.txt) for using Caffe, see this resource for a detailed explanation http://caffe.berkeleyvision.org/gathered/examples/mnist.html

#!/bin/bash
#PBS -N Caffe
#PBS -l nodes=1:ppn=28:gpu
#PBS -l walltime=30:00
#PBS -j oe
#PBS -S /bin/bash 
cd $PBS_O_WORKDIR
. /etc/profile.d/lmod.sh
# Load the modules for Caffe
ml caffe
# Migrate to job temp directory and copy folders over
cd $TMPDIR
cp -r $CAFFE_HOME/{examples,data} .
# Download, create, train
./data/mnist/get_mnist.sh
./examples/mnist/create_mnist.sh
./examples/mnist/train_lenet.sh
# Serialize log files
echo; echo 'LOG 1'
cat convert_mnist_data.bin.$(hostname)*
echo; echo 'LOG 2'
cat caffe.INFO
echo; echo 'LOG 3'
cat convert_mnist_data.bin.INFO
echo; echo 'LOG 4'
cat caffe.$(hostname).*
cp examples/mnist/lenet_iter_10000.* $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Cairo

Cairo is a 2D graphics library with support for multiple output devices. It is designed to produce consistent output on all output media while taking advantage of display hardware acceleration when available.

Availability and Restrictions

Versions

The following versions of Cairo are available on OSC clusters:

VERSION	OAKLEY	Owens
1.12.18	X(GI)
1.14.2	X(GI)*	X(GI)*

* Current default version; G = available with gnu; I = available with intel

You can use module spider cairo to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Cairo is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of Cairo, run the following command: module load cairo. The default version will be loaded. To select a particular Cairo version, use module load cairo/version. For example, use module load cairo/1.14.2 to load Cairo 1.14.2.

Building with Cairo

The Cairo library provides the following variables for use at build time:

Variable	Use
`$CAIRO_CFLAGS`	Use during your compilation step for programs
`$CAIRO_LIBS`	Use during your linking step programs

For example, to build the code myprog.c with the Cairo library you would use:

icc -o myprog myprog.c $CAIRO_LIBS $CAIRO_CFLAGS

Usage on Owens

Set-up

To configure your environment for use of Cairo, run the following command: module load cairo. The default version will be loaded. To select a particular Cairo version, use module load cairo/version. For example, use module load cairo/1.14.2 to load Cairo 1.14.2.

Building with Cairo

The Cairo library provides the following variables for use at build time:

VARIABLE	USE
`$CAIRO_CFLAGS`	Use during your compilation step for programs
`$CAIRO_LIBS`	Use during your linking step programs (Note: Dynamic linking only)

For example, to build the code myprog.c with the Cairo library you would use:

icc -o myprog myprog.c -lcairo

Clustal W

Clustal W is a general purpose multiple sequence alignment program for DNA or proteins.It produces biologically meaningful multiple sequence alignments of divergent sequences. It calculates the best match for the selected sequences, and lines them up so that the identities, similarities and differences can be seen.

Availability and Restrictions

Versions

The following versions of Clustal W are available on OSC systems:

Version	Oakley
2.1	X*

* Current default version

You can use module spider clustalw to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Clustal W is available without restriction to all OSC users.

Usage

Usage on Oakley

Set-up

On the Oakley system load the clustalw module:

module load clustalw

Using Clustal W

Once the clustalw module has been loaded, the commands are available for your use. On the Oakley system, the command is:

clustalw2

The options can be listed interactively by typing clustalw -help or clustalw -check on the command-line.

DATA (sequences)
-INFILE=file.ext                             :input sequences.
-PROFILE1=file.ext  and  -PROFILE2=file.ext  :profiles (old alignment).
VERBS (do things)
-OPTIONS            :list the command line parameters
-HELP  or -CHECK    :outline the command line params.
-ALIGN              :do full multiple alignment 
-TREE               :calculate NJ tree.
-BOOTSTRAP(=n)      :bootstrap a NJ tree (n= number of bootstraps; def. = 1000).
-CONVERT            :output the input sequences in a different file format.
PARAMETERS (set things)
***General settings:****
-INTERACTIVE :read command line, then enter normal interactive menus
-QUICKTREE   :use FAST algorithm for the alignment guide tree
-TYPE=       :PROTEIN or DNA sequences
-NEGATIVE    :protein alignment with negative values in matrix
-OUTFILE=    :sequence alignment file name
-OUTPUT=     :GCG, GDE, PHYLIP, PIR or NEXUS
-OUTORDER=   :INPUT or ALIGNED
-CASE        :LOWER or UPPER (for GDE output only)
-SEQNOS=     :OFF or ON (for Clustal output only)
-SEQNO_RANGE=:OFF or ON (NEW: for all output formats) 
-RANGE=m,n   :sequence range to write starting m to m+n. 
***Fast Pairwise Alignments:***
-KTUPLE=n    :word size
-TOPDIAGS=n  :number of best diags.
-WINDOW=n    :window around best diags.
-PAIRGAP=n   :gap penalty
-SCORE       :PERCENT or ABSOLUTE
***Slow Pairwise Alignments:***
-PWMATRIX=    :Protein weight matrix=BLOSUM, PAM, GONNET, ID or filename
-PWDNAMATRIX= :DNA weight matrix=IUB, CLUSTALW or filename
-PWGAPOPEN=f  :gap opening penalty        
-PWGAPEXT=f   :gap opening penalty
***Multiple Alignments:***
-NEWTREE=      :file for new guide tree
-USETREE=      :file for old guide tree
-MATRIX=       :Protein weight matrix=BLOSUM, PAM, GONNET, ID or filename
-DNAMATRIX=    :DNA weight matrix=IUB, CLUSTALW or filename
-GAPOPEN=f     :gap opening penalty        
-GAPEXT=f      :gap extension penalty
-ENDGAPS       :no end gap separation pen. 
-GAPDIST=n     :gap separation pen. range
-NOPGAP        :residue-specific gaps off  
-NOHGAP        :hydrophilic gaps off
-HGAPRESIDUES= :list hydrophilic res.    
-MAXDIV=n      :% ident. for delay
-TYPE=         :PROTEIN or DNA
-TRANSWEIGHT=f :transitions weighting
***Profile Alignments:***
-PROFILE      :Merge two alignments by profile alignment
-NEWTREE1=    :file for new guide tree for profile1
-NEWTREE2=    :file for new guide tree for profile2
-USETREE1=    :file for old guide tree for profile1
-USETREE2=    :file for old guide tree for profile2
***Sequence to Profile Alignments:***
-SEQUENCES   :Sequentially add profile2 sequences to profile1 alignment
-NEWTREE=    :file for new guide tree
-USETREE=    :file for old guide tree
***Structure Alignments:***
-NOSECSTR1     :do not use secondary structure-gap penalty mask for profile 1 
-NOSECSTR2     :do not use secondary structure-gap penalty mask for profile 2
-SECSTROUT=STRUCTURE or MASK or BOTH or NONE   :output in alignment file
-HELIXGAP=n    :gap penalty for helix core residues 
-STRANDGAP=n   :gap penalty for strand core residues
-LOOPGAP=n     :gap penalty for loop regions
-TERMINALGAP=n :gap penalty for structure termini
-HELIXENDIN=n  :number of residues inside helix to be treated as terminal
-HELIXENDOUT=n :number of residues outside helix to be treated as terminal
-STRANDENDIN=n :number of residues inside strand to be treated as terminal
-STRANDENDOUT=n:number of residues outside strand to be treated as terminal 
***Trees:***
-OUTPUTTREE=nj OR phylip OR dist OR nexus
-SEED=n        :seed number for bootstraps.
-KIMURA        :use Kimura's correction.   
-TOSSGAPS      :ignore positions with gaps.
-BOOTLABELS=node OR branch :position of bootstrap values in tree display

Batch Usage

Sample batch script for the Oakley system:

#PBS -N clustalw
#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=1
#PBS -j oe
cd $PBS_O_WORKDIR
module load clustalw
clustalw2 -INFILE=myfile.seqs -GAPOPEN=2 -GAPEXT=4

Cufflinks

Cufflinks is a program that analyzes RNA -Seq samples. It assembles aligned RNA-Seq reads into a set of transcripts, then inspects the transcripts to estimate abundances and test for differential expression and regulation in the RNA-Seq reads.

Availability and Restrictions

Versions

Cufflinks is available on the Owens Cluster. The versions currently available at OSC are:

VERSION	OWENS
2.2.1	X*

* Current Default Version

You can use module spider cufflinks to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Cufflinks is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of Cufflinks, use command module load cufflinks. This will load the default version.

Darshan

Darshan is a lightweight "scalable HPC I/O characterization tool". It is intended to profile I/O by emitting log files to a consistent log location for systems administrators, and also provides scripts to create summary PDFs to characterize I/O in MPI-based programs.

Availability and Restrictions

Versions

The following versions of Darshan are available on OSC clusters:

version	Owens
3.1.2	X*
3.1.4	X
3.1.5	X

* Current default version

You can use module spider darshan to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Darshan is available to all OSC users without restriction.

Usage

Usage on Owens

Setup on Owens

To configure the Owens cluster for Darshan use the following commands:

module load darshan

Darshan is only supported for the following compiler and MPI implementations:

gnu/4.8.5  mvapich2/2.2
gnu/4.8.5  mvapich2/2.2rc1
gnu/4.8.5  openmpi/1.10
intel/16.0.3  intelmpi/5.1.3
intel/16.0.3  mvapich2/2.2
intel/16.0.3  mvapich2/2.2rc1
intel/16.0.3  openmpi/1.10

Batch Usage on Owens

Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Owens, and Scheduling Policies and Limits for more info.

If you have an MPI-based program the syntax is as simple as

# basic call to darshan 
mpiexec.darshan [args] ./my_mpi_program
# to show evidence that Darshan is working and to see internal timing
mpiexec.darshan.timing [args] ./my_mpi_program

An Example of Using Darshan with MPI-IO

Below is an example batch script (mpiio_with_darshan.qsub) for understanding MPI-IO, see this resource for a detailed explanation: http://beige.ucs.indiana.edu/I590/node29.html. The C program examples have each MPI task write to the same file at different offset locations sequentially. A serial version (1 processor writes to 1 file) is included and timed for comparison. Because the files generated here are large scratch files there is no need to retain them.

Load the Darshan module
Create a function as a build and run harness for each case
Run the cases
Generate PDF reports (using pdflatex and supporting files) from binary log files.
Check job output and read PDF reports.

#!/bin/bash
#PBS -l nodes=1:ppn=28:pfsdir
#PBS -j oe
# one may need to perform 'shopt -s expand_aliases' inside the shell before calling this script to expand darshan aliases
shopt -s expand_aliases
ml r --quiet
module load darshan
ml
function run_darshan() {
COMPILER=mpicc
MPIPROCS=$1
PROGNAME=$2
BLOKSIZE=$3
LOCATION=$4
cp ${DARSHAN_EXAMPLE}/${PROGNAME}.c $LOCATION
cd $LOCATION
$COMPILER -DDEBUG -D_FILE_OFFSET_BITS=64 -D_LARGEFILE64_SOURCE -o $PROGNAME ${PROGNAME}.c
mpiexec_darshan() { eval mpiexec.darshan "$@"; }
mpiexec_darshan -n $MPIPROCS ./$PROGNAME -f test -l $BLOKSIZE
du -sh test # show the human-readable file size
rm test # we are not keeping the large file around
}
# run the parallel version of mpi-io to write to a file but use the local tmp directory
run_darshan $PBS_NP mkrandpfile 256 $TMPDIR
# run the parallel version of mpi-io to write to a file but use the parallel file system scratch location
run_darshan $PBS_NP mkrandpfile 256 $PFSDIR
# for each darshan log generate a PDF report
JOBID=$([[ $tmp =~ ([0-9]*)\..* ]]; echo "${BASH_REMATCH[1]}")
for log in $(ls $DARSHAN_LOG/$(date "+%Y/%-m/%-d")/${USER:0:8}_*_id${JOBID}*.darshan); do
darshan-job-summary.pl $log
done
cp *.pdf $PBS_O_WORKDIR

In order to run it via the batch system, submit the mpiio_with_darshan.qsub file with the following command:

qsub mpiio_with_darshan.qsub

EMBOSS

EMBOSS is "The European Molecular Biology Open Software Suite". EMBOSS is a free Open Source software analysis package specially developed for the needs of the molecular biology (e.g. EMBnet) user community. The software automatically copes with data in a variety of formats and even allows transparent retrieval of sequence data from the web. Also, as extensive libraries are provided with the package, it is a platform to allow other scientists to develop and release software in true open source spirit. EMBOSS also integrates a range of currently available packages and tools for sequence analysis into a seamless whole.

Within EMBOSS you will find around hundreds of programs (applications) covering areas such as:

Sequence alignment,
Rapid database searching with sequence patterns,
Protein motif identification, including domain analysis,
Nucleotide sequence pattern analysis---for example to identify CpG islands or repeats,
Codon usage analysis for small genomes,
Rapid identification of sequence patterns in large scale sequence sets,
Presentation tools for publication

Availability and Restrictions

Versions

The following versions of EMBOSS are available on OSC systems:

Version	Oakley
6.4.0	X*

* Current default version

You can use module spider emboss to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

The EMBOSS software package is available without restriction for all academic OSC users.

Usage

Usage on Oakley

Set-up

For the Oakley system, run the following command:

module load emboss

Using EMBOSS

The EMBOSS programs are typically used to perform one or more tasks on a large number of sequences. Once the emboss module is loaded, all emboss commands are available for your use.

Batch Usage

Coming soon.

Elmer

Elmer is an open source multiphysical simulation software developed by CSC - IT (Finland).

From their Github README:

Elmer is a finite element software for numerical solution of partial
differential equations. Elmer is capable of handling any number of equations
and is therefore ideally suited for the simulation of multiphysical problems.
It includes models, for example, of structural mechanics, fluid dynamics, heat
transfer and electromagnetics. Users can also write their own equations that
can be dynamically linked with the main program.
Elmer consists of several parts. The most important ones are ElmerSolver, the
finite element solver, ElmerGUI, the graphical user interface, and ElmerGrid,
the mesh creation and manipulation tool. Also a visualization tool, ElmerPost,
is included in the package but it is no longer developed.  
Elmer software is licensed under GPL except for the ElmerSolver library which
is licensed under LGPL license. 
Elmer is mainly developed at CSC - IT Center for Science, Finland. However,
there have been numerous contributions from other organizations and developers
as well, and the project is open for new contributions.

Availability and Restrictions

Versions

The following versions of Elmer are available on OSC clusters:

version	oakley	Ruby
8.1	X*	X*

* Current default version

You can use module spider elmer to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

The current versions of Elmer are compiled with GNU 4.8.4 (requirement > 4.8) and BLAS libraries from MKL/11.2.3.

Access

Elmer is available to all OSC users without restriction.

Usage

Usage on Oakley or Ruby

Setup on Oakley or Ruby

To configure the Oakley or Ruby cluster for the use of Elmer 8.1, use the following commands:

module load gnu/4.8.4
module load mkl/11.2.3
module load elmer

Batch Usage on Oakley or Ruby

Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Oakley, Queues and Reservations for Ruby, and Scheduling Policies and Limits for more info.

Converting Mesh with ElmerGrid and Running ElmerSolver

Below is an example batch script (job.txt) for using Elmer:

#PBS -N Elmer
#PBS -l nodes=1:ppn=1 
#PBS -l walltime=30:00
#PBS -j oe 
#PBS -S /bin/bash 
# Load the modules for Elmer
module load gnu/4.8.4
module load mkl/11.2.3
module load elmer
# Move to the problem directory
cd $PBS_O_WORKDIR
# Copy files to $TMPDIR and move there to execute the program
cp -r * $TMPDIR
cd $TMPDIR
# Convert the mesh - see Elmer webpage for more examples of 'ElmerGrid'
ElmerGrid 14 2 beam.msh -autoclean
# Run the solver 
ElmerSolver beam.sif
# Finally, copy files back to your home directory 
cp -r * $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

FASTX-Toolkit

The FASTX-Toolkit is a collection of command line tools for Short-Reads FASTA/FASTQ files preprocessing.

Availability and Restrictions

Verisons

The following versions of FASTX-Toolkit are available on OSC clusters:

Version	Oakley	Owens
0.0.14	X*	X*

*:Current default version

You can use module spider fastx to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

FASTX-Toolkit is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of FASTX-Toolkit, run the following command: module load fastx. The default version will be loaded. To select a particular FASTX-Toolkit version, use module load fastx/version. For example, use module load fastx/0.0.14 to load FASTX-Toolkit 0.0.14.

Usage on Owens

Set-up

To configure your environment for use of FASTX-Toolkit, run the following command: module load fastx. The default version will be loaded. To select a particular FASTX-Toolkit version, use module load fastx/version. For example, use module load fastx/0.0.14 to load FASTX-Toolkit 0.0.14.

FFTW

FFTW is a C subroutine library for computing the Discrete Fourier Transform (DFT) in one or more dimensions, of arbitrary input size, and of both real and complex data. It is portable and performs well on a wide variety of platforms.

Availability and Restrictions

Versions

FFTW is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	OAKLEY	RUBY	Owens	notes
2.1	X
2.1.5		X(GI)
3.1.2
3.3	X			Default version on Oakley prior to 09/15/2015
3.3.1
3.3.2
3.3.3	X(I)
3.3.4	X	X	X(GI)
3.3.5	X*	X*	X*

* Current Default Version; G = available with only gnu; I = available with only intel

NOTE: FFTW2 and FFTW3 are tracked separately in the module system

You can use module spider fftw2 or module spider fftw3 to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

FFTW is available without restriction to all OSC Users.

Usage

Usage on Oakley

Set-up

Initalizing the system for use of the FFTW library is dependent on the system you are using and the compiler you are using. A successful build of your program will depend on an understanding of what module fits your circumstances. To load a particular version, use module load name. For example, use module load fftw3/3.3.3 to load FFTW3 version 3.3.3. You can use module spider fftw to view available modules.

Building with FFTW

The following environment variables are setup when the FFTW library is loaded:

Variable	Use
`$FFTW3_CFLAGS`	Use during your compilation step for C programs.
`$FFTW3_FFLAGS`	Use during your compilation step for Fortran programs.
`$FFTW3_LIBS`	Use during your link step for the sequential version of the library.
`$FFTW3_LIBS_OMP`	Use during your link step for the OpenMP version of the library.
`$FFTW3_LIBS_MPI`	Use during your link step for the MPI version of the library.

below is a set of example commands used to build a file called my-fftw.c.

module load fftw3
icc $FFTW3_CFLAGS my-fftw.c -o my-fftw $FFTW3_LIBS 
ifort $FFTW3_FFLAGS more-fftw.f -o more-fftw $FFTW3_LIBS

Usage on Ruby

Set-up

Initalizing the system for use of the FFTW library is dependent on the system you are using and the compiler you are using. A successful build of your program will depend on an understanding of what module fits your circumstances. To load a particular version, use module load name. For example, use module load fftw3/3.3.4 to load FFTW3 version 3.3.4. You can use module spider fftw to view available modules.

Building with FFTW

The following environment variables are setup when the FFTW library is loaded:

Variable	Use
`$FFTW3_CFLAGS`	Use during your compilation step for C programs.
`$FFTW3_FFLAGS`	Use during your compilation step for Fortran programs.
`$FFTW3_LIBS`	Use during your link step for the sequential version of the library.
`$FFTW3_LIBS_OMP`	Use during your link step for the OpenMP version of the library.
`$FFTW3_LIBS_MPI`	Use during your link step for the MPI version of the library.

below is a set of example commands used to build a file called my-fftw.c.

module load fftw3
icc $FFTW3_CFLAGS my-fftw.c -o my-fftw $FFTW3_LIBS 
ifort $FFTW3_FFLAGS more-fftw.f -o more-fftw $FFTW3_LIBS

Usage on Owens

Set-up

Initalizing the system for use of the FFTW library is dependent on the system you are using and the compiler you are using. A successful build of your program will depend on an understanding of what module fits your circumstances. To load a particular version, use module load name. For example, use module load fftw3/3.3.4 to load FFTW3 version 3.3.4. You can use module spider fftw to view available modules.

Building with FFTW

The following environment variables are setup when the FFTW library is loaded:

Variable	Use
`$FFTW3_CFLAGS`	Use during your compilation step for C programs.
`$FFTW3_FFLAGS`	Use during your compilation step for Fortran programs.
`$FFTW3_LIBS`	Use during your link step for the sequential version of the library.
`$FFTW3_LIBS_OMP`	Use during your link step for the OpenMP version of the library.
`$FFTW3_LIBS_MPI`	Use during your link step for the MPI version of the library.
`$FFTW3_LIBS_THREADS`	Use during your link step for the "threads" version of the library.

below is a set of example commands used to build a file called my-fftw.c.

module load fftw3
icc $FFTW3_CFLAGS my-fftw.c -o my-fftw $FFTW3_LIBS 
ifort $FFTW3_FFLAGS more-fftw.f -o more-fftw $FFTW3_LIBS

FSL

FSL is a library of tools for analyzing FMRI, MRI and DTI brain imaging data.

Availability and Restrictions

Versions

The following versions of FSL are available on OSC clusters:

VERSION	OWENS
5.0.10	X*

* Curent default version

You can use module spider fsl to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

FSL is available to all academic OSC users with proper validation. In order to obtain validation, please contact OSC Help for further instruction.

Usage

Usage on Owens

Set-up

Configure your environment for use of FSL with module load fsl. This will load the default version.

Using FSL GUI

Access the FSL GUI with command

fsl

This will bring up a menu of all FSL tools. For information on individual FSL tools see FSL Overview page.

FastQC

FastQC provides quality control checks of high throughput sequence data that identify areas of the data that may cause problems during further analysis.

Availability and Restrictions

Versions

FastQC is available on the Owens cluster. The versions currently available at OSC are:

VERSION	OWENS
0.11.5	X*

* Current Default Version

You can use module spider fastqc to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

FastQC is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of FastQC, use command module load fastqc. This will load the default version.

FreeSurfer

FreeSurfer is a software package used to anaylze nueroimaging data.

Availability & Restrictions

Versions

The following versions of FreeSurfer are available on OSC clusters:

VERSION	OWENS
5.3.0	X
6.0.0	X*

* Curent default version

You can use module spider freesurfer to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

FreeSurfer is available to all academic OSC users with proper validation. In order to obtain validation, please contact OSC Help for further instruction.

Usage

Usage on Owens

Set-up

Load the FreeSurfer module with module load freesurfer. This will load the default version. Then, to continue configuring your environment, you must source the setup script for Freesurfer. Do this with the following command that corresponds to the Linux shell you are using. If using bash, use:

source $FREESURFER_HOME/SetUpFreeSurfer.sh

If using tcsh, use:

source $FREESURFER_HOME/SetUpFreeSurfer.csh

To finish configuring FreeSurfer, set the the FreeSurfer environment variable SUBJECTS_DIR to the directory of your subject data. The SUBJECTS_DIR variable defaults to the FREESURFER_HOME/subjects directory, so if this is your intended directory to use the enviornment set-up is complete.

To alter the SUBJECTS_DIR variable, however, again use the following command that corresponds to the Linux shell you are using. For bash:

export SUBJECTS_DIR=<path to subject data>

For tcsh:

setenv SUBJECTS_DIR=<path to subject data>

Note that you can set the SUBJECT_DIR variable before or after sourcing the setup script.

FreeType

FreeType is a freely available software library to render fonts. It is written in C, designed to be small, efficient, highly customizable, and portable while capable of producing high-quality output (glyph images) of most vector and bitmap font formats.

Availability and Restrictions

Versions

The following versions of FreeType are available on OSC clusters:

VERSION	OAKLEY
2.5.5	X*

* Current default Version

You can use module spider freetype to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

FreeType is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

To configure your environment for use of FreeType on Oakley, run the following command: module load freetype. The default version will be loaded. To select a particular FreeType version, use module load freetype/version. For example, use module load freetype/2.5.5 to load FreeType 2.5.5 on Oakley.

GAMESS

The General Atomic and Molecular Electronic Structure System (GAMESS) is a flexible ab initio electronic structure program. Its latest version can perform general valence bond, multiconfiguration self-consistent field, Möller-Plesset, coupled-cluster, and configuration interaction calculations. Geometry optimizations, vibrational frequencies, thermodynamic properties, and solution modeling are available. It performs well on open shell and excited state systems and can model relativistic effects. The GAMESS Home Page has additional information.

Availability and Restrictions

Versions

The current versions of GAMESS available on the Oakley and Owens Clusters are:

VERSION	OAKLEY	owens
1 MAY 2013 (R1)	X*
18 AUG 2016 (R1)		X*

* Current default version

You can use module spider gamess to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

GAMESS is available to all OSC users.

Usage

Set-up

GAMESS usage is controlled via modules. Load one of the GAMESS modulefiles at the command line, in your shell initialization script, or in your batch scripts, for example:

module load gamess

Examples

Test input files are in the tests subdirectory
Computational Chemistry Workshop materials are in ~srb/workshops/compchem/gamess

GATK

GATK is a software package for analysis of high-throughput sequencing data. The toolkit offers a wide variety of tools, with a primary focus on variant discovery and genotyping as well as strong emphasis on data quality assurance.

Availability and Restrictions

Versions

The following versions of GATK are available on OSC clusters:

Version	Oakley	Owens
3.5	X*	X*

* Current default version

You can use module spider gatk to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

GATK is available to academic OSC users with proper validation. In order to obtain validation, please contact OSC Help for further instruction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of GATK, run the following command: module load gatk. The default version will be loaded. To select a particular GATK version, use module load gatk/version. For example, use module load gatk/3.5 to load GATK 3.5.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable. From module load gatk, a new environment variable, GATK, will be set. Thus, users can use the software by running the following command: java -jar $GATK {other options}.

Usage on Owens

Set-up

To configure your environment for use of GATK, run the following command: module load gatk. The default version will be loaded. To select a particular GATK version, use module load gatk/version. For example, use module load gatk/3.5 to load GATK 3.5.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable. From module load gatk, a new environment variable, GATK, will be set. Thus, users can use the software by running the following command: java -jar $GATK {other options}.

GLPK

GLPK (GNU Linear Programming Kit) is a set of open source LP (linear programming) and MIP (mixed integer problem) routines written in ANSI C, which can be called from within C programs.

Availability and Restrictions

Versions

The following versions are available on OSC systems:

version	oWENS
4.60	X*

* Current default version

You can use module spider glpk to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

GLPK is available to all OSC users without restriction.

Usage

Set-up

To set up your environment for using GLPK on Oakley, run the following command:

module load glpk

Compiling and Linking

To compile your C code using GLPK API routines, use the environment variable $GLPK_CFLAGS provided by the module:

gcc $GLPK_CFLAGS -c my_prog.c

To link your code, use the variable $GLPK_LIBS:

gcc my_prog.o $GLPK_LIBS -o my_prog

glpsol

Additionally, the GLPK module contains a stand-alone LP/MIP solver, which can be used to process files written in the GNU MathProg modeling language. The solver can be invoked using the following command syntax:

glpsol [options] [filename]

For a complete list of options, use the following command:

glpsol --help

GMAP

GMAP is a genomic mapping and alignment program for mRNA and EST sequences.

Availability and Restrictions

Versions

The following versions of GMAP are available on OSC clusters:

Version	Oakley	Owens
2016-05-25	X*
2016-06-09		X*

* Current default version

You can use module spider gmap to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

GMAP is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of GMAP, run the following command: module load gmap. The default version will be loaded. To select a particular GMAP version, use module load gmap/version. For example, use module load gmap/2016-05-25 to load GMAP 2016-05-25.

Usage on Owens

Set-up

To configure your environment for use of GMAP, run the following command: module load gmap. The default version will be loaded. To select a particular GMAP version, use module load gmap/version. For example, use module load gmap/2016-06-09 to load GMAP 2016-06-09.

GNU Compilers

Fortran, C and C++ compilers produced by the GNU Project.

Availability and Restrictions

Versions

GNU Compiler is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	Owens	notes
4.4.7	X#	X#		Default version on Oakley and Ruby prior to 09/15/2015
4.8.4	X	X
4.8.5	X*	X*	X*#	**See note below.
4.9.1	X	X
5.1.0	X
5.2.0	X	X
6.1.0			X
6.3.0	X	X	X	Libraries have been built for this version

* Current Default Version; # System version

** There is always some version of the GNU compilers in the environment. If you want a specific version you should load the appropriate module. If you don't have a gnu module loaded you will get either the system version or some other version, depending on what modules you do have loaded.

You can use module spider gnu to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

To find out what version of gcc you are using, type gcc --version.

Access

The GNU compilers are available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of the GNU compilers, run the following command (you may have to unload your selected compiler - if an error message appears, it will provide instructions): module load gnu. The default version will be loaded. To select a particular GNU version, use module load gnu/version. For example, use module load gnu/4.8.4 to load GNU 4.8.4.

How to Compile

Once the module is loaded, follow the guides below for compile commands:

Language	non-mpi	mpi
Fortran 90 or 95	`gfortran`	`mpif90`
Fortran 77	`g77`	`mpif77`
c	`gcc`	`mpicc`
c++	`g++`	`mpicxx`

Building Options

The GNU compilers recognize the following command line options :

Compiler Option	Purpose
`-fopenmp`	Enables compiler recognition of OpenMPdirectives (except g77 and mpi77)
`-o FILENAME`	Specifies the name of the object file
`-O0` or no `-O` option	Disable optimization
`-O1` or `-O`	Ligh optimization
`-O2`	Heavy optimization
`-O3`	Most expensive optimization (Recommended)

There are numerous flags that can be used. For more information run man <compiler binary name>.

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named hello.c and the output file named hello_results.
Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=12
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load gnu
cp hello.c $TMPDIR
cd $TMPDIR
gcc -O3 hello.c -o hello
./hello > hello_results
cp hello_results $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=12
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load gnu
mpicc -O3 hello.c -o hello
pbsdcp hello $TMPDIR
cd $TMPDIR
mpiexec hello > hello_results
pbsdcp -g hello_results $PBS_O_WORKDIR

Usage on Ruby

Set-up

To configure your environment for use of the GNU compilers, run the following command (you may have to unload your selected compiler - if an error message appears, it will provide instructions): module load gnu. The default version will be loaded. To select a particular GNU version, use module load gnu/version. For example, use module load gnu/4.9.1 to load GNU 4.9.1.

How to Compile

Once the module is loaded, follow the guides below for compile commands:

Language	non-mpi	mpi
Fortran 90 or 95	`gfortran`	`mpif90`
Fortran 77	`g77`	`mpif77`
c	`gcc`	`mpicc`
c++	`g++`	`mpicxx`

Building Options

The GNU compilers recognize the following command line options :

Compiler Option	Purpose
`-fopenmp`	Enables compiler recognition of OpenMPdirectives (except g77 and mpi77)
`-o FILENAME`	Specifies the name of the object file
`-O0` or no `-O` option	Disable optimization
`-O1` or `-O`	Ligh optimization
`-O2`	Heavy optimization
`-O3`	Most expensive optimization (Recommended)

There are numerous flags that can be used. For more information run man <compiler binary name>.

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=1:00:00

which gives you 20 cores (-l nodes=1:ppn=20) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named hello.c and the output file named hello_results.
Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=20
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load gnu
cp hello.c $TMPDIR
cd $TMPDIR
gcc -O3 hello.c -o hello
./hello > hello_results
cp hello_results $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=20
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load gnu
mpicc -O3 hello.c -o hello
cp hello $TMPDIR
cd $TMPDIR
mpiexec ./hello > hello_results
cp hello_results $PBS_O_WORKDIR

Usage on Owens

Set-up

To configure your environment for use of the GNU compilers, run the following command (you may have to unload your selected compiler - if an error message appears, it will provide instructions): module load gnu. The default version will be loaded. To select a particular GNU version, use module load gnu/version. For example, use module load gnu/4.8.5 to load GNU 4.8.5.

How to Compile

Once the module is loaded, follow the guides below for compile commands:

Language	non-mpi	mpi
Fortran 90 or 95	`gfortran`	`mpif90`
Fortran 77	`g77`	`mpif77`
c	`gcc`	`mpicc`
c++	`g++`	`mpicxx`

Building Options

The GNU compilers recognize the following command line options :

Compiler Option	Purpose
`-fopenmp`	Enables compiler recognition of OpenMPdirectives (except g77 and mpi77)
`-o FILENAME`	Specifies the name of the object file
`-O0` or no `-O` option	Disable optimization
`-O1` or `-O`	Ligh optimization
`-O2`	Heavy optimization
`-O3`	Most expensive optimization (Recommended)

There are numerous flags that can be used. For more information run man <compiler binary name>.

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named hello.c and the output file named hello_results.
Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=28
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load gnu
cp hello.c $TMPDIR
cd $TMPDIR
gcc -O3 hello.c -o hello
./hello > hello_results
cp hello_results $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=28
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load gnu
mpicc -O3 hello.c -o hello
cp hello $TMPDIR
cd $TMPDIR
mpiexec ./hello > hello_results
cp hello_results $PBS_O_WORKDIR

GPU-BLAST

GPU-BLAST is an accelerated version of the popular NCBI-BLAST (www.ncbi.nlm.nih.gov) that uses general-purpose graphics processing unit (GPU). In comparison to the sequential NCBI-BLAST, GPU-BLAST is nearly four times faster, while producing identical results.

Availability and Restrictions

Versions

The following versions of GPU-BLAST are available on OSC systems:

VERSION	NCBI-BLAST VERSION	OAKLEY	RUBY
1.0	2.2.24+	X
1.1	2.2.25+	X
1.1	2.2.26+	X*
1.1	2.2.28+		X*

* Current default version

You can use module spider gpu-blast to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Note that GPU-BLAST is not available on the Owens system. GPU-BLAST requires the CUDA nvcc compiler, but the GPU-BLAST software is only compatible with CUDA versions up to 7.5. CUDA 8, installed on the Owens cluster, is incompatible with GPU-BLAST.

Access

GPU-BLAST is available without restriction to all OSC users.

Usage

Usage on Oakley

Set-up

To set up the environment for gpu-blast on Ruby cluster, do the following:

module load cuda/4.1.28
module load gpu-blast

Batch Usage

Provided is a sample batch script:

#PBS -l nodes=1:ppn=1:gpus=1
#PBS -l walltime=10:00
#PBS -N GPU-Blast
#PBS -S /bin/bash
#PBS -j oe
module load gpu-blast
module load cuda/4.1.28
set -x
cd $PBS_O_WORKDIR
mkdir $PBS_JOBID
cp 100.fasta $TMPDIR
cd $TMPDIR
/usr/bin/time blastn -db nt -query 100.fasta -out test.out
cp * $PBS_O_WORKDIR/$PBS_JOBID

Usage on Ruby

Set-up

To set up the environment for gpu-blast on Ruby cluster, do the following:

module load cuda/7.5.18
module load gpu-blast

Batch Usage

Provided is a sample batch script:

#PBS -l nodes=1:ppn=1:gpus=1
#PBS -l walltime=10:00
#PBS -N GPU-Blast
#PBS -S /bin/bash
#PBS -j oe
module load gpu-blast
module load cuda/7.5.18
set -x
cd $PBS_O_WORKDIR
mkdir $PBS_JOBID
cp 100.fasta $TMPDIR
cd $TMPDIR
/usr/bin/time blastn -db nt -query 100.fasta -out test.out
cp * $PBS_O_WORKDIR/$PBS_JOBID

GROMACS

GROMACS is a versatile package of molecular dynamics simulation programs. It is primarily designed for biochemical molecules, but it has also been used on non-biological systems. GROMACS generally scales well on OSC platforms. Starting with version 4.6 GROMACS includes GPU acceleration.

Availability and Restrictions

Versions

GROMACS is available on Oakley and Owens Clusters. The versions currently available at OSC are the following:

Version	Oakley	Owens	notes
4.5.5	SP		Default version on Oakley prior to 09/15/2015
4.6.3	SPC		Default version on Oakley prior to 02/21/2017
5.1	SPC*
5.1.2		SPC*
2016.4		SPC

* Current default version; S = serial single node executables; P = parallel multinode; C = CUDA

You can use module spider gromacs to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

GROMACS is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To load the module for the default version of GROMACS, which initializes your environment for the GROMACS application, use module load gromacs. To select a particular software version, use module load gromacs/version. For example, use module load gromacs/5.1 to load GROMACS version 5.1; and use module help gromacs/5.1 to view details, such as, compiler prerequisites, additional modules required for specific executables, the suffixes of executables, etc.

Using GROMACS

The syntax of the GROMACS executable(s) changed in version 5. To execute a serial GROMACS versions 4 program interactively, simply run it on the command line, e.g.:

pdb2gmx

To execute a serial GROMACS versions 5 program interactively, simply run it on the command line, e.g.:

gmx pdb2gmx

Note that some serial executables are enabled for single node parallelism via OpenMP. Parallel multinode GROMACS versions 4 programs should be run in a batch environment with mpiexec, e.g.:

mpiexec mdrun_mpi_d

Parallel multinode GROMACS versions 5 programs should be run in a batch environment with mpiexec, e.g.:

mpiexec gmx_mpi_d mdrun

Note that '_mpi' indicates a parallel executable and '_d' indicates a program built with double precision ('_gpu' denotes a GPU executable built with CUDA). See the module help output for specific versions for more details on executable naming conventions.

Batch Usage

When you log into Oakley you are actually connected to a login node. To access the compute nodes, you must submit a job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Oakley, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and input files are available here:

~srb/workshops/compchem/gromacs/

This simple batch script demonstrates some important points:

# GROMACS Tutorial for Solvation Study of Spider Toxin Peptide
# see fwspider_tutor.pdf
#PBS -N fwsinvacuo.oakley
#PBS -l nodes=2:ppn=12
module unload intel
module load intel/12.1.4.319
module load mvapich2/1.7
module load gromacs/4.6.3
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cd $PBS_O_WORKDIR
pbsdcp -p 1OMB.pdb em.mdp $TMPDIR
# Use TMPDIR for best performance.
cd $TMPDIR
pdb2gmx -ignh -ff gromos43a1 -f 1OMB.pdb -o fws.gro -p fws.top -water none
editconf -f fws.gro -d 0.7
editconf -f out.gro -o fws_ctr.gro -center 2.0715 1.6745 1.914
grompp -f em.mdp -c fws_ctr.gro -p fws.top -o fws_em.tpr
mpiexec mdrun_mpi -s fws_em.tpr -o fws_em.trr -c fws_ctr.gro -g em.log -e em.edr
cat em.log
cp -p * $PBS_O_WORKDIR/

Usage on Owens

Set-up

To load the module for the default version of GROMACS, which initializes your environment for the GROMACS application, use module load gromacs. To select a particular software version, use module load gromacs/version. For example, use module load gromacs/5.1.2 to load GROMACS version 5.1.2; and use module help gromacs/5.1.2 to view details, such as, compiler prerequisites, additional modules required for specific executables, the suffixes of executables, etc.

Using GROMACS

To execute a serial GROMACS versions 5 program interactively, simply run it on the command line, e.g.:

gmx pdb2gmx

Parallel multinode GROMACS versions 5 programs should be run in a batch environment with mpiexec, e.g.:

mpiexec gmx_mpi_d mdrun

Note that '_mpi' indicates a parallel executable and '_d' indicates a program built with double precision ('_gpu' denotes a GPU executable built with CUDA). See the module help output for specific versions for more details on executable naming conventions.

Batch Usage

When you log into Owens you are actually connected to a login node. To access the compute nodes, you must submit a job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Owens, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and input files are available here:

~srb/workshops/compchem/gromacs/

This simple batch script demonstrates some important points:

# GROMACS Tutorial for Solvation Study of Spider Toxin Peptide
# see fwspider_tutor.pdf
#PBS -N fwsinvacuo.owens
#PBS -l nodes=2:ppn=28
module load gromacs
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cd $PBS_O_WORKDIR
pbsdcp -p 1OMB.pdb em.mdp $TMPDIR
# Use TMPDIR for best performance.
cd $TMPDIR
pdb2gmx -ignh -ff gromos43a1 -f 1OMB.pdb -o fws.gro -p fws.top -water none
editconf -f fws.gro -d 0.7
editconf -f out.gro -o fws_ctr.gro -center 2.0715 1.6745 1.914
grompp -f em.mdp -c fws_ctr.gro -p fws.top -o fws_em.tpr
mpiexec gmx_mpi mdrun -s fws_em.tpr -o fws_em.trr -c fws_ctr.gro -g em.log -e em.edr
cat em.log
cp -p * $PBS_O_WORKDIR/

Gaussian

Gaussian is the most popular general purpose electronic structure program. Its latest version, g09, can perform density functional theory, Hartree-Fock, Möller-Plesset, coupled-cluster, and configuration interaction calculations among others. Geometry optimizations, vibrational frequencies, magnetic properties, and solution modeling are available. It performs well as black-box software on closed-shell ground state systems.

Availability and Restrictions

Versions

Gaussian is available on Ruby, Oakley, and Owen Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	OWENS	notes
g03d01
g03e01
g09a01
g09b01	X
g09c01	X			Default version on Oakley prior to 09/15/2015
g09d01	X	X
g09e01	X*	X*	X*
g16a03 g16a03-GPU		X X	X

* Current default version

You can use module spider gaussian to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

Use of Gaussian for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Access for Commerical Users

Contact OSC Help for getting access to Gaussian if you are a commerical user.

Usage

Usage on Oakley

Set-up on Oakley

To load the default version of Gaussian module/initalize your environement for use of Gaussian, use module load gaussian. To select a particular software version, use module load gaussian/version. For example, use module load gaussian/g09b01 to load Gaussian version g09b01 on Oakley.

Using Gaussian

To execute Gaussian, simply run the Gaussian binary with the input file on the command line:

g09 < input.com

When the input file is redirected as above ( < ), the output will be standard output; in this form the output can be seen via 'qpeek jobid' when the job is running in a batch queue. Alternatively, Gaussian can be invoked without file redirection:

g09 input.com

in which case the output file will be named 'input.log'; in this form the output cannot be seen via 'qpeek jobid' when the job is running in a batch queue.

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Oakley, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and Gaussian input files are available here:

/users/appl/srb/workshops/compchem/gaussian/

This simple batch script demonstrates the important points:

#PBS -N GaussianJob
#PBS -l nodes=1:ppn=1
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cd $PBS_O_WORKDIR
cp input.com $TMPDIR
# Use TMPDIR for best performance.
cd $TMPDIR
module load gaussian
g16 < input.com
cp -p input.log *.chk $PBS_O_WORKDIR

Note: OSC does not have a functional distributed parallel version (LINDA) of Gaussian. Parallelism of Gaussian at OSC is only via shared memory. Consequently, do not request more than one node for Gaussian jobs on OSC's clusters.

Usage on Ruby

Set-up on Ruby

To load the default version of Gaussian module/initalize your environement for use of Gaussian, use module load gaussian. To select a particular software version, use module load gaussian/version. For example, use module load gaussian/g09d01 to load Gaussian version g09d01 on Ruby.

Using Gaussian

To execute Gaussian, simply run the Gaussian binary with the input file on the command line:

g09 < input.com

When the input file is redirected as above ( < ), the output will be standard output; in this form the output can be seen via 'qpeek jobid' when the job is running in a batch queue. Alternatively, Gaussian can be invoked without file redirection:

g09 input.com

in which case the output file will be named 'input.log'; in this form the output cannot be seen via 'qpeek jobid' when the job is running in a batch queue.

Batch Usage on Ruby

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Ruby, one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=1:00:00

which gives you 20 cores (-l nodes=1:ppn=20) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and Gaussian input files are available here:

/users/appl/srb/workshops/compchem/gaussian/

This simple batch script demonstrates the important points:

#PBS -N GaussianJob
#PBS -l nodes=1:ppn=20
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cd $PBS_O_WORKDIR
cp input.com $TMPDIR
# Use TMPDIR for best performance.
cd $TMPDIR
module load gaussian
g09 < input.com
cp -p input.log *.chk $PBS_O_WORKDIR

Note: OSC does not have a functional distributed parallel version (LINDA) of Gaussian. Parallelism of Gaussian at OSC is only via shared memory. Consequently, do not request more than one node for Gaussian jobs on OSC's clusters.

Running Gaussian jobs with GPU

Gaussian jobs can utilize K40 GPUs that are avaialabe on Ruby. GPUs are not helpful for small jobs but are effective for larger molecules when doing DFT energies, gradients and frequencies (for both ground and excited states). They are also not used effectively by post-SCF calculations such as MP2 or CCSD. For more information, please visit http://gaussian.com/relnotes/?tabid=2

A sample batch script for GPU run will look like as follows:

#PBS -S /bin/tcsh
#PBS -N methane 
#PBS -o methane.log
#PBS -l nodes=1:ppn=20:gpus=1:default
set echo
cd $TMPDIR
set INPUT=methane.com
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cp $PBS_O_WORKDIR/$INPUT .
module load gaussian/g16a03
g16 < ./$INPUT
ls -al
cp *.chk $PBS_O_WORKDIR
cp *.log $PBS_O_WORKDIR

A sample input file for GPU run will look like as follows:

%nproc=20
%mem=8gb
%CPU=0-19
%GPUCPU=0=0
%chk=methane.chk
#b3lyp/6-31G(d) opt
methane B3LYP/6-31G(d) opt freq
0,1
C        0.000000        0.000000        0.000000
H        0.000000        0.000000        1.089000
H        1.026719        0.000000       -0.363000
H       -0.513360       -0.889165       -0.363000
H       -0.513360        0.889165       -0.363000

The above example will utilize CPUs indexed from 0 to 19th, but 0th CPU is associated with 0th GPU.

Usage on Owens

Set-up on Owens

To load the default version of Gaussian module/initalize your environement for use of Gaussian, use module load gaussian. To select a particular software version, use module load gaussian/version. For example, use module load gaussian/g09d01 to load Gaussian version g09d01 on Owens.

Using Gaussian

To execute Gaussian, simply run the Gaussian binary with the input file on the command line:

g09 < input.com

When the input file is redirected as above ( < ), the output will be standard output; in this form the output can be seen via 'qpeek jobid' when the job is running in a batch queue. Alternatively, Gaussian can be invoked without file redirection:

g09 input.com

in which case the output file will be named 'input.log'; in this form the output cannot be seen via 'qpeek jobid' when the job is running in a batch queue.

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Owens, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and Gaussian input files are available here:

/users/appl/srb/workshops/compchem/gaussian/

This simple batch script demonstrates the important points:

#PBS -N GaussianJob
#PBS -l nodes=1:ppn=28
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cd $PBS_O_WORKDIR
cp input.com $TMPDIR
# Use TMPDIR for best performance.
cd $TMPDIR
module load gaussian
g09 < input.com
cp -p input.log *.chk $PBS_O_WORKDIR

Note: OSC does not have a functional distributed parallel version (LINDA) of Gaussian. Parallelism of Gaussian at OSC is only via shared memory. Consequently, do not request more than one node for Gaussian jobs on OSC's clusters.

Gnuplot

Gnuplot is a portable command-line driven data and function plotting utility. It was originally intended to allow scientists and students to visualize mathematical functions and data.

Gnuplot supports many types of plots in two or three dimensions. It can draw using points, lines, boxes, contours, vector fields surfaces and various associated text. It also supports various specialized plot types.

Gnuplot supports many different types of output: interactive screen display (with mouse and hotkey functionality), pen plotters (like hpgl), printers (including postscript and many color devices), and file formats as vectorial pseudo-devices like LaTeX, metafont, pdf, svg, or bitmap png.

Availability and Restrictions

Versions

The current versions of Gnuplot available at OSC are:

VERSION	OAKLEY	OWENS	NOTES
4.6.4	X*
4.6 patchlevel 2		System Install	No module needed.
5.2.2		X*

* Current default version

You can use module spider gnuplot to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Gnuplot is available to all OSC users.

Usage

Usage on Oakley

To start a Gnuplot session, load the module and launch using the following commands:

module load gnuplot

gnuplot

To access the Gnuplot help menu, type ? into the Gnuplot command line.

Usage on Owens

To start a Gnuplot session, load the module and launch using the following commands:

module load gnuplot

gnuplot

To access the Gnuplot help menu, type ? into the Gnuplot command line.

HDF5

HDF5 is a general purpose library and file format for storing scientific data. HDF5 can store two primary objects: datasets and groups. A dataset is essentially a multidimensional array of data elements, and a group is a structure for organizing objects in an HDF5 file. Using these two basic objects, one can create and store almost any kind of scientific data structure, such as images, arrays of vectors, and structured and unstructured grids.

Availability and Restrictions

Versions

HDF5 is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	Owens	notes
1.8.8	X
1.8.12		X(GI)		Default version on Ruby prior to 09/14/2015
1.8.13	X(GI)	X(GI)		Default version on Oakley prior to 09/14/2015
1.8.15	X	X(PI)
1.8.17	X*	X*	X*
1.8.19			X

* Current Default Version; P = available with only pgi; PI = available with only pgi and intel; GI = available with only gnu and intel

You can use module spider hdf5 to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

HDF5 is available without restriction to all OSC users.

Usage

Usage on Oakley

Set-up

Initalizing the system for use of the HDF5 library is dependent on the system you are using and the compiler you are using. To load the default HDF5 library, run the following command: module load hdf5. To load a particular version, use module load hdf5/version. For example, use module load hdf5/1.8.15 to load HDF5 version 1.8.15. You can use module spider hdf5 to view available modules.

Building With HDF5

The HDF5 library provides the following variables for use at build time:

Variable	Use
`$HDF5_C_INCLUDE`	Use during your compilation step for C programs
`$HDF5_CPP_INCLUDE`	Use during your compilation step for C++ programs (serial version only)
`$HDF5_F90_INCLUDE`	Use during your compilation step for FORTRAN programs
`$HDF5_C_LIBS`	Use during your linking step programs
`$HDF5_F90_LIBS`	Use during your linking step for FORTRAN programs

For example, to build the code myprog.c or myprog.f90 with the hdf5 library you would use:

icc -c $HDF5_C_INCLUDE myprog.c
icc -o myprog myprog.o $HDF5_C_LIBS
ifort -c $HDF5_F90_INCLUDE myprog.f90
ifort -o myprog myprog.o $HDF5_F90_LIBS

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is the example batch script that executes a program built with the HDF5 library:

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=12
module load hdf5
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
appname
cp foo_out.h5 $PBS_O_WORKDIR

Usage on Ruby

Set-up

Initalizing the system for use of the HDF5 library is dependent on the system you are using and the compiler you are using. To load the default HDF5 library, run the following command: module load hdf5. To load a particular version, use module load hdf5/version. For example, use module load hdf5/1.8.15 to load HDF5 version 1.8.15. You can use module spider hdf5 to view available modules.

Building With HDF5

The HDF5 library provides the following variables for use at build time:

Variable	Use
`$HDF5_C_INCLUDE`	Use during your compilation step for C programs
`$HDF5_CPP_INCLUDE`	Use during your compilation step for C++ programs (serial version only)
`$HDF5_F90_INCLUDE`	Use during your compilation step for FORTRAN programs
`$HDF5_C_LIBS`	Use during your linking step programs
`$HDF5_F90_LIBS`	Use during your linking step for FORTRAN programs

For example, to build the code myprog.c or myprog.f90 with the hdf5 library you would use:

icc -c $HDF5_C_INCLUDE myprog.c
icc -o myprog myprog.o $HDF5_C_LIBS
ifort -c $HDF5_F90_INCLUDE myprog.f90
ifort -o myprog myprog.o $HDF5_F90_LIBS

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is the example batch script that executes a program built with the HDF5 library:

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=20
module load hdf5
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
appname
cp foo_out.h5 $PBS_O_WORKDIR

Usage on Owens

Set-up

Initalizing the system for use of the HDF5 library is dependent on the system you are using and the compiler you are using. To load the default HDF5 library, run the following command: module load hdf5. To load a particular version, use module load hdf5/version. For example, use module load hdf5/1.8.17 to load HDF5 version 1.8.17. You can use module spider hdf5 to view available modules.

Building With HDF5

The HDF5 library provides the following variables for use at build time:

Variable	Use
`$HDF5_C_INCLUDE`	Use during your compilation step for C programs
`$HDF5_CPP_INCLUDE`	Use during your compilation step for C++ programs (serial version only)
`$HDF5_F90_INCLUDE`	Use during your compilation step for FORTRAN programs
`$HDF5_C_LIBS`	Use during your linking step programs
`$HDF5_F90_LIBS`	Use during your linking step for FORTRAN programs

For example, to build the code myprog.c or myprog.f90 with the hdf5 library you would use:

icc -c $HDF5_C_INCLUDE myprog.c
icc -o myprog myprog.o $HDF5_C_LIBS
ifort -c $HDF5_F90_INCLUDE myprog.f90
ifort -o myprog myprog.o $HDF5_F90_LIBS

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is the example batch script that executes a program built with the HDF5 library:

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=28
module load hdf5
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
appname
cp foo_out.h5 $PBS_O_WORKDIR

HISAT2

HISAT2 is a graph-based alignment program that maps DNA and RNA sequencing reads to a population of human genomes.

Availability and Restrictions

Versions

HISAT2 is available on the Owens Cluster. The versions currently available at OSC are:

VERSION	OWENS
2.1.0	X*

* Current Default Version

You can use module spider hisat2 to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

HISAT2 is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of HISAT2, use command module load hisat2. This will load the default version.

HMMER

Profile hidden Markov models (profile HMMs) can be used to do sensitive database searching using statistical descriptions of a sequence family's consensus. HMMER uses profile HMMs, and can be useful in situations like:

if you are working with an evolutionarily diverse protein family, a BLAST search with any individual sequence may not find the rest of the sequences in the family.
the top hits in a BLAST search are hypothetical sequences from genome projects.
your protein consists of several domains which are of different types.

HMMER (pronounced 'hammer', as in a more precise mining tool than BLAST) was developed by Sean Eddy at Washington University in St. Louis.

HMMER is a very cpu-intensive program and is parallelized using threads, so that each instance of hmmsearch or the other search programs can use all the cpus available on a node. HMMER on OSC clusters are intended for those who need to run HMMER searches on large numbers of query sequences.

Availability and Restrictions

Versions

The following versions of HMMER are available on OSC systems:

Version	Oakley
3.0	X*

* Current default version

You can use module spider hmmer to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

HMMER is available to all OSC users without restriction.

Usage

Set-up

To use HMMER on Oakley, first run the following command:

module load hmmer

Using HMMER

Once the hmmer module is loaded, the following commands will be available for your use:

Single sequence queries: new to HMMER3
phmmer Search a sequence against a sequence database. (BLASTP-like)
jackhmmer Iteratively search a sequence against a sequence database. (PSIBLAST-like)

Replacements for HMMER2’s functionality
hmmbuild Build a profile HMM from an input multiple alignment.
hmmsearch Search a profile HMM against a sequence database.
hmmscan Search a sequence against a profile HMM database.
hmmalign Make a multiple alignment of many sequences to a common profile HMM.

Other utilities
hmmconvert Convert profile formats to/from HMMER3 format.
hmmemit Generate (sample) sequences from a profile HMM.
hmmfetch Get a profile HMM by name or accession from an HMM database.
hmmpress Format an HMM database into a binary format for hmmscan.
hmmstat Show summary statistics for each profile in an HMM database.

If you need to know options for a command, type the command name followed by "-h", for example:

hmmalign -h

Batch Usage

A sample batch job is below:

#PBS -N hmmer
#PBS -j oe
#PBS -l nodes=1:ppn=1
#PBS -S /bin/bash

cd $PBS_O_WORKDIR
hmmalign globins4.align globins45
hmmbuild globins4.hmm globins4.align
hmmsearch globins4.hmm /fdb/fastadb/nr.aa.fas > globins.out

HOMER

HOMER (Hypergeometric Optimization of Motif EnRichment) is a suite of tools for Motif Discovery and ChIP-Seq analysis. It is a collection of command line programs for unix-style operating systems written in mostly perl and c++. Homer was primarily written as a de novo motif discovery algorithm that is well suited for finding 8-12 bp motifs in large scale genomics data.

Availability and Restrictions

Versions

The following versions of HOMER are available on OSC clusters:

Version	Oakley	owens
4.8	X*	X*

* Current default version

You can use module spider homer to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

HOMER is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of HOMER, run the following command: module load homer. The default version will be loaded. To select a particular HOMER version, use module load homer/version. For example, use module load homer/4.8 to load HOMER 4.8.

Usage on Owens

Set-up

To configure your environment for use of HOMER, run the following command: module load homer. The default version will be loaded. To select a particular HOMER version, use module load homer/version. For example, use module load homer/4.8 to load HOMER 4.8.

HTSlib

HTSlib is a C library used for reading and writing high-throughput sequencing data. HTSlib is the core library used by SAMtools. HTSlib also provides the bgzip, htsfile, and tabix utilities.

Availability and Restrictions

Versions

The versions of HTSlib currently available at OSC are:

VERSION	Oakley	OWENS
1.6	X*	X*

* Current Default Version

You can use module spider htslib to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

HTSlib is available to all OSC users.

Usage

Usage on Oakley

Set-up

To set up the environment for HTSlib on the OSC clusters, use the command:

module load htslib/1.6

Usage on Owens

Set-up

To set up the environment for HTSlib on the OSC clusters, use the command:

module load htslib/1.6

Hadoop

A hadoop cluster can be launched within the HPC environment, but managed by the PBS job scheduler using Myhadoop framework developed by San Diego Supercomputer Center. (Please see http://www.sdsc.edu/~allans/MyHadoop.pdf)

Availability and Restrictions

Versions

The following versions of Hadoop are available on OSC systems:

VERSION	OWENS
3.0.0-alpha1	X*

* Current default version

You can use module spider hadoop to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Hadoop is available to all OSC users without restriction.

Usage

Set-up

In order to configure your environment for the usage of Hadoop, run the following command:

module load hadoop

In order to access a particular version of Hadoop, run the following command

module load hadoop/3.0.0-alpha1

Using Hadoop

In order to run Hadoop in batch, reference the example batch script below. This script requests 6 node on the Owens cluster for 1 hour of walltime.

#PBS -N hadoop-example

#PBS -l nodes=6:ppn=12

#PBS -l walltime=01:00:00

setenv WORK $PBS_O_WORKDIR

module load hadoop/3.0.0-alpha1

module load myhadoop/v0.40

setenv HADOOP_CONF_DIR $TMPDIR/mycluster-conf-$PBS_JOBID

cd $TMPDIR

myhadoop-configure.sh -c $HADOOP_CONF_DIR -s $TMPDIR

$HADOOP_HOME/sbin/start-dfs.sh

hadoop dfsadmin -report

hadoop  dfs -mkdir data

hadoop  dfs -put $HADOOP_HOME/README.txt  data/

hadoop  dfs -ls data

hadoop jar $HADOOP_HOME/share/hadoop/mapreduce/hadoop-mapreduce-examples-3.0.0-alpha1.jar wordcount data/README.txt wordcount-out

hadoop  dfs -ls wordcount-out

hadoop  dfs  -copyToLocal -f  wordcount-out  $WORK

$HADOOP_HOME/sbin/stop-dfs.sh

myhadoop-cleanup.sh

Example Jobs

Please check /usr/local/src/hadoop/3.0.0-alpha1/test.osc folder for more examples of hadoop jobs

Intel Compilers

The Intel compilers for both C/C++ and FORTRAN.

Availability and Restrictions

Versions

The versions currently available at OSC are:

Version	Oakley	Ruby	owens	notes
12.1.4.319	X			Default version on Oakley prior to 09/15/2015
13.1.3.192	X
14.0.0.080	X
14.0.3		X
15.0.0		X		Default version on Ruby prior to 09/15/2015
15.0.1		X
15.0.3	X	X
16.0.0	X	X
16.0.3	X*	X*	X*
17.0.2	X	X	X
17.0.5			X
18.0.0			X

* Current Default Version

You can use module spider intel to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

The Intel Compilers are available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

After you ssh to Oakley, the default version of Intel compilers will be loaded for you automatically. To configure your environment for a particular Intel compilers version, use module swap intel/12.1.4.39 intel/version. For example, use module swap intel/12.1.4.39 intel/14.0.0.080 to load Intel compilers 14.0.0.080 on Oakley.

Using the Intel Compilers

Once the intel compiler module has been loaded, the compilers are available for your use. The following table lists common compiler options available in all languages.

Compiler option	Purpose
`-c`	Compile only; do not link
`-DMACRO[=value]`	Defines preprocessor macro MACRO with optional value (default value is 1)
`-g`	Enables debugging; disables optimization
`-I/directory/name`	Add /directory/name to the list of directories to be searched for #include files
`-L/directory/name`	Adds /directory/name to the list of directories to be searched for library files
`-lname`	Adds the library libname.a or libname.so to the list of libraries to be linked
`-o outfile`	Names the resulting executable outfile instead of a.out
`-UMACRO`	Removes definition of MACRO from preprocessor
`-O0`	Disable optimization
`-O1`	Light optimization
`-O2`	Heavy optimization (default)
`-O3`	Aggressive optimization; may change numerical results
`-ipo`	Inline function expansion for calls to procedures defined in separate files
`-funroll-loops`	Loop unrolling
`-parallel`	Automatic parallelization
`-openmp`	Enables translation of OpenMP directives

The following table lists some options specific to C/C++

`-strict-ansi`	Enforces strict ANSI C/C++ compliance
`-ansi`	Enforces loose ANSI C/C++ compliance

The following table lists some options specific to Fortran

`-convert big_endian`	Use unformatted I/O compatible with Sun and SGI systems
`-convert cray`	Use unformatted I/O compatible with Cray systems
`-i8`	Makes 8-byte INTEGERs the default
`-module /dir/name`	Adds /dir/name to the list of directories searched for Fortran 90 modules
`-r8`	Makes 8-byte REALs the default
`-fp-model strict`	Disables optimizations that can change the results of floating point calculations

Some parts of the C++11 and C++14 standards are available by default in various versions of the Intel compilers; others require you to load an extra module.

With the Intel 14 or 15 compilers:

module load cxx11

With the Intel 16 compiler:

module load cxx14

See our compilation guide for suggestions on how to compile your software on our systems.

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session on Oakley, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named hello.c and the output file named hello_results. Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=12
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load intel
cp hello.c $TMPDIR
cd $TMPDIR
icc -O2 hello.c -o hello
./hello > hello_results
cp hello_results $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=12
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load intel
mpicc -O2 hello.c -o hello
pbsdcp hello $TMPDIR
cd $TMPDIR
mpiexec hello > hello_results
pbsdcp -g hello_results $PBS_O_WORKDIR

Usage on Ruby

Set-up on Ruby

After you ssh to Ruby, the default version of Intel compilers will be loaded for you automatically. To configure your environment for a particular Intel compilers version, use module swap intel/15.0.0 intel/version. For example, use module swap intel/15.0.0 intel/15.0.1 to load Intel compilers 15.0.1 on Ruby.

Using the Intel Compilers

Once the intel compiler module has been loaded, the compilers are available for your use. The following table lists common compiler options available in all languages.

Compiler option	Purpose
`-c`	Compile only; do not link
`-DMACRO[=value]`	Defines preprocessor macro MACRO with optional value (default value is 1)
`-g`	Enables debugging; disables optimization
`-I/directory/name`	Add /directory/name to the list of directories to be searched for #include files
`-L/directory/name`	Adds /directory/name to the list of directories to be searched for library files
`-lname`	Adds the library libname.a or libname.so to the list of libraries to be linked
`-o outfile`	Names the resulting executable outfile instead of a.out
`-UMACRO`	Removes definition of MACRO from preprocessor
`-O0`	Disable optimization
`-O1`	Light optimization
`-O2`	Heavy optimization (default)
`-O3`	Aggressive optimization; may change numerical results
`-ipo`	Inline function expansion for calls to procedures defined in separate files
`-funroll-loops`	Loop unrolling
`-parallel`	Automatic parallelization
`-openmp`	Enables translation of OpenMP directives

The following table lists some options specific to C/C++

`-strict-ansi`	Enforces strict ANSI C/C++ compliance
`-ansi`	Enforces loose ANSI C/C++ compliance

The following table lists some options specific to Fortran

`-convert big_endian`	Use unformatted I/O compatible with Sun and SGI systems
`-convert cray`	Use unformatted I/O compatible with Cray systems
`-i8`	Makes 8-byte INTEGERs the default
`-module /dir/name`	Adds /dir/name to the list of directories searched for Fortran 90 modules
`-r8`	Makes 8-byte REALs the default
`-fp-model strict`	Disables optimizations that can change the results of floating point calculations

Some parts of the C++11 and C++14 standards are available by default in various versions of the Intel compilers; others require you to load an extra module.

With the Intel 14 or 15 compilers:

module load cxx11

With the Intel 16 compiler:

module load cxx14

See our compilation guide for suggestions on how to compile your software on our systems.

Batch Usage on Ruby

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Ruby, one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=1:00:00

which gives you 20 cores (-l nodes=1:ppn=20) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named hello.c and the output file named hello_results. Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=20
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load intel
cp hello.c $TMPDIR
cd $TMPDIR
icc -O2 hello.c -o hello
./hello > hello_results
cp hello_results $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=20
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load intel
mpicc -O2 hello.c -o hello
cp hello $TMPDIR
cd $TMPDIR
mpiexec ./hello > hello_results
cp hello_results $PBS_O_WORKDIR

Usage on Owens

Set-up on Owens

After you ssh to Owens, the default version of Intel compilers will be loaded for you automatically.

Using the Intel Compilers

Once the intel compiler module has been loaded, the compilers are available for your use. The following table lists common compiler options available in all languages.

COMPILER OPTION	PURPOSE
`-c`	Compile only; do not link
`-DMACRO[=value]`	Defines preprocessor macro MACRO with optional value (default value is 1)
`-g`	Enables debugging; disables optimization
`-I/directory/name`	Add /directory/name to the list of directories to be searched for #include files
`-L/directory/name`	Adds /directory/name to the list of directories to be searched for library files
`-lname`	Adds the library libname.a or libname.so to the list of libraries to be linked
`-o outfile`	Names the resulting executable outfile instead of a.out
`-UMACRO`	Removes definition of MACRO from preprocessor
`-O0`	Disable optimization
`-O1`	Light optimization
`-O2`	Heavy optimization (default)
`-O3`	Aggressive optimization; may change numerical results
`-ipo`	Inline function expansion for calls to procedures defined in separate files
`-funroll-loops`	Loop unrolling
`-parallel`	Automatic parallelization
`-openmp`	Enables translation of OpenMP directives

The following table lists some options specific to C/C++

`-strict-ansi`	Enforces strict ANSI C/C++ compliance
`-ansi`	Enforces loose ANSI C/C++ compliance

The following table lists some options specific to Fortran

`-convert big_endian`	Use unformatted I/O compatible with Sun and SGI systems
`-convert cray`	Use unformatted I/O compatible with Cray systems
`-i8`	Makes 8-byte INTEGERs the default
`-module /dir/name`	Adds /dir/name to the list of directories searched for Fortran 90 modules
`-r8`	Makes 8-byte REALs the default
`-fp-model strict`	Disables optimizations that can change the results of floating point calculations

Some parts of the C++11 standard are available by default in various versions of the Intel compilers; others require you to load an extra module.

With the Intel 14 or 15 compilers:

module load cxx11

See our compilation guide for suggestions on how to compile your software on our systems.

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session on Owens, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named hello.c and the output file named hello_results. Below is the example batch script (job.txt) for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=28
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load intel
cp hello.c $TMPDIR
cd $TMPDIR
icc -O2 hello.c -o hello
./hello > hello_results
cp hello_results $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=28
#PBS -N jobname
#PBS -j oe
cd $PBS_O_WORKDIR
module load intel
mpicc -O2 hello.c -o hello
cp hello $TMPDIR
cd $TMPDIR
mpiexec ./hello > hello_results
cp hello_results $PBS_O_WORKDIR

Intel MPI

Intel's implementation of the Message Passing Interface (MPI) library. See Intel Compilers for available compiler versions at OSC.

Availability and Restrictions

Versions

Intel MPI may be used as an alternative to - but not in conjunction with - the MVAPICH2 MPI libraries. The versions currently available at OSC are:

Version	Oakley	ruby	owens	notes
4.0.3.008	X			Default version on Oakley prior to 09/15/2015
4.1.0.024	X
4.1.1.036	X
4.1.2.040	X
4.1.3	X	X
4.1.3.049	X	X
5.0.1		X
5.0.1.035		X
5.0.3	X	X
5.1.3	X*	X*	X*
2017.2			X
2018.0			X

* Current Default Version

You can use module spider intelmpi to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Intel MPI is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

To configure your environment for the default version of Intel MPI, use module load intelmpi. To configure your environment for a specific version of Intel MPI, use module load intelmpi/version. For example, use module load intelmpi/4.1.3.049 to load Intel MPI version 4.1.3.049 on Oakley.

You can use module spider intelmpi to view available modules on Oakley.

Note: This module conflicts with the default loaded MVAPICH installations, and Lmod will automatically replace with the correct one when you use module load intelmpi.

Using Intel MPI

Software compiled against this module will use the libraries at runtime.

Building With Intel MPI

On Oakley, we have defined several environment variables to make it easier to build and link with the Intel MPI libraries.

Variable	Use
`$MPI_CFLAGS`	Use during your compilation step for C programs.
`$MPI_CXXFLAGS`	Use during your compilation step for C++ programs.
`$MPI_FFLAGS`	Use during your compilation step for Fortran programs.
`$MPI_F90FLAGS`	Use during your compilation step for Fortran 90 programs.
`$MPI_LIBS`	Use when linking your program to Intel MPI.

In general, for any application already set up to use mpicc compilation should be fairly straightforward.

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will run a program compiled against Intel MPI (called my-impi-application) for five hours on Oakley:

#PBS -N MyIntelMPIJob
#PBS -l nodes=4:ppn=12
#PBS -l walltime=5:00:00
module swap mvapich2 intelmpi
cd $PBS_O_WORKDIR
mpiexec my-impi-application

Usage on Ruby

Set-up on Ruby

To configure your environment for the default version of Intel MPI, use module load intelmpi. To configure your environment for a specific version of Intel MPI, use module load intelmpi/version. For example, use module load intelmpi/4.1.3.049 to load Intel MPI version 4.1.3.049 on Ruby.

You can use module spider intelmpi to view available modules on Ruby.

Note: This module conflicts with the default loaded MVAPICH installations, and Lmod will automatically replace with the correct one when you use module load intelmpi.

Using Intel MPI

Software compiled against this module will use the libraries at runtime.

Building With Intel MPI

On Ruby, we have defined several environment variables to make it easier to build and link with the Intel MPI libraries.

Variable	Use
`$MPI_CFLAGS`	Use during your compilation step for C programs.
`$MPI_CXXFLAGS`	Use during your compilation step for C++ programs.
`$MPI_FFLAGS`	Use during your compilation step for Fortran programs.
`$MPI_F90FLAGS`	Use during your compilation step for Fortran 90 programs.
`$MPI_LIBS`	Use when linking your program to Intel MPI.

In general, for any application already set up to use mpicccompilation should be fairly straightforward.

Batch Usage on Ruby

When you log into ruby.osc.edu you are actually logged into a Linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will run a program compiled against Intel MPI (called my-impi-application) for five hours on Ruby:

#PBS -N MyIntelMPIJob
#PBS -l nodes=4:ppn=20
#PBS -l walltime=5:00:00
module swap mvapich2 intelmpi
cd $PBS_O_WORKDIR
mpiexec my-impi-application

Usage on Owens

Set-up on Owens

To configure your environment for the default version of Intel MPI, use module load intelmpi. To configure your environment for a specific version of Intel MPI, use module load intelmpi/version. For example, use module load intelmpi/5.1.3 to load Intel MPI version 5.1.3 on Owens.

You can use module spider intelmpi to view available modules on Owens.

Note: This module conflicts with the default loaded MVAPICH installations, and Lmod will automatically replace with the correct one when you use module load intelmpi.

Using Intel MPI

Software compiled against this module will use the libraries at runtime.

Building With Intel MPI

On Ruby, we have defined several environment variables to make it easier to build and link with the Intel MPI libraries.

VARIABLE	USE
`$MPI_CFLAGS`	Use during your compilation step for C programs.
`$MPI_CXXFLAGS`	Use during your compilation step for C++ programs.
`$MPI_FFLAGS`	Use during your compilation step for Fortran programs.
`$MPI_F90FLAGS`	Use during your compilation step for Fortran 90 programs.
`$MPI_LIBS`	Use when linking your program to Intel MPI.

In general, for any application already set up to use mpicc, (or similar), compilation should be fairly straightforward.

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will run a program compiled against Intel MPI (called my-impi-application) for five hours on Owens:

#PBS -N MyIntelMPIJob
#PBS -l nodes=4:ppn=28
#PBS -l walltime=5:00:00
module swap mvapich2 intelmpi
cd $PBS_O_WORKDIR
mpiexec my-impi-application

JasPer

Introduction

JasPer is an extensible open source utility designed for the manipulation, conversion, compression and decompression of digital images. Currently supported image formats include bmp, jp2, jpc, jpg, pgx, mif and ras.

Availability

Version 1.900.1 is available on the Oakley cluster.

Usage

In order to use the JasPer library, you must load the JasPer module into your environment. To load the JasPer module, type the following command:

module load jasper

Useful commands

In addition to the JasPer libraries, a few default applications are also provided when the JasPer module is loaded.

To convert an image from one format to another, use the jasper command.

jasper [options]

For a comprehensive list of options available for this command, type

jasper --help

To compare particular qualities of two images, use the imgcmp command.

imgcmp -f [file 1] -F [file 2] -m [metric]

For a full list of acceptable metric arguments, type

imgcmp --help

To view a sequence of one or more images, use the jiv command.

jiv image

Alternatively, if you would like to view multiple images in sequence you can use the --wait option to specify the amount of time between the display of each image, creating a slideshow. The command below will display each image for 5 seconds before displaying the next image.

jiv --wait 5 [image1 image2 image3 ...]

For a full list of options, type

jiv --help

Julia

From julialang.org:

"Julia is a high-level, high-performance dynamic programming language for numerical computing. It provides a sophisticated compiler, distributed parallel execution, numerical accuracy, and an extensive mathematical function library. Julia’s Base library, largely written in Julia itself, also integrates mature, best-of-breed open source C and Fortran libraries for linear algebra, random number generation, signal processing, and string processing. In addition, the Julia developer community is contributing a number of external packages through Julia’s built-in package manager at a rapid pace. IJulia, a collaboration between the Jupyter and Julia communities, provides a powerful browser-based graphical notebook interface to Julia."

Availability and Restrictions

Versions

Julia is available on Owens Clusters. The versions currently available at OSC are:

Version	owens	notes
0.5.1	X*	This version works with Jupyter notebooks

*:Current default version

You can use module spider julia to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Julia is available for use by all OSC users.

Usage

Interactive Julia Notebooks

If you are using OnDemand, you can simply work with Jupyter and the selection of the Julia kernel to use interactive notebooks to work on an Owens compute node!

Navigate to ondemand.osc.edu and select a Jupyter notebook:

Jupyter Notebook

Service:

Software

Fields of Science:

Programming

Kallisto

Kallisto is an RNA-seq quantification program. It quantifies abundances of transcripts from RNA-seq data and uses psedoalignment to determine the compatibility of reads with targets, without needing alignment.

Availability and Restrictions

Versions

Kallisto is available on the Owens Clusters. The versions currently available at OSC are:

VERSION	OWENS
0.43.1	X*

* Current Default Version

You can use module spider kallisto to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Kallisto is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of Salmon, use command module load kallisto. This will load the default version.

LAMMPS

The Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) is a classical molecular dynamics code designed for high-performance simulation of large atomistic systems. LAMMPS generally scales well on OSC platforms, provides a variety of modeling techniques, and offers GPU accelerated computation.

Availability and Restrictions

Versions

LAMMPS is available on Oakley, Ruby, and Owens Clusters. The following versions are currently available at OSC:

Version	Oakley	Ruby	Owens	notes
Feb12	PC			Default version on Oakley prior to 09/15/2015
May12	PC
Feb14	SPC
5Sep14	P	P
7Dec15	PC	PC
14May16	PC*	PC*	P*
31Mar17			PC
16Mar18	PC	PC	PC

* Current default version; S = serial executables; P = parallel; C = CUDA

You can use module spider lammps to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

LAMMPS is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To load the default version of LAMMPS module and set up your environment, use module load lammps. To select a particular software version, use module load lammps/version. For example, use module load lammps/12Feb12 to load LAMMPS version Feb12.

Using LAMMPS

Once a module is loaded, LAMMPS can be run with the following command:

lammps < input.file

To see information on the packages and executables for a particular installation, run the module help command, for example:

module help lammps

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=1 -l walltime=00:20:00

which requests one core (-l nodes=1:ppn=1), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and LAMMPS input files are available here:

~srb/workshops/compchem/lammps/

Below is a sample batch script. It asks for 24 processors and 10 hours of walltime. If the job goes beyond 10 hours, the job would be terminated.

#PBS -N chain  
#PBS -l nodes=2:ppn=12  
#PBS -l walltime=10:00:00  
#PBS -S /bin/bash  
#PBS -j oe  
module load lammps  
cd $PBS_O_WORKDIR  
pbsdcp chain.in $TMPDIR  
cd $TMPDIR  
lammps < chain.in  
pbsdcp -g '*' $PBS_O_WORKDIR

Usage on Ruby

Set-up

To load the default version of LAMMPS module and set up your environment, use module load lammps. To select a particular software version, use module load lammps/version. For example, use module load lammps/5Sep14 to load LAMMPS version 5Sep14.

Using LAMMPS

Once a module is loaded, LAMMPS can be run with the following command:

lammps < input.file

To see information on the packages and executables for a particular installation, run the module help command, for example:

module help lammps

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=20:gpus=1 -l walltime=00:20:00

which requests one whole node with 20 cores (-l nodes=1:ppn=20), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and LAMMPS input files are available here:

~srb/workshops/compchem/lammps/

Below is a sample batch script. It asks for 40 processors and 10 hours of walltime. If the job goes beyond 10 hours, the job would be terminated.

#PBS -N chain  
#PBS -l nodes=2:ppn=20  
#PBS -l walltime=10:00:00  
#PBS -S /bin/bash  
#PBS -j oe  
module load lammps  
cd $PBS_O_WORKDIR  
pbsdcp chain.in $TMPDIR  
cd $TMPDIR  
lammps < chain.in  
pbsdcp -g '*' $PBS_O_WORKDIR

Usage on Owens

Set-up

To load the default version of LAMMPS module and set up your environment, use module load lammps. To select a particular software version, use module load lammps/version. For example, use module load lammps/14May16 to load LAMMPS version 14May16.

Using LAMMPS

Once a module is loaded, LAMMPS can be run with the following command:

lammps < input.file

To see information on the packages and executables for a particular installation, run the module help command, for example:

module help lammps

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=28:gpus=1 -l walltime=00:20:00

which requests one whole node with 28 cores (-l nodes=1:ppn=28), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Parallel Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Sample batch scripts and LAMMPS input files are available here:

~srb/workshops/compchem/lammps/

Below is a sample batch script. It asks for 56 processors and 10 hours of walltime. If the job goes beyond 10 hours, the job would be terminated.

#PBS -N chain  
#PBS -l nodes=2:ppn=28  
#PBS -l walltime=10:00:00  
#PBS -S /bin/bash  
#PBS -j oe  
module load lammps  
cd $PBS_O_WORKDIR  
pbsdcp chain.in $TMPDIR  
cd $TMPDIR  
lammps < chain.in  
pbsdcp -g '*' $PBS_O_WORKDIR

LAPACK

LAPACK (Linear Algebra PACKage) provides routines for solving systems of simultaneous linear equations, least-squares solutions of linear systems of equations, eigenvalue problems, and singular value problems.

Availability and Restrictions

A highly optimized implementation of LAPACK is available on all OSC clusters as part of the Intel Math Kernel Library (MKL). We recommend that you use MKL rather than building LAPACK for yourself. MKL is available to all OSC users.

Usage

See OSC's MKL software page for usage information. Note that there is no library named liblapack.a or liblapack.so. The flag "-llapack" on your link line will not work. You should modify your makefile or build script to link to the MKL libraries instead.

LS-DYNA

LS-DYNA is a general purpose finite element code for simulating complex structural problems, specializing in nonlinear, transient dynamic problems using explicit integration. LS-DYNA is one of the codes developed at Livermore Software Technology Corporation (LSTC).

Availability and Restrictions

Versions

LS-DYNA is available on Owens and Oakley Clusters for both serial (smp solver for single node jobs) and parallel (mpp solver for multipe node jobs) versions. The versions currently available at OSC are:

Version		Oakley	Owens	notes
971-R5.1.1	smp	X		Default version on Oakley prior to 09/15/2015
971-R5.1.1	mpp	X		Default version on Oakley prior to 09/15/2015
971-R7.0.0	smp	X
971-R7.0.0	mpp	X
971-R7.1.1	smp	X
971-R7.1.1	mpp	X
971-R9.0.1	smp	X*	X*
971-R9.0.1	mpp	X*	X*
971-R9.1.0	mpp	X

* Current default version

You can use module spider ls-dyna to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

OSC does not provide LS_DYNA license directly, however users with their own academic departmental license can use it on the OSC clusters. Please contact OSC Help for further instruction.

Access for Commerical Users

Contact OSC Help for getting access to LS-DYNA if you are a commerical user.

Usage

Usage on Oakley

Set-up on Oakley

To view available modules installed on Oakley, use module spider ls-dyna for smp solvers, and use module spider mpp for mpp solvers. In the module name, '_s' indicates single precision and '_d' indicates double precision. For example, mpp-dyna/971_d_R7.0.0 is the mpp solver with double precision on Oakley. Use module load name to load LS-DYNA with a particular software version. For example, use module load mpp-dyna/971_d_R7.0.0 to load LS-DYNA mpp solver version 7.0.0 with double precision on Oakley.

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=00:20:00

which requests one whole node with 12 cores (-l nodes=1:ppn=12), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Please follow the steps below to use LS-DYNA via the batch system:

1) copy your input files (explorer.k in the example below) to your work directory at OSC

2) create a batch script, similar to the following file, saved as job.txt. It uses the smp solver for a serial job (nodes=1) on Oakley:

#PBS -N plate_test
#PBS -l walltime=5:00:00
#PBS -l nodes=1:ppn=12
#PBS -j oe
# The following lines set up the LSDYNA environment
module load ls-dyna/971_d_7.0.0
#
# Move to the directory where the input files are located
#
cd $PBS_O_WORKDIR
#
# Run LSDYNA (number of cpus > 1)
#
lsdyna I=explorer.k NCPU=12

3) submit the script to the batch queue with the command: qsub job.txt.

When the job is finished, all the result files will be found in the directory where you submitted your job ($PBS_O_WORKDIR). Alternatively, you can submit your job from the temporary directory ($TMPDIR), which is faster to access for the system and might be beneficial for bigger jobs. Note that $TMPDIR is uniquely associated with the job submitted and will be cleared when the job ends. So you need to copy your results back to your work directory at the end of your script.

Non-interactive Batch Job (Parallel Run)

Please follow the steps below to use LS-DYNA via the batch system:

1) copy your input files (explorer.k in the example below) to your work directory at OSC

2) create a batch script, similar to the following file, saved as job.txt. It uses the mmp solver for a parallel job (nodes>1) on Oakley:

#PBS -N plate_test
#PBS -l walltime=5:00:00
#PBS -l nodes=2:ppn=12
#PBS -j oe
# The following lines set up the LSDYNA environment
module swap mvapich2/1.7 intelmpi
module load mpp-dyna/971_d_R7.0.0
#
# Move to the directory where the input files are located
#
cd $PBS_O_WORKDIR
#
# Run LSDYNA (number of cpus > 1)
#
mpiexec mpp971 I=explorer.k NCPU=24

3) submit the script to the batch queue with the command: qsub job.txt.

When the job is finished, all the result files will be found in the directory where you submitted your job ($PBS_O_WORKDIR). Alternatively, you can submit your job from the temporary directory ($TMPDIR), which is faster to access for the system and might be beneficial for bigger jobs. Note that $TMPDIR is uniquely associated with the job submitted and will be cleared when the job ends. So you need to copy your results back to your work directory at the end of your script. An example scrip should include the following lines:

...
cd $TMPDIR
cp $PBS_O_WORKDIR/explorer.k .
... #launch the solver and execute
pbsdcp -g '*' $PBS_O_WORKDIR

Usage on Owens

Set-up on Owens

To view available modules installed on Owens, use module spider ls-dyna for smp solvers, and use module spider mpp for mpp solvers. In the module name, '_s' indicates single precision and '_d' indicates double precision. For example, mpp-dyna/971_d_R9.0.1 is the mpp solver with double precision on Owens. Use module load name to load LS-DYNA with a particular software version. For example, use module load mpp-dyna/971_d_R9.0.1 to load LS-DYNA mpp solver version 9.0.1 with double precision on Owens.

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=00:20:00

which requests one whole node with 28 cores (-l nodes=1:ppn=28), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Please follow the steps below to use LS-DYNA via the batch system:

1) copy your input files (explorer.k in the example below) to your work directory at OSC

2) create a batch script, similar to the following file, saved as job.txt. It uses the smp solver for a serial job (nodes=1) on Owens:

#PBS -N plate_test
#PBS -l walltime=5:00:00
#PBS -l nodes=1:ppn=28
#PBS -j oe
# The following lines set up the LSDYNA environment
module load ls-dyna/971_d_9.0.1
#
# Move to the directory where the input files are located
#
cd $PBS_O_WORKDIR
#
# Run LSDYNA (number of cpus > 1)
#
lsdyna I=explorer.k NCPU=28

3) submit the script to the batch queue with the command: qsub job.txt.

When the job is finished, all the result files will be found in the directory where you submitted your job ($PBS_O_WORKDIR). Alternatively, you can submit your job from the temporary directory ($TMPDIR), which is faster to access for the system and might be beneficial for bigger jobs. Note that $TMPDIR is uniquely associated with the job submitted and will be cleared when the job ends. So you need to copy your results back to your work directory at the end of your script.

Non-interactive Batch Job (Parallel Run)

Please follow the steps below to use LS-DYNA via the batch system:

1) copy your input files (explorer.k in the example below) to your work directory at OSC

2) create a batch script, similar to the following file, saved as job.txt). It uses the mmp solver for a parallel job (nodes>1) on Owens:

#PBS -N plate_test
#PBS -l walltime=5:00:00
#PBS -l nodes=2:ppn=28
#PBS -j oe
# The following lines set up the LSDYNA environment
module swap mvapich2/2.2 intelmpi
module load mpp-dyna/971_d_R9.0.1
#
# Move to the directory where the input files are located
#
cd $PBS_O_WORKDIR
#
# Run LSDYNA (number of cpus > 1)
#
mpiexec mpp971 I=explorer.k NCPU=56

3) submit the script to the batch queue with the command: qsub job.txt.

When the job is finished, all the result files will be found in the directory where you submitted your job ($PBS_O_WORKDIR). Alternatively, you can submit your job from the temporary directory ($TMPDIR), which is faster to access for the system and might be beneficial for bigger jobs. Note that $TMPDIR is uniquely associated with the job submitted and will be cleared when the job ends. So you need to copy your results back to your work directory at the end of your script. An example scrip should include the following lines:

...
cd $TMPDIR
cp $PBS_O_WORKDIR/explorer.k .
... #launch the solver and execute
pbsdcp -g '*' $PBS_O_WORKDIR

LS-PrePost

Introduction

LS-PrePost is an advanced pre and post-processor that is delivered free with LS-DYNA. The user interface is designed to be both efficient and intuitive. LS-PrePost runs on Windows, Linux, and Unix utilizing OpenGL graphics to achieve fast rendering and XY plotting. The latest builds can be downloaded from LSTC's FTP Site.

The prefered way of accessing LS-Prepost is through OnDemand's desktop application. This gives you a preconfigured enviornment with GPU acceleration enabled.

Availability

LS-PrePost is available on Oakley cluster. The versions currently available at OSC are:.

Version	Oakley
v3.2	X
v4.0	X*

*: Current default version

Usage

Usage on Oakley

Set-up on Oakley

To view available modules installed on Oakley, use module spider lspp . Use module load name to load LS-PrePost with a particular software version. For example, use module load lspp/3.2 to load LS-PrePost version 3.2 on Oakley.

Running LS-PrePost on Oakley through OnDemand's Destkop

Below are instructions on running LS-PrePost on Oakley through OnDemand desktop interface with GPU acceleration enabled. To run LS-PrePost with a slower X-tunneling procedure, see the specified instructions below.

Log in to OnDemand with your HPC username and password.
Launch Oakley Desktop from "Interactive Apps" menu.
Provide # of hours, use "1" for "Number of nodes", and choose "vis" as "Node type"
Once your job is ready, launch the job, open "Applications menu -> System tools -> terminal", and use the following commands to launch the application:

module load lspp/4.0
module load virtualgl 
vglrun lspp4

At startup LS-PrePost displays a graphical interface for model generation and results post-processing.

Running LS-PrePost on Oakley with X-11 forwarding

The following procedure will result in a much slower and GUI interface, but may be useful if the above instructions are not working. It can be done completely from the command line, with no need for logging into OnDemand. You may need to edit your terminal settings to enabled x11 forwarding.

Login to Oakley with X11 forwarding enabled: ssh -X username@oakley.osc.edu
Submit a Interactive job: qsub -I -X -l nodes=1:ppn=12 -l walltime=hh:mm:ss -A ProjectID
Once the job is stared, load the LS-PrePost module and start up LS-PrePost application:

module load lspp/4.0
lspp4

A x11 window should pop up. If you get a error along the lines of:

Error: Unable to initialize gtk, is DISPLAY set properly?

Double check that you:

1) logged in with x11 forwarding enabled

2) have configured your x11 settings for your terminal

3) included the -X with the qsub command

4) have a x11 client running on your computer (Xming, Xorg, XQuarts, etc.).

Documentation

Documentation for LS-PrePost may be obtained at: http://www.lstc.com/lspp/

Fields of Science:

Aerospace Engineering

Civil Engineering

Manufacturing

Materials

Mechanical Engineering

User-Defined Material for LS-DYNA

This page describes how to specify user defined material to use within LS-DYNA. The user-defined subroutines in LS-DYNA allow the program to be customized for particular applications. In order to define user material, LS-DYNA must be recompiled.

Usage

The first step to running a simulation with user defined material is to build a new executable. The following is an example done with solver version mpp971_s_R7.1.1.

When you log into the Oakley system, load mpp971_s_R7.1.1 with the command:

module load mpp-dyna/R7.1.1

Next, copy the mpp971_s_R7.1.1 object files and Makefile to your current directory:

cp /usr/local/lstc/mpp-dyna/R7.1.1/usermat/* $PWD

Next, update the dyn21.f file with your user defined material model subroutine. Please see the LS-DYNA User's Manual (Keyword version) for details regarding the format and structure of this file.

Once your user defined model is setup correctly in dyn21.f, build the new mpp971 executable with the command:

make

To execute a multi processor (ppn > 1) run with your new executable, execute the following steps:

1) move your input file to a directory on an OSC system (pipe.k in the example below)

2) copy your newly created mpp971 executable to this directory as well

3) create a batch script (lstc_umat.job) like the following:

#PBS -N LSDYNA_umat
#PBS -l walltime=1:00:00
#PBS -l nodes=2:ppn=8
#PBS -j oe
#PBS -S /bin/csh

# This is the template batch script for running a pre-compiled
# MPP 971 v7600 LS-DYNA.  
# Total number of processors is ( nodes x ppn )
#
# The following lines set up the LSDYNA environment
module load mpp-dyna/R7.1.1
#
# Move to the directory where the job was submitted from
# (i.e. PBS_O_WORKDIR = directory where you typed qsub)
#
cd $PBS_O_WORKDIR
#
# Run LSDYNA 
# NOTE: you have to put in your input file name
#
mpiexec mpp971 I=pipe.k NCPU=16

4) Next, submit this job to the batch queue with the command:

       qsub lstc_umat.job

The output result files will be saved to the directory you ran the qsub command from (known as the $PBS_O_WORKDIR_

Documentation

On-line documentation is available on LSTC website.

MAGMA

MAGMA is a collection of next generation linear algebra (LA) GPU accelerated libraries designed and implemented by the team that developed LAPACK and ScaLAPACK. MAGMA is for heterogeneous GPU-based architectures, it supports interfaces to current LA packages and standards, e.g., LAPACK and BLAS, to allow computational scientists to effortlessly port any LA-relying software components. The main benefits of using MAGMA are that it can enable applications to fully exploit the power of current heterogeneous systems of multi/manycore CPUs and multi-GPUs, and deliver the fastest possible time to an accurate solution within given energy constraints.

Availability and Restrictions

Versions

MAGMA is available on Owens, and the following versions are currently available at OSC:

Version	Owens
2.2.0	X(I)*

* Current default version; I = avaible with only intel

You can use module spider magma to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

MAGMA is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To load the default version of MAGMA module, cuda must first be loaded. Use module load cuda toload the default version of cuda, or module load cuda/version to load a specific version. Then use module load magma to load MAGMA. To select a particular software version, use module load magma/version. For example, use module load magma/2.2.0 to load MAGMA version 2.2.0

Using MAGMA

To run MAGMA in the command line, use the Intel compilers (icc, ifort).

icc $MAGMA_CFLAGS example.c

or

ifort $MAGMA_F90FLAGS example.F90

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your MAGMA simulation to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice.

Below is the example batch script (job.txt) for a serial run:

# MAGMA Example Batch Script for the Basic Tutorial in the MAGMA manual
#PBS -N 6pti
#PBS -l nodes=1:ppn=28
#PBS -l walltime=0:20:00
module load cuda
module load magma
# Use TMPDIR for best performance.
cd $TMPDIR
# PBS_O_WORKDIR refers to the directory from which the job was submitted.
cp $PBS_O_WORKDIR/example.c .
icc $MAGMA_CFLAGS example.c

In order to run it via the batch system, submit the job.txt file with the command: qsub job.txt.

MATLAB

MATLAB is a technical computing environment for high-performance numeric computation and visualization. MATLAB integrates numerical analysis, matrix computation, signal processing, and graphics in an easy-to-use environment where problems and solutions are expressed just as they are written mathematically--without traditional programming.

Availability and Restrictions

Versions

MATLAB is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	Owens	notes
2013a	X			Default version on Oakley prior to 09/15/2015
2013b	X
2014a	X	X		Default version on Ruby prior to 09/15/2015
2014b	X	X
2015a	X	X
2015b	X	X	X
2016b	X*	X*	X*
2017a 2018a	X	X X	X X

* Current default version

You can use module spider matlab to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access: Academic Users Only (non-commercial, non-government)

Any academic users at OSC can use Matlab. All users must be added to the license server before using MATLAB. Please contact OSC Help to be granted access.

Usage

Usage on Oakley

Set-up

To load the default version of MATLAB module, use module load matlab . For a list of all available MATLAB versions and the format expected, type: module spider matlab . To select a particular software version, use module load matlab/version . For example, use module load matlab/r2014b to load MATLAB version 2014b.

Running MATLAB

The following command will start an interactive, command line version of MATLAB:

matlab -nodisplay

If you are able to use X-11 forwarding and have enabled it in your SSH client software preferences, you can run MATLAB using the GUI by typing the command matlab . For more information about the matlab command usage, type matlab –h for a complete list of command line options.

The commands listed above will run MATLAB on the login node you are connected to. As the login node is a shared resource, running scripts that require significant computational resources will impact the usability of the cluster for others. As such, you should not use interactive MATLAB sessions on the login node for any significant computation. If your MATLAB script requires significant time, CPU power, or memory, you should run your code via the batch system.

Batch Usage

When you log into oakley.osc.edu you are actually logged into a Linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session using the command line version of MATLAB, one can run the following command:

qsub -I -X -l nodes=1:ppn=12 -l walltime=00:20:00

which requests one whole node with 12 cores ( -l nodes=1:ppn=12 ), for a walltime of 20 minutes ( -l walltime=00:20:00 ). Here you can run MATLAB interactively by loading the MATLAB module and running MATLAB with the options of your choice as described above. The –X flag enables X11 forwarding on the compute node, so you can use the MATLAB GUI if you choose.You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

In order to run MATLAB non-interactively via the batch system, you will require a batch submission script and a MATLAB M-file containing the script that will be run via the batch system. A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is an example batch submission script and a simple M-file. The batch script runs the M-file via the batch system.Example batch script ( job.txt ) for a serial run:

#PBS -N matlab_example
#PBS -l walltime=00:10:00
#PBS -l nodes=1:ppn=12
#PBS -j oe
module load matlab
matlab -nodisplay -nodesktop < hello.m
# end of example file

Example M-file, hello.m:

%Example M-file for Hello World
disp 'Hello World' 
exit   
% end of example file

In order to run hello.m via the batch system, submit the job.txt file with the following command:

qsub job.txt

This will run hello.m via the batch system, and all output from the running of the script will be saved in the output.txt file.

Parallel Processing in MATLAB

MATLAB supports both implicit multiprocessing (multithreading) and explicit multiprocessing across multiple nodes.

Multithreading

Multithreading allows some functions in MATLAB to distribute the work load between cores of the node that your job is running on. By default, all of the current versions of MATLAB available on the OSC clusters have multithreading enabled.

The system will use a number of threads equal to the number of cores you request. Therefore, if you request nodes=1:ppn=4, your job will only spawn four threads.

Multithreading increases the speed of some linear algebra routines, but if you would like to disable multithreading you may request nodes=1:ppn=1 and include the option -singleCompThread when running MATLAB. An example is given below:

#PBS -N disable_multithreading
#PBS -l walltime=00:10:00
#PBS -l nodes=1:ppn=1
#PBS -j oe
module load matlab
matlab -singleCompThread -nodisplay -nodesktop < hello.m
# end of example file

NOTE:

From Apr 1, 2018, we don't support for Matlab Distributed Computing Server (MDCS). If you have any questions, please contact OSC Help .

Usage on Ruby

Set-up

To load the default version of MATLAB module, use module load matlab . For a list of all available MATLAB versions and the format expected, type: module spider matlab . To select a particular software version, use module load matlab/version . For example, use module load matlab/r2014b to load MATLAB version 2014b.

Running MATLAB

The following command will start an interactive, command line version of MATLAB:

matlab -nodisplay

If you are able to use X-11 forwarding and have enabled it in your SSH client software preferences, you can run MATLAB using the GUI by typing the command matlab . For more information about the matlab command usage, type matlab –h for a complete list of command line options.

The commands listed above will run MATLAB on the login node you are connected to. As the login node is a shared resource, running scripts that require significant computational resources will impact the usability of the cluster for others. As such, you should not use interactive MATLAB sessions on the login node for any significant computation. If your MATLAB script requires significant time, CPU power, or memory, you should run your code via the batch system.

Batch Usage

When you log into ruby.osc.edu you are actually logged into a Linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session using the command line version of MATLAB, one can run the following command:

qsub -I -X -l nodes=1:ppn=20 -l walltime=00:20:00

which requests one whole node with 20 cores ( -l nodes=1:ppn=20 ), for a walltime of 20 minutes ( -l walltime=00:20:00 ). Here you can run MATLAB interactively by loading the MATLAB module and running MATLAB with the options of your choice as described above. The –X flag enables X11 forwarding on the compute node, so you can use the MATLAB GUI if you choose.You may adjust the numbers per your need.

Parallel Processing in MATLAB

MATLAB supports both implicit multiprocessing (multithreading) and explicit multiprocessing across multiple nodes.

Multithreading

Multithreading allows some functions in MATLAB to distribute the work load between cores of the node that your job is running on. By default, all of the current versions of MATLAB available on the OSC clusters have multithreading enabled.

The system will run as many threads as there are cores on the nodes requested. For the normal nodes, this mean matlab will start 20 threads to run on the 20 cores of the node. This is due to the whole node scheduling policy of Ruby (see Batch Specifics).

Multithreading increases the speed of some linear algebra routines, but if you would like to disable multithreading you may include the option " -singleCompThread " when running MATLAB. An example is given below:

#PBS -N disable_multithreading
#PBS -l walltime=00:10:00
#PBS -l nodes=1:ppn=20
#PBS -j oe
module load matlab
matlab -singleCompThread -nodisplay -nodesktop < hello.m
# end of example file

NOTE:

From Apr 1, 2018, we don't support for Matlab Distributed Computing Server (MDCS). If you have any questions, please contact OSC Help .

Usage on Owens

Set-up

To load the default version of MATLAB module, use module load matlab . For a list of all available MATLAB versions and the format expected, type: module spider matlab . To select a particular software version, use module load matlab/version . For example, use module load matlab/r2015b to load MATLAB version r2015b.

Running MATLAB

The following command will start an interactive, command line version of MATLAB:

matlab -nodisplay

If you are able to use X-11 forwarding and have enabled it in your SSH client software preferences, you can run MATLAB using the GUI by typing the command matlab . For more information about the matlab command usage, type matlab –h for a complete list of command line options.

The commands listed above will run MATLAB on the login node you are connected to. As the login node is a shared resource, running scripts that require significant computational resources will impact the usability of the cluster for others. As such, you should not use interactive MATLAB sessions on the login node for any significant computation. If your MATLAB script requires significant time, CPU power, or memory, you should run your code via the batch system.

Batch Usage

When you log into owens.osc.edu you are actually logged into a Linux box referred to as the login node. To gain access to the multiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session using the command line version of MATLAB, one can run the following command:

qsub -I -X -l nodes=1:ppn=28 -l walltime=00:20:00

which requests one whole node with 28 cores ( -l nodes=1:ppn=28 ), for a walltime of 20 minutes ( -l walltime=00:20:00 ). Here you can run MATLAB interactively by loading the MATLAB module and running MATLAB with the options of your choice as described above. The –X flag enables X11 forwarding on the compute node, so you can use the MATLAB GUI if you choose.You may adjust the numbers per your need.

Parallel Processing in MATLAB

MATLAB supports both implicit multiprocessing (multithreading) and explicit multiprocessing across multiple nodes.

Multithreading

Multithreading allows some functions in MATLAB to distribute the work load between cores of the node that your job is running on. By default, all of the current versions of MATLAB available on the OSC clusters have multithreading enabled.

The system will run as many threads as there are cores on the nodes requested. For the normal nodes, this mean matlab will start 28 threads to run on the 28 cores of the node. This is due to the whole node scheduling policy of Owens (see Batch Specifics).

Multithreading increases the speed of some linear algebra routines, but if you would like to disable multithreading you may include the option " -singleCompThread " when running MATLAB. An example is given below:

#PBS -N disable_multithreading
#PBS -l walltime=00:10:00
#PBS -l nodes=1:ppn=28
#PBS -j oe
module load matlab
matlab -singleCompThread -nodisplay -nodesktop < hello.m
# end of example file

NOTE:

From Apr 1, 2018, we don't support for Matlab Distributed Computing Server (MDCS). If you have any questions, please contact OSC Help .

MATLAB with a GPU

A GPU can be utilized for MATLAB. You can acquire a GPU by nodes=1:ppn=20;gpus=1 for Ruby or Owens (ppn=28 for Owens). For more detail, please read here.

You can check the GPU assigned to you using:

gpuDeviceCount  # show how many GPUs you have
gpuDevice       # show the details of the GPU

You can replace an array to gpuArray, for example:

A = gpuArray(A)

where A is a regular MATLAB array. This transfers the array data to the GPU memory. Then, you can use the gpuArray variable in GPU supported built-in functions. You can find the full list of GPU supported built-in functions from here. For more information about GPU programming for MATLAB, please read "GPU Computing" from Mathworks.

Toolboxes and Features

OSC's current licenses support the following MATLAB toolboxes and features (please contact OSC Help for license-specific questions):

Polyspace Bug Finder
PolySpace_Bug_Finder_Engine
Aerospace_Blockset
Aerospace_Toolbox
Antenna_Toolbox
Audio_System_Toolbox
Automated_Driving_Toolbox
Bioinformatics_Toolbox
Communications System Toolbox
Computer Vision System Toolbox
Control System Toolbox
Curve_Fitting_Toolbox
DSP System Toolbox
Data Acquisition Toolbox
Database_Toolbox
Datafeed_Toolbox
Econometrics_Toolbox
RTW_Embedded_Coder
Filter_Design_HDL_Coder
Financial Instruments Toolbox
Financial_Toolbox
Fixed_Point_Toolbox
Fuzzy Logic Toolbox
Global Optimization Toolbox
Simulink_HDL_Coder
EDA_Simulator_Link
Image_Acquisition_Toolbox
Image Processing Toolbox
Instrument Control Toolbox
LTE System Toolbox
MATLAB_Coder
MATLAB_Builder_for_Java
Compiler
MATLAB Report Generator
Mapping Toolbox
Model Predictive Control Toolbox
Model-Based Calibration Toolbox
Neural_Network_Toolbox
OPC_Toolbox
Optimization_Toolbox
Distrib_Computing_Toolbox
Partial Differential Equation Toolbox
Phased_Array_System_Toolbox
Polyspace Code Prover
Powertrain_Blockset
RF_Blockset
RF_Toolbox
Risk_Management_Toolbox
Robotics_System_Toolbox
Robust Control Toolbox
Signal Processing Toolbox
SimBiology
SimEvents
Simscape Driveline
Simscape Electronics
Simscape Fluids
Simscape Multibody
Simscape Power Systems
Simscape
Virtual_Reality_Toolbox
Simulink_Code_Inspector
Real-Time_Workshop
Simulink_Control_Design
Simulink Design Optimization
Simulink_Design_Verifier
Simulink Desktop Real-Time
Simulink_PLC_Coder
XPC_Target
Simulink Report Generator
Simulink_Test
Simulink Verification and Validation
Excel_Link
Stateflow
Statistics and Machine Learning Toolbox
Symbolic Math Toolbox
System Identification Toolbox
Trading_Toolbox
Vehicle_Network_Toolbox
Vision_HDL_Toolbox
WLAN_System_Toolbox
Wavelet_Toolbox

MIRA

MIRA - Sequence assembler and sequence mapping for whole genome shotgun and EST / RNASeq sequencing data. Can use Sanger, 454, Illumina and IonTorrent data. PacBio: CCS and error corrected data usable, uncorrected not yet.

Availability and Restrictions

Versions

The following versions of MIRA are available on OSC clusters:

Version	Oakley	Owens
4.0.2	X*	X*

* Current default version

You can use module spider mira to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

MIRA is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of MIRA, run the following command: module load mira. The default version will be loaded. To select a particular MIRA version, use module load mira/version. For example, use module load mira/4.0.2 to load MIRA 4.0.2.

Usage on Owens

Set-up

To configure your environment for use of MIRA, run the following command: module load mira. The default version will be loaded. To select a particular MIRA version, use module load mira/version. For example, use module load mira/4.0.2 to load MIRA 4.0.2.

MKL - Intel Math Kernel Library

Intel Math Kernel Library (MKL) consists of high-performance, multithreaded mathematics libraries for linear algebra, fast Fourier transforms, vector math, and more.

Availability and Restrictions

Versions

OSC supports single-process use of MKL for LAPACK and BLAS levels one through three. For multi-process applications, we also support the ScaLAPACK, FFTW2, and FFTW3 MKL wrappers. MKL modules are available for the Intel, GNU, and PGI compilers. MKL is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	owens	notes
10.3.0	X			Default version on Oakley prior to 09/15/2015
11.1.0	X
11.2.0		X		Default version on Ruby prior to 09/15/2015
11.2.1		X
11.2.3	X	X
11.3.2			X
11.3.3	X*	X*	X*
2017.0.2			X

* Current Default Version

You can use module spider mkl to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

MKL is available to all OSC users.

Usage

Usage on Oakley

Set-up

To load the default MKL, run the following command: module load mkl. To load a particular version, use module load mkl/version. For example, use module load mkl/11.2.3 to load MKL version 11.2.3. You can use module spider mkl to view available modules.

This step is required for both building and running MKL applications.

Exception: The "mkl" module is usually not needed when using the Intel compilers; just use the "-mkl" flag on the compile and link steps.

Building With MKL

The MKL module will automatically configure your environment to locate the appropriate include files. When linking, however, you have a variety of options to choose from depending on your needs. These options come in the form of environment variables defined by the module. Which variable you include when linking will determine what particular configuration of MKL you get.

Environment Variable	Function
`$MKL_CFLAGS`	Include flags for C
`$MKL_FFLAGS`	Include flags for Fortran
`$MKL_LIBS`	Use multithreaded MKL with 32-bit integers. This is the standard configuration.
`$MKL_LIBS_INT64`	Use multithreaded MKL with 64-bit (“ILP64”) integers.
`$MKL_LIBS_SEQ`	Use single-threaded (“ `SEQ` uential”) MKL with with 32-bit integers.
`$MKL_LIBS_SEQ_INT64`	Use single-threaded MKL with 64-bit integers.

Notes:

64-bit integers are not supported by the FFTW2 wrappers. (See below.)
The Intel compilers are specially configured for Intel MKL. If you are using this toolchain, you can use the "-mkl” option when linking in place of $MKL_LIBS.
The default, multithreaded MKL libraries will automatically run in a single thread if they are called from an OpenMP parallel region. If you want to force single-threaded behavior throughout your program, choose one of the “_SEQ” variables from the list above.

Fortran 95 BLAS/LAPACK Wrappers

To compile Fortran 95 programs with modules, add the $MKL_F95FLAGS variable to your compilation step. If you need 64-bit integers, use the $MKL_F95_FLAGS_INT64 instead. When linking, you will also need to use $MKL_F95LIBS (or $MKL_F95LIBS_INT64 if using 64-bit integers). These variables will allow your programs to use the BLAS and LAPACK wrappers for MKL.

FFTW Wrappers

A number of “wrappers” are provided in the form of environment variables that allow you to use FFTW APIs with MKL. Variables ending in FLAGS should be included with your compilation step, while variables ending in LIBS should be included in your linking step.

Environment Variable	Function
`$MKL_FFTW_CFLAGS`	Compile variable for C programs.
`$MKL_FFTW_FFLAGS`	Compile variable for Fortran programs.
`$MKL_FFTW2_D_CLIBS`	Linking variable for double-precision FFTW2 C programs.
`$MKL_FFTW2_D_FLIBS`	Linking variable for double-precision FFTW2 Fortran programs.
`$MKL_FFTW2_S_CLIBS`	Linking variable for single-precision FFTW2 C programs.
`$MKL_FFTW2_S_FLIBS`	Linking variable for single-precision FFTW2 Fortran programs.
`$MKL_FFTW3_CLIBS`	Linking variable for FFTW3 C programs.
`$MKL_FFTW3_FLIBS`	Linking variable for FFTW3 Fortran programs.

Notes:

FFTW3 is the default wrapper and should be picked up by your linker automatically. If you would prefer to use FFTW2, use the appropriate linking variable.
The FFTW2 wrappers do not support 64-bit (“ILP64”) integers.

Examples

The following examples use the Intel compilers, though they should work using the GNU and Portland Group compilers as well.

Compiling a Fortran77 program with MKL:

ifort -c my_prog.f
ifort -o my_prog my_prog.o $MKL_LIBS

Compiling a C program with MKL:

icc -c my_prog.c
icc -o my_prog my_prog.o $MKL_LIBS

Compiling a module-based Fortran90/95 program with MKL:

ifort -c my_prog.f95 $MKL_F95FLAGS
ifort -o my_prog my_prog.o $MKL_F95LIBS $MKL_LIBS

Compiling a C program with the double-precision FFTW2 wrapper:

icc -c $MKL_FFTW_CFLAGS -DMKL_DOUBLE my_prog.c
icc -o my_prog my_prog.o $MKL_FFTW2_D_CLIBS $MKL_LIBS

Compiling a Fortran77 program with the FFTW2 wrapper:

ifort -c $MKL_FFTW_FFLAGS my_prog.f
ifort -o my_prog my_prog.o $MKL_FFTW2_S_FLIBS $MKL_LIBS

Compiling a C program with the FFTW3 wrapper:

icc -c $MKL_FFTW_CFLAGS my_prog.c
icc -o my_prog my_prog.o $MKL_FFTW3_CLIBS $MKL_LIBS

Compiling a Fortran program with the FFTW3 wrapper:

ifort -c $MKL_FFTW_FFLAGS my_prog.f 
ifort -o my_prog my_prog.o $MKL_FFTW3_CLIBS $MKL_LIBS

Advanced Usage

If you are already familiar with building MKL applications with your chosen build tool (GCC, Intel, or PGI) and you do not wish to use the convenience variables discussed above, you may wish to use the $MKLROOT variable instead. This variable points to the installation directory for Intel MKL. All include files can be found in $MKLROOT/include, for example, and the libraries are in $MKLROOT/lib/intel64.

Running MKL Programs

When running an MKL program, you need to be sure to take the following steps.

Load the mkl module:
```
module load mkl
```
If running with parallel MKL, set OMP_NUM_THREADS to match the number of cores per node in your process. In the bash shell, you can accomplish this with:
```
PPNPERNODE=$(expr $(cat $PBS_NODEFILE | wc -l) / $(uniq $PBS_NODEFILE | wc -l))
export OMP_NUM_THREADS=$PPNPERNODE
```

Usage on Ruby

Set-up

To load the default MKL, run the following command: module load mkl. To load a particular version, use module load mkl/version. For example, use module load mkl/11.2.3 to load MKL version 11.2.3. You can use module spider mkl to view available modules.

This step is required for both building and running MKL applications.

Exception: The "mkl" module is usually not needed when using the Intel compilers; just use the "-mkl" flag on the compile and link steps.

Usage on Owens

Set-up

To load the default MKL, run the following command: module load mkl. To load a particular version, use module load mkl/version. For example, use module load mkl/11.3.3 to load MKL version 11.3.3. You can use module spider mkl to view available modules.

This step is required for both building and running MKL applications.

Exception: The "mkl" module is usually not needed when using the Intel compilers; just use the "-mkl" flag on the compile and link steps.

MPJ-Express

MPJ-Express is a Java library that provides message passing capabilities for parallel computing in Java applications.

Availability and Restrictions

Versions

The following versions of MPJ-Express are available on OSC systems:

Version	Oakley
0.38	X
0.44	X*

* Current default version

You can use module spider mpj-express to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Note: We have modified MPJ-Express slightly from the default distribution to work nicely with our infrastructure. Installing your own copy of a newer version will not work. Please contact OSC Help if you need a newer version of this library.

Access

MPJ-Express is available without restriction to all OSC users.

Usage

Set-up

To set up your environment for using the MPJ-Express libraries, you must load the appropriate module. On Oakley this is pretty straightforward:

module load mpj-express

Building With MPJ-Express

When you load the MPJ-Express module, we will modify your CLASSPATH to ensure the JAR files are located.

Batch Usage

Programs built with MPJ-Express can only be run in the batch environment at OSC. For information on writing batch scripts, see Batch Processing at OSC.

If you have loaded a Java module when compiling, be sure to load that same module in your batch script to avoid version mismatches.

Here is a basic "Hello World" program, which needs to be in a file called "HelloWorld.java", which we use in our example batch script:

import mpi.*;
public class HelloWorld
{
     public static void main(String[] args) 
                     throws Exception
     {
          MPI.Init(args);
          int me = MPI.COMM_WORLD.Rank();
          int size = MPI.COMM_WORLD.Size();
          System.out.println
            ("Hi from " + me + " of " + size + "\n");
          MPI.Finalize();
     }
}

Below we have included a sample batch script:

#PBS -N mpj-world
#PBS -l walltime=0:05:00
#PBS -l nodes=2:ppn=12
#PBS -S /bin/bash

# Set environment
module load mpj-express/0.44

# Change to the directory you submitted the job from
cd $PBS_O_WORKDIR

# Set up the directory for the daemon logs to go in
export MPJ_LOG="$PBS_O_WORKDIR/log"
mkdir $MPJ_LOG

# Re-compile our application - not strictly necessary,
# if the source is unchanged.
javac HelloWorld.java

# Start MPJ
mpjboot $PBS_NODEFILE

# Launch your parallel program
mpjrun.sh -np $PBS_NP -dev native HelloWorld

# Shutdown MPJ
mpjhalt $PBS_NODEFILE

MVAPICH2

MVAPICH2 is a standard library for performing parallel processing using a distributed-memory model.

Availability and Restrictions

Versions

The following versions of MVAPICH2 are available on OSC systems:

Version	Oakley	Ruby	Owens
1.7	X
1.8.1	X
1.9	X	X
2.1	X	X	X
2.2	X*	X*	X*
2.3b			X

* Current default version

You can use module spider mvapich2 to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

MPI is available without restriction to all OSC users.

Usage

Set-up

To set up your environment for using the MPI libraries, you must load the appropriate module:

module load mvapich2

You will get the default version for the compiler you have loaded.

Note:Be sure to swap the intel compiler module for the gnu module if you're using gnu.

Building With MPI

To build a program that uses MPI, you should use the compiler wrappers provided on the system. They accept the same options as the underlying compiler. The commands are shown in the following table.

C	`mpicc`
C++	`mpicxx`
FORTRAN 77	`mpif77`
Fortran 90	`mpif90`

For example, to build the code my_prog.c using the -O2 option, you would use:

mpicc -o my_prog -O2 my_prog.c

In rare cases you may be unable to use the wrappers. In that case you should use the environment variables set by the module.

Variable	Use
`$MPI_CFLAGS`	Use during your compilation step for C programs.
`$MPI_CXXFLAGS`	Use during your compilation step for C++ programs.
`$MPI_FFLAGS`	Use during your compilation step for Fortran 77 programs.
`$MPI_F90FLAGS`	Use during your compilation step for Fortran 90 programs.
`$MPI_LIBS`	Use when linking your program to the MPI libraries.

For example, to build the code my_prog.c without using the wrappers you would use:

mpicc -c $MPI_CFLAGS my_prog.c

mpicc -o my_prog my_prog.o $MPI_LIBS

Batch Usage

Programs built with MPI can only be run in the batch environment at OSC. For information on starting MPI programs using the mpiexec command, see Batch Processing at OSC.

Be sure to load the same compiler and mvapich modules at execution time as at build time.

Maven (Apache)

Apache Maven is a plugin-based build automation tool, similar in purpose to GNU Make or Apache Ant. It is most commonly used with Java projects, but also supports other languages such as .NET and C/C++.

Availability & Compatibility

Versions

Version	Oakley
3.0.4	X*

* Current default version

You can use module spider maven to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

There are no restrictions for this software; any OSC user may make use of the Apache Maven.

Usage

Set-up

To use Maven, load the the "maven" module with the following command.

module load maven

To test that the install worked correctly, run "mvn --version". You should see output similar to that shown below:

$ mvn --version
Apache Maven 3.0.4 (r1232337; 2012-01-17 03:44:56-0500)
Maven home: /usr/local/maven/3.0.4
Java version: 1.6.0_29, vendor: Sun Microsystems Inc.
Java home: /usr/lib/jvm/java-1.6.0-sun-1.6.0.29.x86_64/jre
Default locale: en_US, platform encoding: UTF-8
OS name: "linux", version: "2.6.32-131.17.1.el6.x86_64", arch: "amd64", family: "unix"

Basic Usage

Once the module is loaded, you can use Maven just as you would on your local machine. For example, the session below illustrates initializing a new project call "my-app":

$ mvn archetype:generate -DgroupId=com.mycompany.app -DartifactId=my-app -DarchetypeArtifactId=maven-archetype-quickstart -DinteractiveMode=false
[INFO] Scanning for projects...
[INFO]                                                                         
[INFO] ------------------------------------------------------------------------
[INFO] Building Maven Stub Project (No POM) 1
[INFO] ------------------------------------------------------------------------
[INFO] 
[INFO] >>> maven-archetype-plugin:2.2:generate (default-cli) @ standalone-pom >>>
[INFO] 
[INFO] <<< maven-archetype-plugin:2.2:generate (default-cli) @ standalone-pom <<<
[INFO] 
[INFO] --- maven-archetype-plugin:2.2:generate (default-cli) @ standalone-pom ---
[INFO] Generating project in Batch mode
[INFO] ----------------------------------------------------------------------------
[INFO] Using following parameters for creating project from Old (1.x) Archetype: maven-archetype-quickstart:1.0
[INFO] ----------------------------------------------------------------------------
[INFO] Parameter: groupId, Value: com.mycompany.app
[INFO] Parameter: packageName, Value: com.mycompany.app
[INFO] Parameter: package, Value: com.mycompany.app
[INFO] Parameter: artifactId, Value: my-app
[INFO] Parameter: basedir, Value: /nfs/12/jmccance
[INFO] Parameter: version, Value: 1.0-SNAPSHOT
[INFO] project created from Old (1.x) Archetype in dir: /nfs/12/jmccance/my-app
[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 4.429s
[INFO] Finished at: Fri Jun 08 14:30:57 EDT 2012
[INFO] Final Memory: 9M/361M
[INFO] ------------------------------------------------------------------------

MrBayes

MrBayes is a program for the Bayesian estimation of phylogeny. Bayesian inference of phylogeny is based upon a quantity called the posterior probability distribution of trees, which is the probability of a tree conditioned on the observations. The conditioning is accomplished using Bayes's theorem. The posterior probability distribution of trees is impossible to calculate analytically; instead, MrBayes uses a simulation technique called Markov chain Monte Carlo (or MCMC) to approximate the posterior probabilities of trees.

Availability and Restrictions

Versions

The following versions of MrBayes are available:

Version	Oakley
3.2.1	X*

* Current default version

You can use module spider mrbayes to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

MrBayes is available without restriction to all OSC users.

Usage

Set-up

To load the MrBayes software on the Oakley system, use the following command:

module load mrbayes

Using MrBayes

Once mrbayes module is loaded, the commands are available for your use. The command for serial version is

mb

Parallel Mrbayes can only be used via batch system. The command is

mb-parallel

Batch Usage

Below is a sample batch script for running serial mrbayes job. It asks for 1 processor and 30 minutes of walltime.

#PBS -l walltime=30:00
#PBS -l nodes=1:ppn=1
#PBS -N mb
#PBS -j oe
set -x
date
cd $PBS_O_WORKDIR
cp ./primates.nex $TMPDIR
cp ./primates.nxs $TMPDIR
cd $TMPDIR
module load mrbayes/3.2.1
/usr/bin/time mb ./primates.nxs > mb.log
cp * $PBS_O_WORKDIR
date

Below is a sample batch script for running a parallel job on Oakley. It asks for 2 nodes, 24 processors and 30 minutes of walltime.

#PBS -l walltime=30:00
#PBS -l nodes=2:ppn=12
#PBS -N mb-parallel
#PBS -j oe
set -x
date
cd $PBS_O_WORKDIR
pbsdcp ./primates.nex $TMPDIR
pbsdcp ./primates.nxs $TMPDIR
cd $TMPDIR
module load mrbayes
/usr/bin/time mpiexec mb-parallel ./primates.nxs > mb-paralllel.log
pbsdcp -g '*' $PBS_O_WORKDIR
date

MuTect

MuTect is a method developed at the Broad Institute for the reliable and accurate identification of somatic point mutations in next generation sequencing data of cancer genomes.

Availability and Restrictions

Versions

The following versions of MuTect are available on OSC clusters:

Version	Oakley	owens
1.1.4	X
1.1.7	X*	X*

* Current default version

You can use module spider mutect to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

MuTect is available to academic OSC users with proper validation. In order to obtain validation, please contact OSC Help for further instruction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of MuTect, run the following command: module load mutect. The default version will be loaded. To select a particular MuTect version, use module load mutect/version. For example, use module load mutect/1.1.4 to load MuTect 1.1.4.

NOTE: Java will also need to be loaded in order to use MuTect: module load java.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load mutect, a new environment variable, MUTECT, will be set.

Thus, users can use the software by running the following command: java -jar $MUTECT {other options}.

Usage on Owens

Set-up

To configure your environment for use of MuTect, run the following command: module load mutect. The default version will be loaded. To select a particular MuTect version, use module load mutect/version. For example, use module load mutect/1.1.4 to load MuTect 1.1.4.

NOTE: Java 1.7.0 will also need to be loaded in order to use MuTect: module load java/1.7.0.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load mutect, a new environment variable, MUTECT, will be set.

Thus, users can use the software by running the following command: java -jar $MUTECT {other options}.

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD generally scales well on OSC platforms and offers a variety of modelling techniques. NAMD is file-compatible with AMBER, CHARMM, and X-PLOR.

Availability and Restrictions

Versions

The following versions of NAMD are available:

Version	Oakley	ruby	Owens
2.9	X
2.10	X
2.11	X		X
2.12	X*	X*	X*

* Current default version

You can use module spider namd to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

NAMD is available without restriction to all OSC users.

Usage

Set-up

To load the NAMD software on the system, use the following command: module load namd/"version" where "version" is the version of NAMD you require. The following will load the default or latest version of NAMD: module load namd.

Using NAMD

NAMD is rarely executed interactively because preparation for simulations is typically performed with extraneous tools, such as, VMD.

Batch Usage

Sample batch scripts and input files are available here:

~srb/workshops/compchem/namd/

The simple batch script for Oakley below demonstrates some important points. It requests 24 processors and 2 hours of walltime. If the job goes beyond 2 hours, the job would be terminated.

#PBS -N apoa1
#PBS -l nodes=2:ppn=12
#PBS -l walltime=2:00:00
#PBS -S /bin/bash
#PBS -j oe
module load namd
cd $PBS_O_WORKDIR
pbsdcp -p apoa1.namd apoa1.pdb apoa1.psf *.xplor $TMPDIR
cd $TMPDIR
run_namd apoa1.namd
pbsdcp -pg '*' $PBS_O_WORKDIR

NOTE: ppn should =20 on Ruby and =28 on Owens.

GPU support

We have GPU support with NAMD 2.12 for Owens and Ruby clusters, and with NAMD 2.11 for Oakley. These temporarily use pre-compiled binaries due to installation issues. For more detail, please read the corresponding example script:

~srb/workshops/compchem/namd/apoa1.namd212nativecuda.owens.pbs  # for Owens
~srb/workshops/compchem/namd/apoa1.namd212nativecuda.ruby.pbs   # for Ruby
~srb/workshops/compchem/namd/apoa1.namd211nativecuda.oakley.pbs  # for Oakley

NCL/NCARG

NCAR Graphics is a Fortran and C based software package for scientific visualization. NCL (The NCAR Command Language), is a free interpreted language designed specifically for scientific data processing and visualization. It is a product of the Computational & Information Systems Laboratory at the National Center for Atmospheric Research (NCAR) and sponsored by the National Science Foundation. NCL has robust file input and output: it can read and write netCDF-3, netCDF-4 classic, HDF4, binary, and ASCII data, and read HDF-EOS2, GRIB1, and GRIB2. The graphics are based on NCAR Graphics.

Availability and Restrictions

Versions

NCL/NCAR Graphics is available on Oakley and Owens Cluster. The versions currently available at OSC are:

version	oakley	Owens
6.0.0	X
6.2.1	X
6.3.0	X(GI)*	X(GI)*

* Current default version; G = available with gnu; I = available with intel

You can use module spider ncarg to view available NCL/NCAR Graphics modules. Feel free to contact OSC Help if you need other versions for your work.

Access

NCL/NCAR Graphics is available for use by all OSC users.

Usage

Usage on Oakley

Set-up on Oakley

To load the default version, first load the cairo module, then use either module load ncarg. To select a particular version, use module load ncarg/version. For example, use module load ncarg/6.3.0 to load NCARG version 6.3.0 on Oakley. For the default version of ncarg, use

module load cairo
module load ncarg

Batch Usage on Oakley

Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Oakley and Scheduling Policies and Limits for more info.

Interactive Batch Session

For an interactive batch session on Oakley, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named interp1d_1.ncl.

Below is the example batch script job.txt for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=12
#PBS -N jobname
#PBS -j oe
module load cairo
module load ncarg
cp interp1d_1.ncl $TMPDIR
cd $TMPDIR
ncl interp1d_1.ncl
cp interp1d.ps $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Usage on Owens

Set-up on Owens

To load the default version use module load ncarg. To select a particular version, use module load ncarg/version. For example, use module load ncarg/6.3.0 to load NCARG version 6.3.0 on Owens. For the default version of ncarg, use

module load ncarg

Batch Usage on Owens

Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Owens and Scheduling Policies and Limits for more info.

Interactive Batch Session

For an interactive batch session on Owens, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. The following example batch script file will use the input file named interp1d_1.ncl.

Below is the example batch script job.txt for a serial run:

#PBS -l walltime=1:00:00
#PBS -l nodes=1:ppn=28
#PBS -N jobname
#PBS -j oe
module load ncarg
cp interp1d_1.ncl $TMPDIR
cd $TMPDIR
ncl interp1d_1.ncl
cp interp1d.ps $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

NWChem

NWChem aims to provide its users with computational chemistry tools that are scalable both in their ability to treat large scientific computational chemistry problems efficiently, and in their use of available parallel computing resources from high-performance parallel supercomputers to conventional workstation clusters.

Availability and Restrictions

Versions

The following versions of NWChem are available on OSC clusters:

Version	Oakley	Owens
6.1	X*
6.6		X*

* Current default version

You can use module spider nwchem to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

NWChem is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of NWChem, run the following command: module load nwchem. The default version will be loaded. To select a particular NWChem version, use module load nwchem/version. For example, use module load nwchem/6.1 to load NWChem 6.1.

Usage on Owens

Set-up

To configure your environment for use of NWChem, run the following command: module load nwchem. The default version will be loaded. To select a particular NWChem version, use module load nwchem/version. For example, use module load nwchem/6.6 to load NWChem 6.6.

Ncview

Ncview is a visual browser for netCDF format files. Typically you would use ncview to get a quick and easy, push-button look at your netCDF files. You can view simple movies of the data, view along various dimensions, take a look at the actual data values, change color maps, invert the data, etc.

Availability and Restrictions

Versions

The following versions of Ncview are available on OSC clusters:

VERSION	Owens
2.1.7	X*

* Current default version

You can use module spider ncview to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Ncview is available to all OSC users without restriction.

Usage

Usage on Owens

Set-up

To configure your environment for use of Ncview, run the following command: module load ncview. The default version will be loaded. To select a particular Ncview version, use module load ncview/version. For example, use module load ncview/2.1.7 to load Ncview 2.1.7.

NetCDF

NetCDF (Network Common Data Form) is an interface for array-oriented data access and a library that provides an implementation of the interface. The netcdf library also defines a machine-independent format for representing scientific data. Together, the interface, library, and format support the creation, access, and sharing of scientific data.

Availability and Restrictions

Versions

NetCDF is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	owens	notes
4.1.3	X			Default version on Oakley prior to 09/15/2015
4.3.2		X
4.3.3.1	X*	X*	X*

* Current default version

You can use module spider netcdf to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Additionally, the C++ interface version 4.3.0 and the Fortran interface version 4.4.2 is included in the netcdf/4.3.3.1 module.

Access

NetCDF is available without restriction to all OSC users.

Usage

Usage on Oakley

Set-up

Initalizing the system for use of the NetCDF is dependent on the system you are using and the compiler you are using. To load the default NetCDF, run the following command: module load netcdf. To use the parallel implementation of NetCDF, run the following command instead: module load pnetcdf. To load a particular version, use module load netcdf/version. For example, use module load netcdf/4.3.3.1 to load NetCDF version 4.3.3.1. You can use module spider netcdf to view available modules.

Building With NetCDF

With the netcdf library loaded, the following environment variables will be available for use:

Variable	Use
$NETCDF_CFLAGS	Use during your compilation step for C or C++ programs.
$NETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$NETCDF_LIBS	Use when linking your program to NetCDF.

Similarly, when the pnetcdf module is loaded, the following environment variables will be available:

VARIABLE	USE
$PNETCDF_CFLAGS	Use during your compilation step for C programs.
$PNETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$PNETCDF_LIBS	Use when linking your program to NetCDF.

For example, to build the code myprog.c with the netcdf library you would use:

icc -c $NETCDF_CFLAGS myprog.c
icc -o myprog myprog.o $NETCDF_LIBS

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. You must load the netcdf or pnetcdf module in your batch script before executing a program which is built with the netcdf library. Below is the example batch script that executes a program built with NetCDF:

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=12
module load netcdf
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
appname < foo.dat > foo.out
cp foo.out $PBS_O_WORKDIR

Usage on Ruby

Set-up

Initalizing the system for use of the NetCDF is dependent on the system you are using and the compiler you are using. To load the default NetCDF, run the following command: module load netcdf. To use the parallel implementation of NetCDF, run the following command instead: module load pnetcdf. To load a particular version, use module load netcdf/version. For example, use module load netcdf/4.3.3.1 to load NetCDF version 4.3.3.1. You can use module spider netcdf to view available modules.

Building With NetCDF

With the netcdf library loaded, the following environment variables will be available for use:

Variable	Use
$NETCDF_CFLAGS	Use during your compilation step for C or C++ programs.
$NETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$NETCDF_LIBS	Use when linking your program to NetCDF.

Similarly, when the pnetcdf module is loaded, the following environment variables will be available:

VARIABLE	USE
$PNETCDF_CFLAGS	Use during your compilation step for C programs.
$PNETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$PNETCDF_LIBS	Use when linking your program to NetCDF.

For example, to build the code myprog.c with the netcdf library you would use:

icc -c $NETCDF_CFLAGS myprog.c
icc -o myprog myprog.o $NETCDF_LIBS

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. You must load the netcdf or pnetcdf module in your batch script before executing a program which is built with the netcdf library. Below is the example batch script that executes a program built with NetCDF:

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=20
module load netcdf
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
appname < foo.dat > foo.out
cp foo.out $PBS_O_WORKDIR

Usage on Owens

Set-up

Initalizing the system for use of the NetCDF is dependent on the system you are using and the compiler you are using. To load the default NetCDF, run the following command: module load netcdf. To use the parallel implementation of NetCDF, run the following command instead: module load pnetcdf. To load a particular version, use module load netcdf/version. For example, use module load netcdf/4.3.3.1 to load NetCDF version 4.3.3.1. You can use module spider netcdf to view available modules.

Building With NetCDF

With the netcdf library loaded, the following environment variables will be available for use:

Variable	Use
$NETCDF_CFLAGS	Use during your compilation step for C or C++ programs.
$NETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$NETCDF_LIBS	Use when linking your program to NetCDF.

Similarly, when the pnetcdf module is loaded, the following environment variables will be available:

VARIABLE	USE
$PNETCDF_CFLAGS	Use during your compilation step for C programs.
$PNETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$PNETCDF_LIBS	Use when linking your program to NetCDF.

For example, to build the code myprog.c with the netcdf library you would use:

icc -c $NETCDF_CFLAGS myprog.c
icc -o myprog myprog.o $NETCDF_LIBS

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. You must load the netcdf or pnetcdf module in your batch script before executing a program which is built with the netcdf library. Below is the example batch script that executes a program built with NetCDF:

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=28
module load netcdf
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
appname < foo.dat > foo.out
cp foo.out $PBS_O_WORKDIR

Octave

Octave is a high-level language, primarily intended for numerical computations. It provides a convenient command-line interface for solving linear and nonlinear problems numerically, and for performing other numerical experiments using a language that is mostly compatible with Matlab. It may also be used as a batch-oriented language.

Octave has extensive tools for solving common numerical linear algebra problems, finding the roots of nonlinear equations, integrating ordinary functions, manipulating polynomials, and integrating ordinary differential and differential-algebraic equations. It is easily extensible and customizable via user-defined functions written in Octave's own language, or using dynamically loaded modules written in C++, C, Fortran, or other languages.

Availability and Restrictions

Versions

The following versions of Octave are available on OSC clusters:

Version	Oakley	Ruby	Owens
3.6.1	X
3.6.4	X*
4.0.0		X*
4.0.3			X*

* Current default version

You can use module spider octave to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Octave is available to all OSC users without restriction.

Usage

Set-up

To initialize either cluster for the use of Octave, run the following command:

module load octave

Using Octave

To run Octave, simply run the following command:

octave

Batch Usage

The following example batch script will an octave code file mycode.o via the batch processing system. The script requests one full node of cores on Oakley and 1 hour of walltime.

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=12
#PBS -l walltime=01:00:00
#PBS -l software=appname

module load octave
cd $PBS_O_WORKDIR
cp mycode.o $TMPDIR
cd $TMPDIR

octave < mycode.o > data.out

cp data.out $PBS_O_WORKDIR

Working with Packages

See the Octave 4.0.1 documentation on working with packages.

To install a package, launch an Octave session and type the pkg list command to see if there are any packages within your user scope. There is an issue where global packages may not be seen by particular Octave versions. To see the location of the global packages file use the command pkg global_list.

If you are having trouble installing your own packages, you can use the system-wide packages. Due to issues with system-wide installation, you will need to copy the system-wide package installation file to your local package installation file with cp $OCTAVE_PACKAGES $HOME/.octave_packages.

Then via pkg list you should see packages that you can load. This is clearly not portable and needs to be reperformed within a job script if you are using packages across multiple clusters.

OpenACC

OpenACC is a standard for parallel programming on accelerators, such as Nvidia GPUs and Intel Phi. It consists primarily of a set of compiler directives for executing code on the accelerator, in C and Fortran. OpenACC is currently only supported by the PGI compilers installed on OSC systems.

Availability and Restrictions

OpenACC is available without restriction to all OSC users. It is supported by the PGI compilers.

Usage

Set-up

OpenACC support is built into the compilers. There is no separate module to load.

Building With OpenACC

To build a program with OpenACC, use the compiler flag appropriate to your compiler. The correct libraries are included implicitly.

Compiler Family	Flag
PGI	`-acc -ta=nvidia -Minfo=accel`

Batch Usage

An OpenACC program will not run without an acelerator present. You need to ensure that your PBS resource request includes GPUs. For example, to run an OpenACC program on Oakley, your resource request should look something like this: #PBS -l nodes=1:ppn=12:gpus=2.

OpenFOAM

OpenFOAM is a suite of computational fluid dynamics applications. It contains myriad solvers, both compressible and incompressible, as well as many utilities and libraries.

Availability and Restrictions

Versions

The following versions of OpenFOAM are available on OSC clusters:

version	oakley	Ruby	Owens
2.3.0	X	X
2.4.0	X	X
3.0.0	X	X
4.1 5.0			X X

The location of OpenFOAM may be dependent on the compiler/MPI software stack, in that case, you should use both of the following commands (adjusting the version number) to learn how to load the appropriate modules:

module spider openfoam
module spider openfoam/3.0.0

Feel free to contact OSC Help if you need other versions for your work.

Access

OpenFOAM is available to all OSC users without restriction.

Basic Structure for an OpenFOAM Case

The basic directory structure for an OpenFOAM case is:

/home/yourusername/OpenFOAM_case
|-- 0
|-- U
|-- epsilon
|-- k
|-- p
`-- nut 
|-- constant
|-- RASProperties
|-- polyMesh
|   |-- blockMeshDict
|   `-- boundary
|-- transportProperties
`-- turbulenceProperties
|-- system
|-- controlDict
|-- fvSchemes
|-- fvSolution
`-- snappyHexMeshDict

IMPORTANT: To run in parallel, you need to also create the decomposeParDict file in the system directory. If you do not create this file, the decomposePar command will fail.

Usage

Usage on Oakley

Setup on Oakley

To configure the Oakley cluster for the use of OpenFOAM 1.7.x, use the following commands:

. /usr/local/OpenFOAM/OpenFOAM-1.7.x/etc/bashrc

To configure the Oakley cluster for the use of OpenFOAM 2.1.0, use the following commands:

. /usr/local/OpenFOAM/OpenFOAM-2.1.0/etc/bashrc

For OpenFOAM 2.3.0, use the following commands:

module load openfoam/2.3.0 # permissible with default Intel and MVAPICH2 compilers

For OpenFOAM 3.0.0, use the following commands:

module load openmpi/1.10 # currently only 3.0.0 is installed using OpenMPI libraries
module load openfoam/3.0.0

Batch Usage on Oakley

Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems.

On Oakley, refer to Queues and Reservations for Oakley and Scheduling Policies and Limits for more info.

Interactive Batch Session

For an interactive batch session on Oakley, one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=1:00:00

which gives you 12 cores (-l nodes=1:ppn=12) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor in a working directory on the system of your choice. Below is the example batch script (job.txt) for a serial run, note that this would use the default installation of OpenFOAM 2.3.0 on Oakley:

#PBS -N serial_OpenFOAM 
#PBS -l nodes=1:ppn=1 
#PBS -l walltime=24:00:00 
#PBS -j oe 
#PBS -S /bin/bash 
# Initialize OpenFOAM on Oakley Cluster
# This only works if you are using default modules
module load openfoam
# Move to the case directory, where the 0, constant and system directories reside
cd $PBS_O_WORKDIR
# Copy files to $TMPDIR and move there to execute the program
cp * $TMPDIR
cd $TMPDIR
# Mesh the geometry
blockMesh
# Run the solver
icoFoam
# Finally, copy files back to your home directory
cp * $PBS_O_WORKDIR

To run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -N parallel_OpenFOAM
#PBS -l nodes=2:ppn=8
#PBS -l walltime=6:00:00
#PBS -j oe
#PBS -S /bin/bash 
# Initialize OpenFOAM on Oakley Cluster
# This only works if you are using default modules
module load openfoam
#Move to the case directory, where the 0, constant and system directories reside
cd $PBS_O_WORKDIR
#Mesh the geometry
blockMesh
#Decompose the mesh for parallel run
decomposePar
#Run the solver
mpiexec simpleFoam -parallel 
#Reconstruct the parallel results
reconstructPar

Usage on Ruby

Setup on Ruby

To configure the Ruby cluster for the use of OpenFOAM 2.3.0, use the following commands:

module load openfoam/2.3.0 # permissible with default Intel and MVAPICH2 compilers

For OpenFOAM 3.0.0, use the following commands:

module load openmpi/1.10 # currently only 3.0.0 is installed using OpenMPI libraries
module load openfoam/3.0.0

Batch Usage on Ruby

Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems.

On Ruby, refer to Queues and Reservations for Ruby and Scheduling Policies and Limits for more info.

Interactive Batch Session

For an interactive batch session on Ruby, one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=1:00:00

which gives you 20 cores (-l nodes=1:ppn=20) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is the example batch script (job.txt) for a serial run:

#PBS -N serial_OpenFOAM 
#PBS -l nodes=1:ppn=1 
#PBS -l walltime=24:00:00 
#PBS -j oe 
#PBS -S /bin/bash 
# Initialize OpenFOAM on Ruby Cluster
# This only works if you are using default modules
module load openfoam
# Move to the case directory, where the 0, constant and system directories reside
cd $PBS_O_WORKDIR
# Copy files to $TMPDIR and move there to execute the program
cp * $TMPDIR
cd $TMPDIR
# Mesh the geometry
blockMesh
# Run the solver
icoFoam
# Finally, copy files back to your home directory
cp * $PBS_O_WORKDIR

To run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -N parallel_OpenFOAM
#PBS -l nodes=2:ppn=8
#PBS -l walltime=6:00:00
#PBS -j oe
#PBS -S /bin/bash 
# Initialize OpenFOAM on Ruby Cluster
# This only works if you are using default modules
module load openfoam
# Move to the case directory, where the 0, constant and system directories reside
cd $PBS_O_WORKDIR
# Mesh the geometry
blockMesh
# Decompose the mesh for parallel run
decomposePar
# Run the solver
mpiexec simpleFoam -parallel 
# Reconstruct the parallel results
reconstructPar

Usage on Owens

Setup on Owens

To configure the Owens cluster for the use of OpenFOAM 4.1, use the following commands:

module load openmpi/1.10-hpcx # currently only 4.1 is installed using OpenMPI libraries
module load openfoam/4.1

Batch Usage on Owens

Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems.

On Owens, refer to Queues and Reservations for Owens and Scheduling Policies and Limits for more info.

Interactive Batch Session

For an interactive batch session on Owens, one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=1:00:00

which gives you 28 cores (-l nodes=1:ppn=28) with 1 hour (-l walltime=1:00:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. Below is the example batch script (job.txt) for a serial run:

#PBS -N serial_OpenFOAM 
#PBS -l nodes=1:ppn=1
#PBS -l walltime=24:00:00 
#PBS -j oe 
#PBS -S /bin/bash 
# Initialize OpenFOAM on Owens Cluster
module load openmpi/1.10-hpcx
module load openfoam
# Move to the case directory, where the 0, constant and system directories reside
cd $PBS_O_WORKDIR
# Copy files to $TMPDIR and move there to execute the program
cp * $TMPDIR
cd $TMPDIR
# Mesh the geometry
blockMesh
# Run the solver
icoFoam
# Finally, copy files back to your home directory
cp * $PBS_O_WORKDIR

To run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

Non-interactive Batch Job (Parallel Run)

Below is the example batch script (job.txt) for a parallel run:

#PBS -N parallel_OpenFOAM
#PBS -l nodes=2:ppn=28
#PBS -l walltime=6:00:00
#PBS -j oe
#PBS -S /bin/bash 
# Initialize OpenFOAM on Ruby Cluster
# This only works if you are using default modules
module load openmpi/1.10-hpcx 
module load openfoam/2.3.0 
# Move to the case directory, where the 0, constant and system directories reside
cd $PBS_O_WORKDIR
# Mesh the geometry
blockMesh
# Decompose the mesh for parallel run
decomposePar
# Run the solver
mpiexec simpleFoam -parallel 
# Reconstruct the parallel results
reconstructPar

OpenMP

OpenMP is a standard for parallel programming on shared-memory systems, including multicore systems. It consists primarily of a set of compiler directives for sharing work among multiple threads. OpenMP is supported by all the Fortran, C, and C++ compilers installed on OSC systems.

Availability and Restrictions

OpenMP is available without restriction to all OSC users. It is supported by the Intel, PGI, and gnu compilers.

Usage

Set-up

OpenMP support is built into the compilers. There is no separate module to load.

Building With OpenMP

To build a program with OpenMP, use the compiler flag appropriate to your compiler. The correct libraries are included implicitly.

Compiler Family	Flag
Intel	`-qopenmp` or `-openmp`
gnu	`-fopenmp`
PGI	`-mp`

Batch Usage

An OpenMP program by default will use a number of threads equal to the number of processor cores available. To use a different number of threads, set the environment variable OMP_NUM_THREADS.

OpenMPI

MPI is a standard library for performing parallel processing using a distributed memory model. The Oakley, Ruby, and Owens clusters at OSC can use the OpenMPI implementation of the Message Passing Interface (MPI).

Availability and Restrictions

Versions

Installations are available for the Intel, PGI, and GNU compilers. The following versions of OpenMPI are available on OSC systems:

Version	Oakley	Ruby	Owens	COMPILERS SUPPORTED
1.10	X*	X	X	gnu/4.8.5 gnu/5.2.0 (oakley) intel/15.0.3 (oakley, ruby) intel/16.0.3 (oakley, owens) pgi/15.4.0 (oakley, ruby) pgi/16.5.0 (oakley, owens)
1.10-hpcx			X*	gnu/4.8.5 intel/16.0.3
1.10.7			X	gnu/4.8.5 intel/16.0.3 pgi/16.5.0
1.4.5	X			intel/15.0.3
2.0		X*	X	gnu/4.8.5 gnu/6.3.0 intel/16.0.3 pgi/16.5.0
2.0-hpcx		X	X	gnu/4.8.5 gnu/6.3.0
2.0.3			X	gnu/4.8.5 intel/16.0.3
2.1.2			X	gnu/4.8.5 intel/16.0.3

* Current default version

You can use module spider openmpi to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

OpenMPI is available without restriction to all OSC users.

Usage

Setup on OSC Clusters

To set up your environment for using the MPI libraries, you must load the appropriate module. On any OSC system, this is performed by:

module load openmpi

You will get the default version for the compiler you have loaded.

Building With MPI

To build a program that uses MPI, you should use the compiler wrappers provided on the system. They accept the same options as the underlying compiler. The commands are shown in the following table:

C	`mpicc`
C++	`mpicxx`
FORTRAN 77	`mpif77`
Fortran 90	`mpif90`

For example, to build the code my_prog.c using the -O2 option, you would use:

mpicc -o my_prog -O2 my_prog.c

In rare cases, you may be unable to use the wrappers. In that case, you should use the environment variables set by the module.

Variable	Use
`$MPI_CFLAGS`	Use during your compilation step for C programs.
`$MPI_CXXFLAGS`	Use during your compilation step for C++ programs.
`$MPI_FFLAGS`	Use during your compilation step for Fortran 77 programs.
`$MPI_F90FLAGS`	Use during your compilation step for Fortran 90 programs.

Batch Usage

Programs built with MPI can only run in the batch environment at OSC. For information on starting MPI programs using the command mpiexec see Job Scripts.

Be sure to load the same compiler and OpenMPI modules at execution time as at build time.

MPIEXEC

The mpiexec and mpirun commands are not part of the MPI standard and differ slightly between MPI implementations.

OpenMPI uses a different mpiexec implementation than other MPI libraries available at OSC. Basic functionality is the same but some options and defaults differ. To see all available options use mpiexec -h (with the openmpi module loaded) or see Open MPI Documentation.

Two differences in particular are worth noting.

By default the mpiexec process will spawn one MPI process per CPU core requested in a batch job. To specify the number of processes per node use the -npernode procs option. For one process per node use either -npernode 1 or -pernode.

If you run a hybrid OpenMPI / OpenMP program you should turn off binding with --bind-to none. Otherwise you may find your program using only half the available cores, leaving the others idle. This happens in particular with -npernode 1.

PAPI

PAPI provides the tool designer and application engineer with a consistent interface and methodology for use of the performance counter hardware found in most major microprocessors. PAPI enables software engineers to see, in near real time, the relation between software performance and processor events.

This software will be of interest only to HPC experts.

Availability and Restrictions

Versions

PAPI is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	Owens
5.4.1	X*	X*
5.6.0			X*

* Current default version

You can use module spider papi to view available modules for a given machine. For now, PAPI is available only with the Intel and gnu compilers. Feel free to contact OSC Help if you need other versions for your work.

Access

PAPI is available without restriction to all OSC users.

Usage

Usage on Oakley

Set-up

To load the default PAPI library, run the following command: module load papi. To load a particular version, use module load papi/version. For example, use module load papi/5.4.1 to load PAPI version 5.4.1. You can use module spider papi to view available modules.

Building With PAPI

The PAPI library provides the following variables for use at build time:

Variable	Use
`$PAPI_CFLAGS`	Use during your compilation step for C/C++ programs
`$PAPI_FFLAGS`	Use during your compilation step for FORTRAN programs
`$PAPI_LIB`	Use during your linking step programs

For example, to build the code myprog.c with the PAPI library you would use:

module load papi
gcc -c myprog.c $PAPI_CFLAGS
gcc -o myprog myprog.o $PAPI_LIB

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Usage on Ruby

Set-up

To load the default PAPI library, run the following command: module load papi. To load a particular version, use module load papi/version. For example, use module load papi/5.4.1 to load PAPI version 5.4.1. You can use module spider papi to view available modules.

Building With PAPI

The PAPI library provides the following variables for use at build time:

Variable	Use
`$PAPI_CFLAGS`	Use during your compilation step for C/C++ programs
`$PAPI_FFLAGS`	Use during your compilation step for FORTRAN programs
`$PAPI_LIB`	Use during your linking step programs

For example, to build the code myprog.c with the PAPI library you would use:

module load papi
gcc -c myprog.c $PAPI_CFLAGS
gcc -o myprog myprog.o $PAPI_LIB

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Usage on Owens

Set-up

Since PAPI version 5.2.0 is a System Install, no module is needed to run the application. To load a different version of the PAPI library, run the following command: module load papi. To load a particular version, use module load papi/version. For example, use module load papi/5.6.0 to load PAPI version 5.6.0. You can use module spider papi to view available modules.

Building With PAPI

To build the code myprog.c with the PAPI 5.2.0 library you would use:

gcc -c myprog.c -lpapi
gcc -o myprog myprog.o -lpapi

For other versions, the PAPI library provides the following variables for use at build time:

VARIABLE	USE
`$PAPI_CFLAGS`	Use during your compilation step for C/C++ programs
`$PAPI_FFLAGS`	Use during your compilation step for FORTRAN programs
`$PAPI_LIB`	Use during your linking step programs

For example, to build the code myprog.c with the PAPI version 5.6.0 library you would use:

module load papi
gcc -c myprog.c $PAPI_CFLAGS
gcc -o myprog myprog.o $PAPI_LIB

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

PAUP

PAUP is a leading program for performing phylogenetic analysis for bioinformatics sequences. PAUP currently runs as a single processor program. No further enhancements are suggested.

Availability and Restrictions

Versions

The following versions of PAUP are available on OSC systems:

Version	Oakley
4b10	X*

* Current default version

You can use module spider paup to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

PAUP is available to all OSC users without restriction.

Usage

Set-up

To configure your environment for using PAUP, run the following command:

module load paup

Using PAUP

After loading the PAUP module, PAUP can be run with commands similar to the following:

paup nexus_file > nexus_file.out

Batch Usage

PAUP is best run via the batch processing system. The following example batch script file will use the input file nexus_file and the output file nexus_file.out

#PBS -l walltime=10:00:00
#PBS -l nodes=1:ppn=1
#PBS -N paup
#PBS -j oe

cd $PBS_O_WORKDIR
cp ./nexus_file $TMPDIR
cd $TMPDIR

module load paup
paup nexus_file > nexus_file.out
cp nexus_file.out $PBS_O_WORKDIR

PGI Compilers

Fortran, C, and C++ compilers exist from the Portland Group.

Availability and Restrictions

PGI compilers are available on the Oakley, Ruby, and Owens Clusters. Here are the versions currently available at OSC:

Version	Oakley	ruby	Owens	notes
12.10	X			Default version on Oakley prior to 09/15/2015
13.5.0		X
14.9.0		X		Default version on Ruby prior to 09/15/2015
15.4.0	X	X
16.5.0	X*	X*	X*
17.3.0 17.10.0	X X	X X	X X

* : Current Default Version

You can use module spider pgi to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

The PGI Compilers are available to all OSC users without restriction on OSC clusters. If you would like to install the PGI compilers on your local computers, you may use the PGI Community Edition of the compiler for academic users for free at here.

Usage on Oakley

Set-up

To configure your environment for the default version of the PGI Compilers, use module load pgi . To configure your environment for a particular PGI compiler version, use module load pgi/version . For example, use module load pgi/12.10 to load the PGI compiler version 12.10.

Using the PGI Compilers

Once the module is loaded, compiling with the PGI compilers requires understanding which binary should be used for which type of code. Specifically, use the pgcc binary for C codes, the pgCC binary for C++ codes, the pgf77 for FORTRAN 77 codes and the pgf90 for Fortran 90 codes.

See our compilation guide for a more detailed breakdown of the compilers.

Building with the PGI Compilers

The PGI compilers recognize the following command line options (this list is not exhaustive, for more information run man <compiler binary name> ):

Compiler option	Purpose
-c	Compile into object code only; do not link
-DMACRO[=value]	Defines preprocessor macro MACRO with optional value (default value is 1)
-g	Enables debugging; disables optimization
-I/directory/name	Add /directory/name to the list of directories to be searched for #include files
-L/directory/name	Adds /directory/name to the list of directories to be searched for library files
-lname	Adds the library libname.a or libname.so to the list of libraries to be linked
-o outfile	Names the resulting executable outfile instead of a.out
-UMACRO	Removes definition of MACRO from preprocessor
-O0	Disable optimization; default if -g is specified
-O1	Light optimization; default if -g is not specified
-O or -O2	Heavy optimization
-O3	Aggressive optimization; may change numerical results
-Mipa	Inline function expansion for calls to procedures defined in separate files; implies -O2
-Munroll	Loop unrolling; implies -O2
-Mconcur	Automatic parallelization; implies -O2
-mp	Enables translation of OpenMP directives

Usage on Ruby

Set-up

To configure your environment for the default version of the PGI Compilers, use module load pgi . To configure your environment for a particular PGI compiler version, use module load pgi/version . For example, use module load pgi/14.9.0 to load the PGI compiler version 14.9.0.

Using the PGI Compilers

Once the module is loaded, compiling with the PGI compilers requires understanding which binary should be used for which type of code. Specifically, use the pgcc binary for C codes, the pgCC binary for C++ codes, the pgf77 for FORTRAN 77 codes and the pgf90 for Fortran 90 codes.

See our compilation guide for a more detailed breakdown of the compilers.

Building with the PGI Compilers

The PGI compilers recognize the following command line options (this list is not exhaustive, for more information run man <compiler binary name> ):

Compiler option	Purpose
-c	Compile into object code only; do not link
-DMACRO[=value]	Defines preprocessor macro MACRO with optional value (default value is 1)
-g	Enables debugging; disables optimization
-I/directory/name	Add /directory/name to the list of directories to be searched for #include files
-L/directory/name	Adds /directory/name to the list of directories to be searched for library files
-lname	Adds the library libname.a or libname.so to the list of libraries to be linked
-o outfile	Names the resulting executable outfile instead of a.out
-UMACRO	Removes definition of MACRO from preprocessor
-O0	Disable optimization; default if -g is specified
-O1	Light optimization; default if -g is not specified
-O or -O2	Heavy optimization
-O3	Aggressive optimization; may change numerical results
-Mipa	Inline function expansion for calls to procedures defined in separate files; implies -O2
-Munroll	Loop unrolling; implies -O2
-Mconcur	Automatic parallelization; implies -O2
-mp	Enables translation of OpenMP directives

Usage on Owens

Set-up

To configure your environment for the default version of the PGI Compilers, use module load pgi . To configure your environment for a particular PGI compiler version, use module load pgi/version . For example, use module load pgi/16.5.0 to load the PGI compiler version 16.5.0.

Using the PGI Compilers

Once the module is loaded, compiling with the PGI compilers requires understanding which binary should be used for which type of code. Specifically, use the pgcc binary for C codes, the pgc++ binary for C++ codes, the pgf77 for FORTRAN 77 codes and the pgf90 for Fortran 90 codes.

See our compilation guide for a more detailed breakdown of the compilers.

Building with the PGI Compilers

The PGI compilers recognize the following command line options (this list is not exhaustive, for more information run man <compiler binary name> ):

COMPILER OPTION	PURPOSE
-c	Compile into object code only; do not link
-DMACRO[=value]	Defines preprocessor macro MACRO with optional value (default value is 1)
-g	Enables debugging; disables optimization
-I/directory/name	Add /directory/name to the list of directories to be searched for #include files
-L/directory/name	Adds /directory/name to the list of directories to be searched for library files
-lname	Adds the library libname.a or libname.so to the list of libraries to be linked
-o outfile	Names the resulting executable outfile instead of a.out
-UMACRO	Removes definition of MACRO from preprocessor
-O0	Disable optimization; default if -g is specified
-O1	Light optimization; default if -g is not specified
-O or -O2	Heavy optimization
-O3	Aggressive optimization; may change numerical results
-Mipa	Inline function expansion for calls to procedures defined in separate files; implies -O2
-Munroll	Loop unrolling; implies -O2
-Mconcur	Automatic parallelization; implies -O2
-mp	Enables translation of OpenMP directives

PNG

PNG (Portable Network Graphics) is an extensible file format for the lossless, portable, well-compressed storage of raster images. PNG provides a patent-free replacement for GIF and can also replace many common uses of TIFF. Indexed-color, grayscale, and truecolor images are supported, plus an optional alpha channel. Sample depths range from 1 to 16 bits.

Availability and Restrictions

Versions

The following versions of PNG are available on OSC clusters:

VERSION	OAKLEY
1.2.52	X
1.5.9	X*

* Current default version

You can use module spider png to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

PNG is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

To configure your environment for use of PNG on Oakley, run the following command: module load png. The default version will be loaded. To select a particular PNG version, use module load png/version. For example, use module load png/1.2.52 to load PNG 1.2.52 on Oakley.

ParMETIS / METIS

ParMETIS (Parallel Graph Partitioning and Fill-reducing Matrix Ordering) is an MPI-based parallel library that implements a variety of algorithms for partitioning unstructured graphs, meshes, and for computing fill-reducing orderings of sparse matrices. ParMETIS extends the functionality provided by METIS and includes routines that are especially suited for parallel AMR computations and large scale numerical simulations. The algorithms implemented in ParMETIS are based on the parallel multilevel k-way graph-partitioning, adaptive repartitioning, and parallel multi-constrained partitioning schemes developed in Karypis lab.

METIS (Serial Graph Partitioning and Fill-reducing Matrix Ordering) is a set of serial programs for partitioning graphs, partitioning finite element meshes, and producing fill reducing orderings for sparse matrices. The algorithms implemented in METIS are based on the multilevel recursive-bisection, multilevel k-way, and multi-constraint partitioning schemes developed in Karypis lab.

Availability and Restrictions

Versions

ParMETIS is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	owens
4.0	X
4.0.3	X*	X*	X*

* Current default version

METIS is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	Ruby	Owens
5.0	X
5.1.0	X*	X*	X*

* Current default version

You can use module -r spider '.*metis.*' to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

ParMETIS / METIS is available without restriction to all OSC users.

Usage

Usage on Oakley

Set-up

To load ParMETIS, run the following command: module load parmetis. To use the serial implementation, METIS, run the following command instead: module load metis . You can use module spider metis and module spider parmetis to view available modules. Use module spider metis/versionand module spider parmetis/version to check what modules should be loaded before load ParMETIS / METIS.

Building With ParMETIS / METIS

With the ParMETIS library loaded, the following environment variables will be available for use:

Variable	Use
$PARMETIS_CFLAGS	Use during your compilation step for C or C++ programs.
$PARMETIS_LIBS	Use when linking your program to ParMETIS.

Similarly, when the METIS module is loaded, the following environment variables will be available:

VARIABLE	USE
$METIS_CFLAGS	Use during your compilation step for C programs.
$METIS_LIBS	Use when linking your program to METIS.

For example, to build the code myprog.cc with the METIS library you would use:

g++ -o myprog myprog.cc $METIS_LIBS

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. You must load the ParMETIS / METIS module in your batch script before executing a program which is built with the ParMETIS / METIS library. Below is the example batch script that executes a program built with ParMETIS:

#PBS -N myprogJob
#PBS -l nodes=1:ppn=12
module load gnu/4.8.4
module load parmetis
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
myprog < foo.dat > foo.out
cp foo.out $PBS_O_WORKDIR

Usage on Ruby

Set-up

To load ParMETIS, run the following command: module load parmetis. To use the serial implementation, METIS, run the following command instead: module load metis. You can use module spider metis and module spider parmetis to view available modules. Use module spider metis/version and module spider parmetis/version to check what modules should be loaded before load ParMETIS / METIS.

Building With ParMETIS / METIS

With the ParMETIS library loaded, the following environment variables will be available for use:

Variable	Use
$PARMETIS_CFLAGS	Use during your compilation step for C or C++ programs.
$PARMETIS_LIBS	Use when linking your program to ParMETIS.

Similarly, when the METIS module is loaded, the following environment variables will be available:

VARIABLE	USE
$METIS_CFLAGS	Use during your compilation step for C programs.
$METIS_LIBS	Use when linking your program to METIS.

For example, to build the code myprog.cc with the METIS library you would use:

g++ -o myprog myprog.cc $METIS_LIBS

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. You must load the ParMETIS / METIS module in your batch script before executing a program which is built with the ParMETIS / METIS library. Below is the example batch script that executes a program built with ParMETIS:

#PBS -N myprogJob
#PBS -l nodes=1:ppn=20
module load gnu/4.8.4
module load parmetis
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
myprog < foo.dat > foo.out
cp foo.out $PBS_O_WORKDIR

Usage on Owens

Set-up

To load ParMETIS, run the following command: module load parmetis. To use the serial implementation, METIS, run the following command instead: module load metis. You can use module spider metis and module spider parmetis to view available modules. Use module spider metis/version and module spider parmetis/version to check what modules should be loaded before load ParMETIS / METIS.

Building With ParMETIS / METIS

With the ParMETIS library loaded, the following environment variables will be available for use:

Variable	Use
$PARMETIS_CFLAGS	Use during your compilation step for C or C++ programs.
$PARMETIS_LIBS	Use when linking your program to ParMETIS.

Similarly, when the METIS module is loaded, the following environment variables will be available:

VARIABLE	USE
$METIS_CFLAGS	Use during your compilation step for C programs.
$METIS_LIBS	Use when linking your program to METIS.

For example, to build the code myprog.cc with the METIS library you would use:

g++ -o myprog myprog.cc $METIS_LIBS

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. You must load the ParMETIS / METIS module in your batch script before executing a program which is built with the ParMETIS / METIS library. Below is the example batch script that executes a program built with ParMETIS:

#PBS -N myprogJob
#PBS -l nodes=1:ppn=28
module load gnu/4.8.5
module load parmetis
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
myprog < foo.dat > foo.out
cp foo.out $PBS_O_WORKDIR

ParaView

ParaView is an open-source, multi-platform application designed to visualize data sets of size varying from small to very large. ParaView was developed to support distributed computational models for processing large data sets and to create an open, flexible user interface.

Availability and Restrictions

Versions

ParaView is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

VERSION	Oakley	Ruby	Owens
3.14.1	X
4.3.1		X
4.4.0	X*	X*	X*
5.3.0	X		X

* Current default version

You can use module spider paraview to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

ParaView is available for use by all OSC users.

Usage

Usage on Oakley

Set-up

To load the default version of ParaView module, use module load paraview. To select a particular software version, use module load paraview/version. For example, use module load paraview/4.4.0 to load ParaView version 4.4.0. Following a successful loading of the ParaView module, you can access the ParaView program:

paraview

Usage on Ruby

Set-up

To load the default version of ParaView module, use module load paraview. To select a particular software version, use module load paraview/version. For example, use module load paraview/4.3.1 to load ParaView version 4.3.1. Following a successful loading of the ParaView module, you can access the ParaView program:

paraview

Usage on Owens

Set-up

To load the default version of ParaView module, use module load paraview. To select a particular software version, use module load paraview/version. For example, use module load paraview/4.4.0 to load ParaView version 4.4.0. Following a successful loading of the ParaView module, you can access the ParaView program:

paraview

Using ParaView with OSC OnDemand

Using ParaView with OSC OnDemand requires VirtualGL. To begin, connect to OSC OnDemand and luanch a virtual desktop, either a Virtual Desktop Interface (VDI) or an Interactive HPC desktop. In the desktop open a terminal and load the ParaView and VirtualGL modules with module load paraview and module load virtualgl. You can then access the ParaView program with:

vglrun paraview

Note that using ParaView with OSC OnDemand does not work on all clusters.

Perl

Perl is a family of programming languages.

Availability and Restrictions

Versions

A system default Perl is available on Ruby, Oakley, and Owens. A Perl module is available on the Owens Clusters. The non-system Perl versions currently available at OSC are:

VERSION	oakley	Ruby	OWENS	Notes
5.10.1	X*#	X*#
5.16.3			X#
5.26.1			X*	**See note below.

* Current default version; # System version

** There is always some version of Perl in the environment. If you want a specific version you should load the approriate module. If you don't have a Perl module loaded you will get the system version.

You can use module spider perl to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Perl is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of Perl, use command module load perl. This will load the default version.

Installing Perl Modules

To install your own Perl modules locally, use CPAN Minus. Instructions for installing modules for system Perl are available here. Note that you do not need to load the cpanminus module if you are using a non-system Perl.

Picard

Picard is a set of command line tools for manipulating high-throughput sequencing (HTS) data and formats such as SAM/BAM/CRAM and VCF.

Availability and Restrictions

Versions

The following versions of Picard are available on OSC clusters:

Version	Oakley	Owens
2.3.0	X*	X*

* Current default version

You can use module spider picard to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Picard is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of Picard, run the following command: module load picard. The default version will be loaded. To select a particular Picard version, use module load picard/version. For example, use module load picard/2.3.0. to load Picard 2.3.0.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load picard, a new environment variable, PICARD, will be set. Thus, users can use the software by running the following command: java -jar $PICARD {other options}.

Usage on Owens

Set-up

To configure your environment for use of Picard, run the following command: module load picard. The default version will be loaded. To select a particular Picard version, use module load picard/version. For example, use module load picard/2.3.0. to load Picard 2.3.0.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load picard, a new environment variable, PICARD, will be set. Thus, users can use the software by running the following command: java -jar $PICARD {other options}.

Pixman

Pixman is a low-level software library for pixel manipulation, providing features such as image compositing and trapezoid rasterization. Important users of pixman are the cairo graphics library and the X server. Pixman is implemented as a library in the C programming language. It runs on many platforms, including Linux, BSD Derivatives, MacOS X, and Windows.

Availability and Restrictions

Versions

The following versions of Pixman are available on OSC clusters:

VERSION	OAKLEY
0.32.6	X*

*Current default version

You can use module spider pixman to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Pixman is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

To configure your environment for use of Pixman on Oakley, run the following command: module load pixman. The default version will be loaded. To select a particular Pixman version, use module load pixman/version. For example, use module load pixman/0.32.6 to load Pixman 0.32.6 on Oakley.

PnetCDF

PnetCDF is a library providing high-performance parallel I/O while still maintaining file-format compatibility with Unidata's NetCDF, specifically the formats of CDF-1 and CDF-2. Although NetCDF supports parallel I/O starting from version 4, the files must be in HDF5 format. PnetCDF is currently the only choice for carrying out parallel I/O on files that are in classic formats (CDF-1 and 2). In addition, PnetCDF supports the CDF-5 file format, an extension of CDF-2, that supports more data types and allows users to define large dimensions, attributes, and variables (>2B elements).

Availability and Restrictions

Versions

The following versions of PnetCDF are available at OSC:

Version	Ruby	Owens
1.5.0	X*
1.7.0		X*

* Current default version

You can use module spider pnetcdf to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

PnetCDF is available without restriction to all OSC users.

Usage

Usage on Ruby

Set-up

To initalize the system prior to using PnetCDF, run the following comand:

module load pnetcdf

Building With PnetDCF

With the PnetCDF module loaded, the following environment variables will be available for use:

Variable	Use
$PNETCDF_CFLAGS	Use during your compilation step for C or C++ programs.
$PNETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$PNETCDF_LIBS	Use when linking your program to PnetCDF.
$PNETCDF	Path to the PnetCDF installation directory

For example, to build the code myprog.c with the pnetcdf library you would use:

mpicc -c $PNETCDF_CFLAGS myprog.c
mpicc -o myprog myprog.o $PNETCDF_LIBS

Batch Usage

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=20
cd $PBS_O_WORKDIR
mpiexec ./myprog

Usage on Owens

Set-up

To initalize the system prior to using PnetCDF, run the following comand:

module load pnetcdf

Building With PnetDCF

With the PnetCDF module loaded, the following environment variables will be available for use:

VARIABLE	USE
$PNETCDF_CFLAGS	Use during your compilation step for C or C++ programs.
$PNETCDF_FFLAGS	Use during your compilation step for Fortran programs.
$PNETCDF_LIBS	Use when linking your program to PnetCDF.
$PNETCDF	Path to the PnetCDF installation directory

For example, to build the code myprog.c with the pnetcdf library you would use:

mpicc -c $PNETCDF_CFLAGS myprog.c
mpicc -o myprog myprog.o $PNETCDF_LIBS

Batch Usage

#PBS -N AppNameJob
#PBS -l nodes=1:ppn=28
cd $PBS_O_WORKDIR
mpiexec ./myprog

Python

Python is a high-level, multi-paradigm programming language that is both easy to learn and useful in a wide variety of applications. Python has a large standard library as well as a large number of third-party extensions, most of which are completely free and open source.

Availability and Restrictions

Versions

Python is available on Oakley, Ruby, and Owens Clusters. The versions currently available at OSC are:

Version	Oakley	ruby	owens	notes
2.7			X
2.7.1	X			Default version on Oakley prior to 09/15/2015
2.7.8	X(A)
2.7.9		X(A)		On Ruby module load 2.7.8 will actually load version 2.7.9
3.4.2	X(A)*	X(A)*
3.5			X*
3.6			X

* Current default version; A = installed as an integrated package Anaconda

You can use module spider python to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Python is available for use by all OSC users.

Usage

Usage on Oakley

Set-up

To load the default version of Python module, use module load python. To select a particular software version, use module load python/version. For example, use module load python/2.7.1 to load Python version 2.7.1. After the module is loaded, you can run the interpreter by using the command python. To unload the Python 2.7.1 module, use the command module unload python/2.7.1.

Installed Modules

We have installed a number of Python packages and tuned them for optimal performance on our systems. When Python module is loaded, executing module help python will help you view the current list of packages available.

NOTE:

numpy and scipy have been compiled to use MKL for best performance
Due to architecture differences between our supercomputers, we recommend NOT installing your own packages in ~/.local. Instead, you should install them in some other directory and set $PYTHONPATH in your default environment. For more information about installing your own Python modules, please see our HOWTO.

Batch Usage

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=00:20:00

which requests one whole node with 12 cores (-l nodes=1:ppn=12), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Non-interactive Batch Job (Serial Run)

Python can technically be used in both interactive and non-interactive batch jobs, but you will most likely only want to run a Python script non-interactively. To do this, you will need to write a batch submission script and a Python script using any text editor you like in a working directory on the system of your choice. A batch script can be created and submitted for a serial or parallel run. Below is a sample batch submission script( job.txt) and a simple “Hello World” program (hello.py) written in Python.

Example batch submission script job.txt:

#PBS –N hello
#PBS –l nodes=1:ppn=12
#PBS –l walltime=1:00:00
#PBS –j oe
module load python/2.7.1
python hello.py
#end of sample.job

Example Python script, hello.py:

import sys
hello = [“Hello”, “World”]
for i in hello:
print i
sys.exit(0)
#end of hello.py

Usage on Ruby

Set-up

To load the default version of Python module, use module load python. To select a particular software version, use module load python/version. For example, use module load python/3.4.2 to load Python version 3.4.2. After the module is loaded, you can run the interpreter by using the command python. To unload the Python 3.4.2 module, use the command module unload python/3.4.2.

Installed Modules

We have installed a number of Python packages and tuned them for optimal performance on our systems. When using the Anaconda distributions of python you can run conda list to view the installed packages.

NOTE:

Due to architecture differences between our supercomputers, we recommend NOT installing your own packages in ~/.local. Instead, you should install them in some other directory and set $PYTHONPATH in your default environment. For more information about installing your own Python modules, please see our HOWTO.

Batch Usage

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=20 -l walltime=00:20:00

which requests one whole node with 20 cores (-l nodes=1:ppn=20), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Usage on Owens

Set-up

To load the default version of Python module, use module load python . To select a particular software version, use module load python/version. For example, use module load python/3.5 to load the latest version of Python 3.5. After the module is loaded, you can run the interpreter by using the command python. To unload the latest version of Python 3.5 module, use the command module unload python/3.5.

Installed Modules

We have installed a number of Python packages and tuned them for optimal performance on our systems. When using the Anaconda distributions of python you can run conda list to view the installed packages.

NOTE:

Due to architecture differences between our supercomputers, we recommend NOT installing your own packages in ~/.local. Instead, you should install them in some other directory and set $PYTHONPATH in your default environment. For more information about installing your own Python modules, please see our HOWTO.

Batch Usage

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=00:20:00

which requests one whole node with 28 cores (-l nodes=1:ppn=28), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Q-Chem

Q-Chem is a general purpose ab initio electronic structure program. Its latest version emphasizes Self-Consistent Field, especially Density Functional Theory, post Hartree-Fock, and innovative algorithms for fast performance and reduced scaling calculations. Geometry optimizations, vibrational frequencies, thermodynamic properties, and solution modeling are available. It performs reasonably well within its single reference paradigm on open shell and excited state systems. The Q-Chem Home Page has additional information.

Availability and Restrictions

Versions

Q-Chem is available on the Oakley Cluster. The versions currently available at OSC are:

VERSION	OAKLEY	ruby	owens	notes
4.0.1	X			Default version on Oakley prior to 09/15/2015
4.2.1	X
4.2.2	X
4.3	X
4.3.1	X
4.4	X
4.4.1	X*	X*	X*
4.4.1-openmp	X	X	X

* Current default version

You can use module spider qchem to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

The use of Q-Chem requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Usage

Usage on Oakley

Set-up on Oakley

Q-Chem usage is controlled via modules. Load one of the Q-Chem modulefiles at the command line, in your shell initialization script, or in your batch scripts. To load the default version of Q-Chem module, use module load qchem. To select a particular software version, use module load qchem/version. For example, use module load qchem/4.3 to load Q-Chem version 4.3 on Oakley.

Examples

The root of the Q-Chem directory tree is /usr/local/qchem/
Example Q-Chem input files are in the samples subdirectory
Computational Chemistry Workshop materials are in /home/srb/workshops/compchem/qchem

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=1 -l walltime=00:20:00

which requests one core (-l nodes=1:ppn=1), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Usage on Ruby

Set-up on Ruby

Q-Chem usage is controlled via modules. Load one of the Q-Chem modulefiles at the command line, in your shell initialization script, or in your batch scripts. To load the default version of Q-Chem module, use module load qchem. To select a particular software version, use module load qchem/version. For example, use module load qchem/4.4.1 to load Q-Chem version 4.4.1 on Ruby.

Examples

The root of the Q-Chem directory tree is /usr/local/qchem/
Example Q-Chem input files are in the samples subdirectory
Computational Chemistry Workshop materials are in /home/srb/workshops/compchem/qchem

Batch Usage on Ruby

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=1 -l walltime=00:20:00

which requests one core (-l nodes=1:ppn=1), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Usage on Owens

Set-up on Owens

Q-Chem usage is controlled via modules. Load one of the Q-Chem modulefiles at the command line, in your shell initialization script, or in your batch scripts. To load the default version of Q-Chem module, use module load qchem. To select a particular software version, use module load qchem/version. For example, use module load qchem/4.4.1 to load Q-Chem version 4.4.1 on Owens.

Examples

The root of the Q-Chem directory tree is /usr/local/qchem/
Example Q-Chem input files are in the samples subdirectory
Computational Chemistry Workshop materials are in /home/srb/workshops/compchem/qchem

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=1 -l walltime=00:20:00

which requests one core (-l nodes=1:ppn=1), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Quantum Espresso

Quantum ESPRESSO (QE) is a program package for ab-initio molecular dynamics (MD) simulations and electronic structure calculations. It is based on density-functional theory, plane waves, and pseudopotentials.

Availability and Restrictions

Versions

The following versions are available on OSC systems:

Version	Oakley	ruby	owens
5.0.3	X
5.2.1	X*	X*	X*
6.1			X
6.2.1			X

* Current default version

You can use module spider espresso to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

NOTE: 5.2.1 version on Oakley and 6.1 on Owens include the WEST package. Version 6.1 on Owens is available for Intel compiler versions 16.0.3 and 17.0.2. The 17.0.2 installation should be preferred because it addresses potential issues with MKL in 16.0.3 (a sympton of these issues is an error message starting with: Fatal error in PMPI_).

Access

Quantum ESPRESSO is open source and available without restriction. We recommend that Owens be used.

Usage

Set-up

You can configure your environment for the usage of Quantum ESPRESSO by running the following command:

module load espresso

In the case of multiple compiled versions load the appropriate compiler first, e.g., on Owens to select the most recently compiled QE 6.1 version use the following commands:

module load intel/17.0.2
module load espresso/6.1

Batch Usage

Sample batch scripts and input files are available here:

~srb/workshops/compchem/espresso/

R

R is a language and environment for statistical computing and graphics. It is similar to the S language and environment developed at Bell Laboratories (formerly AT&T, now Lucent Technologies). R provides a wide variety of statistical and graphical techniques and is highly extensible.

Availability and Restrictions

Versions

The following versions of R are available on OSC systems:

Version	Oakley	Owens
3.0.1	X
3.1.3	X
3.2.0	X
3.3.1	X*	X
3.3.2		X*
3.4.0		X
3.4.2		X
3.5.0		X

* Current default version

You can use module spider R to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

R is available to all OSC users without restriction.

Usage

Set-up

In order to configure your environment for the usage of R, run the following command:

module load R

Using R

Once your environment is configured, R can be started simply by entering the following command:

For a listing of command line options, run:

R --help

Batch Usage

Running R interactively on a login node for extended computations is not recommended and may violate OSC usage policy. In order to run R in batch, reference the example batch script below. This script requests one full node on the Oakley cluster for 1 hour of walltime.

#PBS -N R_ExampleJob
#PBS -l nodes=1:ppn=12

module load R
cd $PBS_O_WORKDIR
cp in.dat $TMPDIR
cd $TMPDIR

R CMD BATCH test.R test.Rout

cp test.Rout $PBS_O_WORKDIR

NOTE: Owens users should use ppn=28 instead of ppn=12.

HOWTO: Install Local R Packages

Please see our HOWTO guide on installing Local R packages.

R Parallel Cluster Job Submission

In this demonstration, we will submit a parallel job to the cluster using R. Most parallelization concepts in R are centered around loop-level parallelism with independence, where each iteration acts as a separate simulation. There are a variety of strategies based on the concept of:

foreach - enumerating through the contents of a collection
apply - application of a function to a collection

This is a toy example, hence we use a function that will generate values sampled from a normal distribution and summing the vector of those results; every call to the function is a separate simulation. But sometimes simple is the best way to demonstrate using OSC resources, and to consider a variety of approaches, afforded by the packages that provide foreach, mclapply, and clusterApply. For sufficient parallelism, we generate 100M elements and run this function 100 times.

For reference, this example was heavily adapted from TACC. We then combine several of the strategies in one single R script.

Background: Indirectly, MPI bindings are provided with the Rmpi package, compiled with OpenMPI. On the OSC Owens cluster, this is automatically selected for you if you use the system packages provided with R/3.3.1+.

The code is presented below in 100 lines of parallel_testing.R which you can toggle below.

parallel_testing.R

myProc <- function(size=100000000) {
  # Load a large vector
  vec <- rnorm(size)
  # Now sum the vec values
  return(sum(vec))
}

detachDoParallel <- function() {
  detach("package:doParallel")
  detach("package:foreach")
  detach("package:parallel")
  detach("package:iterators")
}

max_loop <- 100

# version 1: use mclapply (multicore) - warning - generates zombie processes
library(parallel)

tick <- proc.time()
result <- mclapply(1:max_loop, function(i) myProc(), mc.cores=detectCores())
tock <- proc.time() - tick

cat("\nmclapply/multicore test times using", detectCores(), "cores: \n")
tock

# version 2: use foreach with explicit MPI cluster on one node
library(doParallel, quiet = TRUE)
library(Rmpi)

slaves <- detectCores() - 1
{ sink("/dev/null"); cl_onenode <- makeCluster(slaves, type="MPI"); sink(); } # number of MPI tasks to use
registerDoParallel(cl_onenode)

tick <- proc.time()
result <- foreach(i=1:max_loop, .combine=c) %dopar% {
    myProc()
}
tock <- proc.time() - tick

cat("\nforeach w/ Rmpi test times using", slaves, "MPI slaves: \n")
tock

invisible(stopCluster(cl_onenode))
detachDoParallel()

# version 3: use foreach (multicore)
library(doParallel, quiet = TRUE)

cores <- detectCores()
cl <- makeCluster(cores)
registerDoParallel(cl)

tick <- proc.time()
result <- foreach(i=1:max_loop, .combine=c) %dopar% {
    myProc()
}
tock <- proc.time() - tick

cat("\nforeach w/ fork times using", cores, "cores: \n")
tock

invisible(stopCluster(cl))
detachDoParallel()

## version 4: use foreach (doSNOW/Rmpi)
library(doParallel, quiet = TRUE)
library(Rmpi)

slaves <- as.numeric(Sys.getenv(c("PBS_NP")))-1
{ sink("/dev/null"); cl <- makeCluster(slaves, type="MPI"); sink(); } # number of MPI tasks to use
registerDoParallel(cl)

tick <- proc.time()
result <- foreach(i=1:max_loop, .combine=c) %dopar% {
    myProc()
}
tock <- proc.time() - tick

cat("\nforeach w/ Rmpi test times using", slaves, "MPI slaves: \n")
tock

detachDoParallel() # no need to stop cluster we will use it again

## version 5: use snow backed by Rmpi (cluster already created)
library(Rmpi) # for mpi.*
library(snow) # for clusterExport, clusterApply

#slaves <- as.numeric(Sys.getenv(c("PBS_NP")))-1
clusterExport(cl, list('myProc'))

tick <- proc.time()
result <- clusterApply(cl, 1:max_loop, function(i) myProc())
tock <- proc.time() - tick

cat("\nsnow w/ Rmpi test times using", slaves, "MPI slaves: \n")
tock

invisible(stopCluster(cl))
mpi.quit()

Now, we set up a job to submit R on the cluster, the script again you can toggle below:

owens_job.sh

#!/bin/bash
#PBS -l walltime=10:00
#PBS -l nodes=2:ppn=28
#PBS -j oe

cd $PBS_O_WORKDIR
ml openmpi/1.10
ml R/3.3.1

# parallel R: submit job with one MPI master
mpirun -np 1 R --slave < parallel_testing.R

Then you can submit to the cluster via of qsub owens_job.sh.

If everything is good, you should see some output like:

mclapply/multicore test times using 28 cores:
   user  system elapsed
 59.162   1.577  27.099

foreach w/ Rmpi test times using 27 MPI slaves:
   user  system elapsed
 12.617  15.204  27.808

foreach w/ fork times using 28 cores:
   user  system elapsed
  0.072   0.031  27.003

foreach w/ Rmpi test times using 55 MPI slaves:
   user  system elapsed
  8.242   7.230  16.048

snow w/ Rmpi test times using 55 MPI slaves:
   user  system elapsed
  7.494   8.052  15.581

This example shows that MPI has the potential to be effectively parallelized on 1 and 2 nodes. In another series, we will look at a more realistic example using more than a single data structure.

Takeaways

When using more than one node use Rmpi or Rmpi SNOW (like in version 4 or version 5). Also, keep in mind that in independent simulation scenarios the use of a do loop means that there are logical processor alignments with the value of max_loop which is set to 100. Because we are setting a fixed problem size (100 times 100M), the benefits of parallelism are chunked by the granularity of the iterations.

Typically, one sets the max_loop to the number of MPI slaves, or where MPI slaves is a factor of max_loop, because each MPI slave will have its own iteration to perform. However, this all depends on whether or not you have sufficient work on a single iteration. With max_loop set to 100, we are interested in strong scaling effects for a fixed problem size. For this example, the chart below, using 4 nodes of 28 cores (112 cores total available), makes it clear that you should set max_loop equal to MPI slaves.

Wall Time Performance of R Parallel Example with 100 Loops

parallel_testing_version5_only.R

myProc <- function(size=100000000) {
      # Load a large vector
      vec <- rnorm(size)
  # Now sum the vec values
  return(sum(vec))
}

max_loop <- 100

## version 5: use snow backed by Rmpi
library(Rmpi) # for mpi.*
library(snow) # for clusterExport, clusterApply

slaves <- as.numeric(Sys.getenv(c("PBS_NP")))-1
{ sink("/dev/null"); cl <- makeCluster(slaves, type="MPI"); sink(); } # MPI tasks to use
clusterExport(cl, list('myProc'))

tick <- proc.time()
result <- clusterApply(cl, 1:max_loop, function(i) myProc())
tock <- proc.time() - tick

cat("\nsnow w/ Rmpi test times using", slaves, "MPI slaves: \n")
tock

stopCluster(cl)
mpi.quit()

RAxML

"RAxML is a fast implementation of maximum-likelihood (ML) phylogeny estimation that operates on both nucleotide and protein sequence alignments."

Availability and Restrictions

Versions

The following versions of RAxML are available on OSC systems:

Version	Oakley
7.4.2	X*

* Current default version

You can use module spider raxml to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

RAxML is available to all OSC users without restriction.

Usage

Set-up

On the Oakley Cluster RAxML is accessed by executing the following commands:

module load biosoftw
module load RAxML

Using RAxML

raxmlHPC[-MPI|-PTHREADS] -s sequenceFileName -n outputFileName -m substitutionModel [-a weightFileName] [-b bootstrapRandomNumberSeed] [-c numberOfCategories][-d] [-e likelihoodEpsilon] [-E excludeFileName] [-f a|b|c|d|e|g|h|i|j|m|n|o|p|s|t|w|x] [-g groupingFileName] [-h] [-i initialRearrangementSetting] [-j] [-k] [-l sequenceSimilarityThreshold] [-L sequenceSimilarityThreshold] [-M] [-o outGroupName1[,outGroupName2[,...]]] [-p parsimonyRandomSeed] [-P proteinModel] [-q multipleModelFileName] [-r binaryConstraintTree] [-t userStartingTree] [-T numberOfThreads] [-u multiBootstrapSearches] [-v][-w workingDirectory] [-x rapidBootstrapRandomNumberSeed][-y][-z multipleTreesFile] [-#|-N numberOfRuns]

Options

-a    Specify a column weight file name to assign individual weights to each column of the alignment. Those weights must be integers separated by any type and number of whitespaces within a separate file, see file "example_weights" for an example.
-b    Specify an integer number (random seed) and turn on bootstrapping.  DEFAULT: OFF
-c    Specify number of distinct rate catgories for RAxML when modelOfEvolution is set to GTRCAT or GTRMIX.  Individual per-site rates are categorized into numberOfCategories rate categories to accelerate computations.  DEFAULT: 25
-d    start ML optimization from random starting tree.  DEFAULT: OFF
-e    set model optimization precision in log likelihood units for final optimization of tree topology under MIX/MIXI or GAMMA/GAMMAI.  DEFAULT: 0.1   for models not using proportion of invariant sites estimate 0.001 for models using proportion of invariant sites estimate
-E    specify an exclude file name, that contains a specification of alignment positions you wish to exclude.  Format is similar to Nexus, the file shall contain entries like "100-200 300-400", to exclude a single column write, e.g., "100-100", if you use a mixed model, an appropriately adapted model file will be written.
-f    select algorithm:
      "-f a": rapid Bootstrap analysis and search for best-scoring ML tree in one program run
      "-f b": draw bipartition information on a tree provided with "-t" based on multiple trees (e.g. form a bootstrap) in a file specifed by "-z"
      "-f c": check if the alignment can be properly read by RAxML
      "-f d": new rapid hill-climbing
      "-f e": optimize model+branch lengths for given input tree under GAMMA/GAMMAI only
      "-f g": compute per site log Likelihoods for one ore more trees passed via "-z" and write them to a file that can be read by CONSEL
      "-f h": compute log likelihood test (SH-test) between best tree passed via "-t" and a bunch of other trees passed via "-z"
      "-f i": perform a really thorough bootstrap, refinement of final BS tree under GAMMA and a more exhaustive algorithm
      "-f j": generate a bunch of bootstrapped alignment files from an original alignment file
      "-f m": Compare bipartitions between two bunches of trees passed via "-t" and "-z" respectively. This will return the Pearson correlation between all bipartitions found in the two tree files. A file called RAxML_bipartitionFrequencies.outpuFileName will be printed that contains the pair-wise bipartition frequencies of the two sets
      "-f n": Compute the log likelihood score of all trees contained in a tree file provided by "-z" under GAMMA or GAMMA+P-Invar
      "-f o": old and slower rapid hill-climbing
      "-f p": perform pure stepwise MP addition of new sequences to an incomplete starting tree
      "-f s": split up a multi-gene partitioned alignment into the respective subalignments
      "-f t": do randomized tree searches on one fixed starting tree
      "-f w": compute ELW test on a bunch of trees passed via "-z"
      "-f x": compute pair-wise ML distances, ML model parameters will be estimated on an MP starting tree or a user-defined tree passed via "-t", only allowed for GAMMA-based models of rate heterogeneity.  DEFAULT: new rapid hill climbing
-g    specify the file name of a multifurcating constraint tree this tree does not need to be comprehensive, i.e. must not contain all taxa
-h    Display this help message.
-i    Initial rearrangement setting for the subsequent application of topological changes phase.  DEFAULT: determined by program
-j    Specifies if checkpoints will be written by the program. If checkpoints (intermediate tree topologies) shall be written by the program specify "-j"  DEFAULT: OFF
-k    Specifies that bootstrapped trees should be printed with branch lengths.  The bootstraps will run a bit longer, because model parameters will be optimized at the end of each run. Use with CATMIX/PROTMIX or GAMMA/GAMMAI.  DEFAULT: OFF
-l    Specify a threshold for sequence similarity clustering. RAxML will then print out an alignment to a file called sequenceFileName.reducedBy.threshold that only contains sequences <= the specified thresold that must be between  0.0 and 1.0.  RAxML uses the QT-clustering algorithm to perform this task. In addition, a file called RAxML_reducedList.outputFileName will be written that contains clustering information.  DEFAULT: OFF
-L    Same functionality as "-l" above, but uses a less exhasutive and thus faster clustering algorithm.  This is intended for very large datasets with more than 20,000-30,000 sequences.  DEFAULT: OFF
-m    Model of Nucleotide or Amino Acid Substitution:
NUCLEOTIDES:
    "-m GTRCAT"                      : GTR + Optimization of substitution rates + Optimization of site-specific evolutionary rates which are categorized into numberOfCategories distinct rate categories for greater computational efficiency if you do a multiple analysis with "-#" or "-N" but without bootstrapping the program will use GTRMIX instead
    "-m GTRGAMMA"                    : GTR + Optimization of substitution rates + GAMMA model of rate heterogeneity (alpha parameter will be estimated)
    "-m GTRMIX"                      : Inference of the tree under GTRCAT and thereafter evaluation of the final tree topology under GTRGAMMA
    "-m GTRCAT_GAMMA"                : Inference of the tree with site-specific evolutionary rates.  However, here rates are categorized using the 4 discrete GAMMA rates. Evaluation of the final tree topology under GTRGAMMA
    "-m GTRGAMMAI"                   : Same as GTRGAMMA, but with estimate of proportion of invariable sites
    "-m GTRMIXI"                     : Same as GTRMIX, but with estimate of proportion of invariable sites
    "-m GTRCAT_GAMMAI"               : Same as GTRCAT_GAMMA, but with estimate of proportion of invariable sites
AMINO ACIDS:
    "-m PROTCATmatrixName[F]"        : specified AA matrix + Optimization of substitution rates + Optimization of site-specific evolutionary rates which are categorized into numberOfCategories distinct rate categories for greater computational efficiency if you do a multiple analysis with  "-#" or "-N" but without bootstrapping the program will use PROTMIX... instead
    "-m PROTGAMMAmatrixName[F]"      : specified AA matrix + Optimization of substitution rates + GAMMA model of rate heterogeneity (alpha parameter will be estimated)
    "-m PROTMIXmatrixName[F]"        : Inference of the tree under specified AA matrix + CAT and thereafter evaluation of the final tree topology under specified AA matrix + GAMMA
    "-m PROTCAT_GAMMAmatrixName[F]"  : Inference of the tree under specified AA matrix and site-specific evolutionary rates.  However, here rates are categorized using the 4 discrete GAMMA rates. Evaluation of the final tree topology under specified AA matrix + GAMMA
    "-m PROTGAMMAImatrixName[F]"     : Same as PROTGAMMAmatrixName[F], but with estimate of proportion of invariable sites
    "-m PROTMIXImatrixName[F]"       : Same as PROTMIXmatrixName[F], but with estimate of proportion of invariable sites
    "-m PROTCAT_GAMMAImatrixName[F]" : Same as PROTCAT_GAMMAmatrixName[F], but with estimate of proportion of invariable sites
Available AA substitution models: DAYHOFF, DCMUT, JTT, MTREV, WAG, RTREV, CPREV, VT, BLOSUM62, MTMAM, GTR
With the optional "F" appendix you can specify if you want to use empirical base frequencies.  Please not that for mixed models you can in addition specify the per-gene AA model in the mixed model file (see manual for details)
-M    Switch on estimation of individual per-partition branch lengths. Only has effect when used in combination with "-q". Branch lengths for individual partitions will be printed to separate files.  A weighted average of the branch lengths is computed by using the respective partition lengths.  DEFAULT: OFF
-n    Specifies the name of the output file.
-o    Specify the name of a single outgrpoup or a comma-separated list of outgroups, eg "-o Rat" or "-o Rat,Mouse", in case that multiple outgroups are not monophyletic the first name in the list will be selected as outgroup, don't leave spaces between taxon names!
-q    Specify the file name which contains the assignment of models to alignment partitions for multiple models of substitution. For the syntax of this file please consult the manual.
-p    Specify a random number seed for the parsimony inferences. This allows you to reproduce your results and will help me debug the program. This option HAS NO EFFECT in the parallel MPI version
-P    Specify the file name of a user-defined AA (Protein) substitution model. This file must contain 420 entries, the first 400 being the AA substitution rates (this must be a symmetric matrix) and the last 20 are the empirical base frequencies
-r    Specify the file name of a binary constraint tree.  This tree does not need to be comprehensive, i.e. must not contain all taxa
-s    Specify the name of the alignment data file in PHYLIP format
-t    Specify a user starting tree file name in Newick format
-T    PTHREADS VERSION ONLY! Specify the number of threads you want to run.  Make sure to set "-T" to at most the number of CPUs you have on your machine, otherwise, there will be a huge performance decrease!
-u    Specify the number of multiple BS searches per replicate to obtain better ML trees for each replicate.  DEFAULT: One ML search per BS replicate
-v    Display version information
-w    Name of the working directory where RAxML will write its output files.  DEFAULT: current directory
-x    Specify an integer number (random seed) and turn on rapid bootstrapping
-y    If you want to only compute a parsimony starting tree with RAxML specify "-y", the program will exit after computation of the starting tree.  DEFAULT: OFF
-z    Specify the file name of a file containing multiple trees e.g. from a bootstrap that shall be used to draw bipartition values onto a tree provided with "-t", It can also be used to compute per site log likelihoods in combination with "-f g" and to read a bunch of trees for a couple of other options ("-f h", "-f m", "-f n").
-#|-N Specify the number of alternative runs on distinct starting trees.  In combination with the "-b" option, this will invoke a multiple bootstrap analysis.  Note that "-N" has been added as an alternative since "-#" sometimes caused problems with certain MPI job submission systems, since "-#" is often used to start comments.  DEFAULT: 1 single analysis

Example

The file brown.philip is from the PAML example and modified to be in PHYLIP format.

#PBS -N RAxML_test
#PBS -l walltime=10:00
#PBS -l nodes=1:ppn=4

cd $PBS_O_WORKDIR
module load biosoftw
module load RAxML
raxmlHPC-PTHREAD -T 4 -m GTRCAT -n test -s brown.phylip

If you want to compute more than one tree add the following line before calling the program:

module load pvfs2

And substitute raxmlHPC-MPI for raxmlHPC-PTHREAD.

RNA-SeQC

RNA-SeQC is a java program which computes a series of quality control metrics for RNA-seq data. The input can be one or more BAM files. The output consists of HTML reports and tab delimited files of metrics data. This program can be valuable for comparing sequencing quality across different samples or experiments to evaluate different experimental parameters. It can also be run on individual samples as a means of quality control before continuing with downstream analysis.

Availability and Restrictions

Versions

The following versions of RNA-SeQC are available on OSC clusters:

Version	Oakley	Owens
1.1.8	X*	X*

* Current default version

You can use module spider rna-seqc to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

RNA-SeQC is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of RNA-SeQC, run the following command: module load rna-seqc. The default version will be loaded. To select a particular RNA-SeQC version, use module load rna-seqc/version. For example, use module load rna-seqc/1.1.8 to load RNA-SeQC 1.1.8.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load rna-seqc, a new environment variable, RNA_SEQC, will be set. Thus, users can use the software by running the following command: java -jar $RNA_SEQC {other options}.

Usage on Owens

Set-up

To configure your environment for use of RNA-SeQC, run the following command: module load rna-seqc. The default version will be loaded. To select a particular RNA-SeQC version, use module load rna-seqc/version. For example, use module load rna-seqc/1.1.8 to load RNA-SeQC 1.1.8.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load rna-seqc, a new environment variable, RNA_SEQC, will be set. Thus, users can use the software by running the following command: java -jar $RNA_SEQC {other options}.

Relion

Relion (REgularised LIkelihood OptimisatioN) is a stand-alone computer program for the refinement of 3D reconstructions or 2D class averages in electron cryo-microscopy.

Availability and Restrictions

Versions

Relion is available on the Owens cluster. The versions currently available at OSC are:

VERSION	OWENS
2.0	X*

* Current Default Version

You can use module spider relion to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Relion is available to all OSC users.

Usage

Usage on Owens

Set-up

To set up the environment for Relion on the Owens cluster, use the command:

module load relion2/2.0

SAMtools

SAM format is a generic format for storing large nucleotide sequence alignments. SAMtools provide various utilities for manipulating alignments in the SAM format, including sorting, merging, indexing and generating alignments in a per-position format.

Availability and Restrictions

The following versions of SAMtools are available on OSC clusters:

Version	Oakley	Owens
1.3.1	X*	X*
1.6	X	X

* Current default version

You can use module spider samtools to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

SAMtools is available to all OSC users without restriction.

Usage on Oakley

Set-up

To configure your environment for use of SAMtools, run the following command: module load samtools . The default version will be loaded. To select a particular SAMtools version, use module load samtools/version . For example, use module load samtools/1.3.1 to load SAMtools 1.3.1.

Usage on Owens

Set-up

To configure your environment for use of SAMtools, run the following command: module load samtools . The default version will be loaded. To select a particular SAMtools version, use module load samtools/version . For example, use module load samtools/1.3.1 to load SAMtools 1.3.1.

SIESTA

SIESTA is both a method and its computer program implementation, to perform efficient electronic structure calculations and ab initio molecular dynamics simulations of molecules and solids. More information can be found from here.

Availability and Restrictions

Versions

SIESTA is available on the Owens and Oakley clusters. A serial and a parallel build were created in order to meet users' computational needs.

VERSION	OAKLEY	Owens
3.2-pl4	X
3.2-pl5	X*
4.0		X*

* Current default version

You can use module spider siesta to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

The SIESTA program is distributed freely to academics. Use of SIESTA for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further insruction. Commercial users need to show proof of license in order to use SIESTA at OSC.

Usage

Batch Usage

When you log into oakley.osc.edu or owens.osc.edu, you are actually logged into a linux box referred to as the login node. To gain access to the 4000+ processors in the computing environment, you must submit your SIESTA job to the batch system for execution.

Assume that you have a test case in your work directory (where you submit your job, represented by $PBS_O_WORKDIR), with the input file 32_h2o.fdf. A batch script can be created and submitted for a serial or parallel run. The following are the sample batch scripts for running serial and parallel SIESTA jobs.

Sample Batch Script for Serial Jobs

#PBS -l walltime=0:30:00
#PBS -l nodes=1:ppn=12
#PBS -N siesta
#PBS -j oe
#
cd $PBS_O_WORKDIR
#
# Set up the package environment
module load siesta
#
# Execute the serial solver (nodes=1, ppn<=12)
siesta <32_h2o.fdf> output
exit

NOTE: Change ppn to = 28 for Owens

Sample Batch Script for Parallel Jobs

#PBS -l walltime=0:30:00
#PBS -l nodes=2:ppn=12
#PBS -N siesta
#PBS -j oe
#
cd $PBS_O_WORKDIR
#
# Set up the package environment
module swap intel/12.1.4.319 intel/13.1.3.192
module load siesta_par
#
# Execute the parallel solver (nodes>1, ppn=12)
mpiexec -np 24 siesta <32_h2o.fdf> output
exit

NOTE: Change ppn to = 28 for Owens

Citations

1. “Self-consistent order-N density-functional calculations for very large systems”, P. Ordejón, E. Artacho and J. M. Soler, Phys. Rev. B (Rapid Comm.) 53, R10441-10443 (1996).

2. “The SIESTA method for ab initio order-N materials simulation”, J. M. Soler, E. Artacho,J. D. Gale, A. García, J. Junquera, P. Ordejón, and D. Sánchez-Portal, J. Phys.: Condens. Matt. 14, 2745-2779 (2002).

Service:

Software

Fields of Science:

Bioinformatics & Biology

Chemical Engineering & Chemistry

Electrical Engineering

Physics

SPRNG

Computational stochastic approaches (Monte Carlo methods) based on random sampling are becoming extremely important research tools not only in their "traditional" fields such as physics, chemistry or applied mathematics but also in social sciences and, recently, in various branches of industry. An indication of importance is, for example, the fact that Monte Carlo calculations consume about one half of the supercomputer cycles. One of the indispensable and important ingredients for reliable and statistically sound calculations is the source of pseudo random numbers. SPRNG provides a scalable package for parallel pseudo random number generation which will be easy to use on a variety of architectures, especially in large-scale parallel Monte Carlo applications.

SPRNG 1.0 provides the user the various SPRNG random number generators each in its own library. For most users this is acceptable, as one rarely uses more than one type of generator in a single program. However, if the user desires this added flexibility, SPRNG 2.0 provides it. In all other respects, SPRNG 1.0 and SPRNG 2.0 are identical.

Availability and Restrictions

Versions

SPRNG is available on Oakley Cluster. The versions currently available at OSC are:

Version	Oakley
2.0	X*

* Current default version

You can use module spider sprng to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

SPRNG is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

Initalizing the system for use of SPRNG is dependent on the system you are using and the compiler you are using. To configure your environment for use of SPRNG 2.0 on Oakley, use the following command: module load sprng/2.0. You can use module avail sprng to view available modules on Oakley.

Note: SPRNG is dependent on the compiler and mpi module and may not be supported for all combinations.

Building With SPRNG 2.0 on Oakley

When SPRNG 2.0 is loaded, the following variables will be defined:

VARIABLE	Function
`$SPRNG_INCLUDE`	Include path for SPRNG 2.0 header files
`$SPRNG_LIBS`	Library path and library for SPRNG 2.0 generators

SRA Toolkit

The Sequence Read Archive (SRA Toolkit) stores raw sequence data from "next-generation" sequencing technologies including 454, IonTorrent, Illumina, SOLiD, Helicos and Complete Genomics. In addition to raw sequence data, SRA now stores alignment information in the form of read placements on a reference sequence. Use SRA Toolkit tools to directly operate on SRA runs.

Availability and Restrictions

The following versions of SRA Toolkit are available on OSC clusters:

Version	Oakley	Owens
2.6.3	X*	X*

* Current default version

You can use module spider sratoolkit to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

SRA Toolkit is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of SRA Toolkit, run the following command: module load sratoolkit. The default version will be loaded. To select a particular SRA Toolkit version, use module load sratoolkit/version. For example, use module load sratoolkit/2.6.3 to load SRA Toolkit 2.6.3.

Usage on Owens

Set-up

To configure your environment for use of SRA Toolkit, run the following command: module load sratoolkit. The default version will be loaded. To select a particular SRA Toolkit version, use module load sratoolkit/version. For example, use module load sratoolkit/2.6.3 to load SRA Toolkit 2.6.3.

STAR

STAR: Spliced Transcripts Alignment to a Reference.

Availability and Restrictions

Versions

The following versions of STAR are available on OSC clusters:

Version	Oakley	Owens
2.5.2a	X*	X*

* Current default version

You can use module spider star to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

STAR is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of STAR, run the following command: module load star. The default version will be loaded. To select a particular STAR version, use module load star/version. For example, use module load star/2.5.2a to load STAR 2.5.2a.

Usage on Owens

Set-up

To configure your environment for use of STAR, run the following command: module load star. The default version will be loaded. To select a particular STAR version, use module load star/version. For example, use module load star/2.5.2a to load STAR 2.5.2a.

STAR-CCM+

STAR-CCM+ provides the world’s most comprehensive engineering physics simulation inside a single integrated package. Much more than a CFD code, STAR‑CCM+ provides an engineering process for solving problems involving flow (of fluids and solids), heat transfer and stress. STAR‑CCM+ is unrivalled in its ability to tackle problems involving multi‑physics and complex geometries. Support is provided by CD-adapco. CD-adapco usually releases new version of STAR-CCM+ every four months.

Important Note: From March 14, 2018, STAR-CCM+ is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

Versions

STAR-CCM+ is available on Oakley Cluster. The versions currently available at OSC are:

VERSION	ruby	Owens
10.04.009	X
11.02.010	X	X
11.06.011	X*	X*
12.04.010		X
12.06.010		X
13.02.011		X

* Current default version

We have STAR-CCM+ Academic Pack, which includes STAR-CCM+, STAR-innovate, CAD Exchange, STAR-NX, STAR-CAT5, STAR-Inventor, STAR-ProE, JTOpen Reader, EHP, Admixturs, Vsim, CAT, STAR-ICE, Battery Design Studio, Sattery Simulation Module, SPEED, SPEED/Enabling PC-FEA, SPEED/Optimate, DARS, STAR-CD, STAR-CD/Reactive Flow Models, STAR-CD/Motion, esiece, and pro-STAR.

You can use module spider starccm to view available modules for a given machine. The default versions are in double precision. Please check with module spider starccm to see if there is a mixed precision version available. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

Academic users can use STAR-CCM+ on OSC machines if the user or user's institution has proper STAR-CCM+ license. Currently, users from Ohio State University, University of Cincinnati, University of Akron, and University of Toledo can access the OSC's license.

Use of STAR-CCM+ for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Currently, OSC has a 50 seat license (ccmpsuite, which allows up to 50 concurrent users), with 10,000 additional core licenses (hpcdomains) for academic users.

Access for Commerical Users

Contact OSC Help for getting access to STAR-CCM+ if you are a commerical user.

Usage

Usage on Ruby

Set-up on Ruby

We recommend to run STAR-CCM+ on only the compute nodes. Thus, all STAR-CCM+ jobs should be submitted via the batch scheduling system, either as interactive or non-interactive batch jobs. To load the default version of STAR-CCM+ module on Owens, use module load starccm. To select a particular software version, use module load starccm/version. For example, use module load starccm/11.02.010 to load STAR-CCM+ version 11.02.010 on Owens.

Batch Usage on Ruby

When you log into ruby.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your STAR-CCM+ analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used. STAR-CCM+ can be run on OSC clusters in either interactive mode or in non-interactive batch mode.

Interactive Batch Session

Interactive mode is similar to running STAR-CCM+ on a desktop machine in that the graphical user interface (GUI) will be sent from OSC and displayed on the local machine. To run interactive STAR-CCM+, it is suggested to request necessary compute resources from the login node, with X11 forwarding. The intention is that users can run STAR-CCM+ interactively for the purpose of building their model, preparing input file (.sim file), and checking results. Once developed this input file can then be run in no-interactive batch mode. For example, the following line requests one node with 20 cores(-l nodes=1:ppn=20), for a walltime of one hour (-l walltime=1:00:00), with one STAR-CCM+ base license token (-l software=starccm+1) on Owens:

qsub -I -X -l walltime=1:00:00 -l nodes=1:ppn=20 -l software=starccm+1

This job will queue until resources become available. Once the job is started, you're automatically logged in on the compute node; and you can launch STAR-CCM+ GUI with the following commands:

module load starccm
starccm+ -mesa

Non-interactive Batch Job (Serial Run using 1 Base Token)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice.

Below is the example batch script (job.txt) for a serial run with an input file (starccm.sim) on Owens:

#PBS -N star-ccm_test  
#PBS -l walltime=10:00:00  
#PBS -l nodes=1:ppn=1  
#PBS -l software=starccm+1
#PBS -j oe
#PBS -S /bin/bash
cd $TMPDIR  
cp $PBS_O_WORKDIR/starccm.sim .  
module load starccm  
starccm+ -batch starccm.sim >&output.txt  
cp output.txt $PBS_O_WORKDIR

To run this job on OSC batch system, the above script is to be submitted with the command:

qsub job.txt

Non-interactive Batch Job (Parallel Run using HPC Tokens)

To take advantage of the powerful compute resources at OSC, you may choose to run distributed STAR-CCM+ for large problems. Multiple nodes and cores can be requested to accelerate the solution time. The following shows an example script if you need 2 nodes with 20 cores per node on Owens using the inputfile named starccm.sim:

#PBS -N starccm_test 
#PBS -l walltime=3:00:00 
#PBS -l nodes=2:ppn=20
#PBS -l software=starccm+1%starccmpar+40
#PBS -j oe
#PBS -S /bin/bash
cd $PBS_O_WORKDIR
cp starccm.sim $TMPDIR
cd $TMPDIR
module load starccm
starccm+ -np 40 -rsh /usr/local/bin/pbsrsh -batch -machinefile $PBS_NODEFILE starccm.sim >&output.txt
cp output.txt $PBS_O_WORKDIR

Note that one needs to use "-rsh /usr/local/bin/pbsrsh" for the parallel runs. In addition to requesting the STAR-CCM+ base license token (-l software=starccm+1), you need to request copies of the starccmpar license, i.e., HPC tokens (-l software=starccm+1%starccmpar+[n]), where [n] is equal to the number of cores.

Usage on Owens

Set-up on Owens

We recommend to run STAR-CCM+ on only the compute nodes. Thus, all STAR-CCM+ jobs should be submitted via the batch scheduling system, either as interactive or non-interactive batch jobs. To load the default version of STAR-CCM+ module on Owens, use module load starccm. To select a particular software version, use module load starccm/version. For example, use module load starccm/11.02.010 to load STAR-CCM+ version 11.02.010 on Owens.

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your STAR-CCM+ analysis to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used. STAR-CCM+ can be run on OSC clusters in either interactive mode or in non-interactive batch mode.

Interactive Batch Session

Interactive mode is similar to running STAR-CCM+ on a desktop machine in that the graphical user interface (GUI) will be sent from OSC and displayed on the local machine. To run interactive STAR-CCM+, it is suggested to request necessary compute resources from the login node, with X11 forwarding. The intention is that users can run STAR-CCM+ interactively for the purpose of building their model, preparing input file (.sim file), and checking results. Once developed this input file can then be run in no-interactive batch mode. For example, the following line requests one node with 28 cores(-l nodes=1:ppn=28), for a walltime of one hour (-l walltime=1:00:00), with one STAR-CCM+ base license token (-l software=starccm+1) on Owens:

qsub -I -X -l walltime=1:00:00 -l nodes=1:ppn=28 -l software=starccm+1

This job will queue until resources become available. Once the job is started, you're automatically logged in on the compute node; and you can launch STAR-CCM+ GUI with the following commands:

module load starccm
starccm+ -mesa

Non-interactive Batch Job (Serial Run using 1 Base Token)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice.

Below is the example batch script (job.txt) for a serial run with an input file (starccm.sim) on Owens:

#PBS -N star-ccm_test  
#PBS -l walltime=10:00:00  
#PBS -l nodes=1:ppn=1  
#PBS -l software=starccm+1
#PBS -j oe
#PBS -S /bin/bash
cd $TMPDIR  
cp $PBS_O_WORKDIR/starccm.sim .  
module load starccm  
starccm+ -batch starccm.sim >&output.txt  
cp output.txt $PBS_O_WORKDIR

To run this job on OSC batch system, the above script is to be submitted with the command:

qsub job.txt

Non-interactive Batch Job (Parallel Run using HPC Tokens)

To take advantage of the powerful compute resources at OSC, you may choose to run distributed STAR-CCM+ for large problems. Multiple nodes and cores can be requested to accelerate the solution time. The following shows an example script if you need 2 nodes with 28 cores per node on Owens using the inputfile named starccm.sim :

#PBS -N starccm_test 
#PBS -l walltime=3:00:00 
#PBS -l nodes=2:ppn=28
#PBS -l software=starccm+1%starccmpar+56
#PBS -j oe
#PBS -S /bin/bash
cd $PBS_O_WORKDIR
cp starccm.sim $TMPDIR
cd $TMPDIR
module load starccm
starccm+ -np 56 -rsh /usr/local/bin/pbsrsh -batch -machinefile $PBS_NODEFILE starccm.sim >&output.txt
cp output.txt $PBS_O_WORKDIR

Note that one needs to use "-rsh /usr/local/bin/pbsrsh" for the parallel runs on Ownes. In addition to requesting the STAR-CCM+ base license token (-l software=starccm+1), you need to request copies of the starccmpar license, i.e., HPC tokens (-l software=starccm+1%starccmpar+[n]), where [n] is equal to the number of cores.

Run STAR-CCM+ to STAR-CCM+ Coupling

This documentation is to discuss how to run STAR-CCM+ to STAR-CCM+ Coupling simulation in batch job at OSC. The following example demonstrates the process of using STAR-CCM+ version 11.02.010 on Owens. Depending on the version of STAR-CCM+ and cluster you work on, there mighe be some differences from the example. Feel free to contact OSC Help if you have any questions.

Prepare Lagging Simulation

Launch the STAR-CCM+ GUI following the instructions on this page
Load the simulation that lags and prepare the lagging simulation following the STAR-CCM+ User Guide
- Active a co-simulation model
- Set "Concurrency mode -> Method" to Lag
- Other setups
Save the lagging simulation and name it for example as lag.sim

Prepare Leading Simulation

Load the simulation that leads and prepare the leading simulation following the STAR-CCM+ User Guide
- Active a co-simulation model
- Set "Concurrency mode -> Method" to Lead
Go to the "Connect Method" node by selecting "Co-Simulations -> <name of co-simulation> -> Conditions". Click "Edit" of "Connect Method". In "Connect Method" node, select "Launch Application and Connect" under method. Under "Launch Application and Connect", put the following information as "Launch Command":

/usr/local/starccm/11.02.010/STAR-CCM+11.02.010-R8/star/bin/starccm+ -load -server -rsh /usr/local/bin/pbsrsh lag.sim

See the picture below:

connect method

Save the leading simulation and name it for example as lead.sim

Prepare Job Script

In the job script, use the following command to run the co-simulation:

starccm+ -np N,M -rsh /usr/local/bin/pbsrsh -batch -machinefile $PBS_NODEFILE lead.sim

where N is # of cores for the leading simulation and M is # of cores for the lagging simulation, and the summation of N and M should be the total number of cores you request in the job.

Once the job is completed, the output results of the leading simulation will be returned, while the lagging simulation runs on the background server and the final results won't be saved.

Service:

Software

Fields of Science:

Aerospace Engineering

Mechanical Engineering

STAR-Fusion

STAR-Fusion is a component of the Trinity Cancer Transcriptome Analysis Toolkit (CTAT). STAR-Fusion uses the STAR aligner to identify candidate fusion transcripts supported by Illumina reads. STAR-Fusion further processes the output generated by the STAR aligner to map junction reads and spanning reads to a reference annotation set.

Availability and Restrictions

Versions

The following versions of STAR-Fusion are available on OSC clusters:

Version	Oakley	Owens
0.7.0	X*	X*

* Current default version

You can use module spider star-fusion to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

STAR-Fusion is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of STAR-Fusion, run the following command: module load star-fusion. The default version will be loaded. To select a particular STAR-Fusion version, use module load star-fusion/version. For example, use module load star-fusion/0.7.0 to load STAR-Fusion 0.7.0.

Usage on Owens

Set-up

To configure your environment for use of STAR-Fusion, run the following command: module load star-fusion. The default version will be loaded. To select a particular STAR-Fusion version, use module load star-fusion/version. For example, use module load star-fusion/0.7.0 to load STAR-Fusion 0.7.0.

Salmon

Salmon is a tool for quantifying the expression of transcripts using RNA-seq data.

Availability and Restrictions

Versions

Salmon is available on the Owens Cluster. The versions currently available at OSC are:

VERSION	OWENS
0.8.2	X*

* Current default version

You can use module spider salmonto view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Salmon is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of Salmon, use command module load salmon. This will load the default version.

ScaLAPACK

ScaLAPACK is a library of high-performance linear algebra routines for clusters supporting MPI. It contains routines for solving systems of linear equations, least squares problems, and eigenvalue problems.

This page documents usage of the ScaLAPACK library installed by OSC from source. An optimized implementation of ScaLAPACK is included in MKL; see the software documentation page for Intel Math Kernel Library for usage information.

Availability and Restrictions

Versions

The versions currently available at OSC are:

Version	Oakley	ruby	owens	notes
2.0.1	X			Default version on Oakley prior to 09/15/2015
2.0.2	X*	X*	X*

* Current default version

You can use module spider scalapack to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

ScaLAPACK is available to all OSC users without restriction. If you need high performance, we recommend using MKL instead of the standalone ScaLAPACK module.

Usage

Usage on Oakley

Set-up

Initalizing the system for use of the ScaLAPACK libraries is dependent on the system you are using and the compiler you are using. To use the ScaLAPACK libraries in your compilation, run the following command: module load scalapack. To load a particular version, use module load scalapack/version. For example, use module load scalapack/2.0.2 to load ScaLAPACK version 2.0.2. You can use module spider scalapack to view available modules.

Building with ScaLAPACK

Once loaded, the ScaLAPACK libraries can be linked in with your compilation. To do this, use the following environment variables. You must also link with MKL. With the Intel compiler, just add -mkl to the end of the link line. With other compilers, load the mkl module and add $MKL_LIBS to the end of the link line.

Variable	Use
`$SCALAPACK_LIBS`	Used to link ScaLAPACK into either Fortran or C

Usage on Ruby

Set-up

Initalizing the system for use of the ScaLAPACK libraries is dependent on the system you are using and the compiler you are using. To use the ScaLAPACK libraries in your compilation, run the following command: module load scalapack. To load a particular version, use module load scalapack/version. For example, use module load scalapack/2.0.2 to load ScaLAPACK version 2.0.2. You can use module spider scalapack to view available modules.

Building with ScaLAPACK

Once loaded, the ScaLAPACK libraries can be linked in with your compilation. To do this, use the following environment variables. You must also link with MKL. With the Intel compiler, just add -mkl to the end of the link line. With other compilers, load the mkl module and add $MKL_LIBS to the end of the link line.

Variable	Use
`$SCALAPACK_LIBS`	Used to link ScaLAPACK into either Fortran or C

Usage on Owens

Set-up

Initalizing the system for use of the ScaLAPACK libraries is dependent on the system you are using and the compiler you are using. To use the ScaLAPACK libraries in your compilation, run the following command: module load scalapack. To load a particular version, use module load scalapack/version. For example, use module load scalapack/2.0.2 to load ScaLAPACK version 2.0.2. You can use module spider scalapack to view available modules.

Building with ScaLAPACK

Once loaded, the ScaLAPACK libraries can be linked in with your compilation. To do this, use the following environment variables. You must also link with MKL. With the Intel compiler, just add -mkl to the end of the link line. With other compilers, load the mkl module and add $MKL_LIBS to the end of the link line.

Variable	Use
`$SCALAPACK_LIBS`	Used to link ScaLAPACK into either Fortran or C

Schrodinger

The Schrodinger molecular modeling software suite includes a number of popular programs focused on drug design and materials science but of general applicability, for example Glide, Jaguar, and MacroModel. Maestro is the graphical user interface for the suite. It allows the user to construct and graphically manipulate both simple and complex chemical structures, to apply molecular mechanics and dynamics techniques to evaluate the energies and geometries of molecules in vacuo or in solution, and to display and examine graphically the results of the modeling calculations.

Important Note: From March 14, 2018, Schrodinger is available from Owens. If you have any questions, please contact OSC Help.

Availability and Restrictions

Versions

The Schrodinger suite is available on Oakley and Owens. The versions currently available at OSC are:

Version	Owens
15	X
16	X*

* Current default version

You can use module spider schrodinger to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Schrodinger is available to all academic users.

To use Schrodinger you will have to be added to the license server first. Please contact OSC Help to be added. Please note that if you are a non-OSU user, we need to send your name, contact email, and affiliation information to Schrodinger in order to grant access. Currently, we have license for following features:

CANVAS_ELEMENTS
CANVAS_MAIN
CANVAS_SHARED
COMBIGLIDE_MAIN
EPIK_MAIN
FFLD_OPLS_MAIN
GLIDE_MAIN
GLIDE_XP_DESC
IMPACT_MAIN
KNIME_MAIN
LIGPREP_MAIN
MAESTRO_MAIN
MMLIBS
MMOD_CONFGEN
MMOD_MACROMODEL
MMOD_MBAE
QIKPROP_MAIN

You need to use one of following software flags in order to use the particular feature of the software without license errors.

maestro, macromodel, glide, ligprep, qikprep, epik

You can add -l software=glide+1 to your job script if you use GLIDE for example. When you use this software flag, then your job won't start until it secures available licenses. Please read the batch script examples down below.

Usage

Usage on Owens

To set up your environment for schrodinger load one of its modulefiles:

module load schrodinger/schrodinger16

Using schrodinger interactively requires an X11 connection. Typically one will launch the graphical user interface:

maestro

Here is an example batch script that uses schrodinger non-interactively via the batch system:

# Example glide single node batch script.
#PBS -N glidebatch
#PBS -j oe
#PBS -m ae
#PBS -M srb@osc.edu
#PBS -l walltime=0:10:00
#PBS -l nodes=1:ppn=28
#PBS -l software=glide+1
#PBS -S /bin/sh
qstat -f $PBS_JOBID
export
module load schrodinger
cd $PBS_O_WORKDIR
pbsdcp -p '*' $TMPDIR
cd $TMPDIR
host=`cat $PBS_NODEFILE|head -1`
nproc=`cat $PBS_NODEFILE|wc -l`
glide -WAIT -HOST ${host}:${nproc} receptor_glide.in
ls -l
pbsdcp -p '*' $PBS_O_WORKDIR

The glide command passes control to the Schrodinger Job Control utility which processes the two options: The WAIT option forces the glide command to wait until all tasks of the command are completed. This is necessary for PBS batch jobs to run effectively. The HOST option specifies how tasks are distributed over processors.

Singularity

Singularity is a container system designed for use on High Performance Computing (HPC) systems. It allows users to run both Docker and Singularity containers.

From the Docker website: "A container image is a lightweight, stand-alone, executable package of a piece of software that includes everything needed to run it: code, runtime, system tools, system libraries, settings."

Availability and Restrictions

Versions

Singularity is available on the Owens cluster. Only one version is available at any given time. To find out the current version:

module load singularity
singularity --version

Access

Singularity is available to all OSC users.

Usage

Set-up

To set up the environment for Singularity on the OSC clusters, use the command:

module load singularity

Using Singularity

See HOWTO: Use Docker and Singularity Containers at OSC for information about using Singularity on the Owens cluster, including some site-specific caveats.

Example: Run a container from the Singularity hub

[owens-login01]$ singularity run shub://vsoch/hello-world
Progress |===================================| 100.0%
RaawwWWWWWRRRR!!
[owens-login01]$

SnpEff

SnpEff is a variant annotation and effect prediction tool. It annotates and predicts the effects of variants on genes (such as amino acid changes).

Availability and Restrictions

Versions

The following versions of SnpEff are available on OSC clusters:

Version	Oakley	owens
4.2	X*	X*

* Current default version

You can use module spider snpeff to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

SnpEff is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of SnpEff, run the following command: module load snpeff. The default version will be loaded. To select a particular SnpEff version, use module load snpeff/version. For example, use module load snpeff/4.2 to load SnpEff 4.2.

Usage

This software consists of Java executable .jar files; thus, it is not possible to add to the PATH environment variable.

From module load snpeff, new environment variables, SNPEFF and SNPSIFT, will be set. Thus, users can use the software by running the following command: java -jar $SNPEFF {other options}, or java -jar $SNPSIFT {other options}.

Usage on Owens

Set-up

To configure your environment for use of SnpEff, run the following command: module load snpeff. The default version will be loaded. To select a particular SnpEff version, use module load snpeff/version. For example, use module load snpeff/4.2 to load SnpEff 4.2.

Usage

This software consists of Java executable .jar files; thus, it is not possible to add to the PATH environment variable.

From module load snpeff, new environment variables, SNPEFF and SNPSIFT, will be set. Thus, users can use the software by running the following command: java -jar $SNPEFF {other options}, or java -jar $SNPSIFT {other options}.

Spark

Apache Spark is an open source cluster-computing framework originally developed in the AMPLab at University of California, Berkeley but was later donated to the Apache Software Foundation where it remains today. In contrast to Hadoop's disk-based analytics paradigm, Spark has multi-stage in-memory analytics. Spark can run programs upto 100x faster than Hadoop’s MapReduce in memory or 10x faster on disk. Spark support applications written in python, java, scala and R

Availability and Restrictions

Versions

The following versions of Spark are available on OSC systems:

VERSION	OAKLEY	OWENS
1.5.2	X
1.6.1	X
2.0.0	X*	X*
2.1.0		X
2.3.0		X

* Current default version

You can use module spider spark to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Spark is available to all OSC users without restriction.

Usage

Set-up

In order to configure your environment for the usage of Spark, run the following command:

module load spark

A particular version of Spark can be loaded as follows

module load spark/2.3.0

Using Spark

In order to run Spark in batch, reference the example batch script below. This script requests 6 node on the Owens cluster for 1 hour of walltime. The script will submit the pyspark script called test.py using pbs-spark-submit command into the PBS queue.

#PBS -N Spark-example

#PBS -l nodes=6:ppn=28

#PBS -l walltime=01:00:00

module load spark

cd $PBS_O_WORKDIR

cp test.py $TMPDIR

cd $TMPDIR 

pbs-spark-submit test.py  > test.log

cp * $PBS_O_WORKDIR

pbs-spark-submit script is used for submitting Spark jobs into PBS queue. For more options, please run,

pbs-spark-submit --help

Running Spark interactively in batch

To run Spark interactively, but in batch on Owens please run the following command,

qsub -I -l nodes=4:ppn=28 -l walltime=01:00:00

When your interactive shell is ready, please launch spark cluster using the pbs-spark-submit script

pbs-spark-submit

You can then launch pyspark by connecting to Spark master node as follows.

pyspark --master spark://nodename.ten.osc.edu:7070

Launching Jupyter+Spark on OSC OnDemand

Instructions on how to launch Spark on OSC OnDemand web interface is here. https://www.osc.edu/content/launching_jupyter_spark_app

Example Jobs

Please check /usr/local/src/spark/2.0.0/test.osc folder for more examples of pyspark script, job submission script and output files.

Stata

Stata is a complete, integrated statistical package that provides everything needed for data analysis, data management, and graphics. Release 13 32-processor SMP is currently available at OSC.

Availability and Restrictions

Versions

The following versions of Stata are available on OSC systems:

Version	Oakley
13	X*

* Current default version

You can use module spider stata to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Only academic use for OSU users is allowed. Please contact OSC Help to get validated for using Stata.

Usage

Set-up

To configure your environment on Oakley for the usage of Stata, run the following command:

module load stata

Using Stata

Due to licensing restrictions, Stata may ONLY be used via the batch system on Oakley. See below for information on how this is done.

Batch Usage

OSC has a 2-user license. However, there is no enforcement mechanism built into Stata. In order for us to stay within the 2-user limit, we require you to run in the context of PBS and to include this option when starting your batch job (the PBS system will enforce the 2 user limit):

#PBS -l software=stata

Non-Interactive batch example

Use the script below as a template for your usage.

#PBS -l walltime=01:00:00
#PBS -l nodes=1:ppn=12
#PBS -l software=stata
#PBS -N stata
#PBS -j oe

module load stata
cd $PBS_O_WORKDIR
stata-mp -b do bigjob

Interactive batch example

The following illustrates how to run Stata via an interactive batsh session. Please note that this requires an X11 server on the client system.

oakley01: qsub -X -I -l nodes=1:ppn=12 -l software=stata 
qsub: waiting for job 5860.oak-batch.osc.edu to start
qsub: job 5860.oak-batch.osc.edu ready

n0533: module load stata
n0533: xstata-mp
n0533: exit

Stata can take full advantage of our large 32 core, 1TB RAM node. Please see the Oakley documentation for information about how to access this node.

StringTie

StringTie assembles aligned RNA-Seq reads into transcripts that represent splice variants in RNA-Seq samples.

Availability and Restrictions

Versions

StringTie is available on the Owens Cluster. The versions currently available at OSC are:

VERSION	OWENS
1.3.3b	X*

* Current Default Version

You can use module spider stringtieto view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

StringTie is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of StringTie, use command module load stringtie. This will load the default version.

Subread

The Subread package comprises a suite of software programs for processing next-gen sequencing read data like Subread, Subjunc, featureCounts, and exactSNP.

Availability and Restrictions

Versions

The following versions of Subread are available on OSC clusters:

Version	Oakley	owens
1.5.0-p2	X*	X*

* Current default version

You can use module spider subread to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Subread is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of Subread, run the following command: module load subread. The default version will be loaded. To select a particular Subread version, use module load subread/version. For example, use module load subread/1.5.0-p2 to load Subread 1.5.0-p2.

Usage on Owens

Set-up

To configure your environment for use of Subread, run the following command: module load subread. The default version will be loaded. To select a particular Subread version, use module load subread/version. For example, use module load subread/1.5.0-p2 to load Subread 1.5.0-p2.

SuiteSparse

SuiteSparse is a suite of sparse matrix algorithms, including: UMFPACK(multifrontal LU factorization), CHOLMOD(supernodal Cholesky, with CUDA acceleration), SPQR(multifrontal QR) and many other packages.

Availability and Restrictions

Versions

OSC supports most packages in SuiteSparse, including UMFPACK, CHOLMOD, SPQR, KLU and BTF, Ordering Methods (AMD, CAMD, COLAMD, and CCOLAMD) and CSparse. SuiteSparse modules are available for the Intel, GNU, and Portland Group compilers. The following versions of SuiteSparse are available at OSC.

Version	oakley
4.4.4	X
4.5.3	X*

* Current default version

You can use module spider suitesparse to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

NOTE: SuiteSparse library on our clusters is built without METIS, which might matter if CHOLMOD package is included in your program.

Access

SuiteSparse is available without restriction to all OSC users.

Usage

Usage on Oakley

Set-up on Oakley

To use SuitSparse, ensure the correct compiler is loaded. User module spider suitesparse/version to view compatible compilers. Before loading the SuiteSparse library, MKL is also required. Load the MKL library with module load mkl. Then with the following command, SuiteSparse library is ready to be used: module load suitesparse.

Building With SuiteSparse

With the SuiteSparse library loaded, the following environment variables will be available for use:

Variable	Use
$SUITESPARSE_CFLAGS	Include flags for C or C++ programs.
$SUITESPARSE_LIBS	Use when linking your program to SuiteSparse library.

For example, to build the code my_prog.c with the SuiteSparse library you would use:

icc -c my_prog.c
icc -o my_prog my_prog.o $SUITESPARSE_LIBS

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Non-interactive Batch Job (Serial Run)

A batch script can be created and submitted for a serial or parallel run. You can create the batch script using any text editor you like in a working directory on the system of your choice. You must load the SuiteSparse module in your batch script before executing a program which is built with the SuiteSparse library. Below is the example batch script that executes a program built with SuiteSparse:

#PBS -N MyProgJob
#PBS -l nodes=1:ppn=12
module load gnu/4.8.4
module load mkl
module load suitesparse
cd $PBS_O_WORKDIR
cp foo.dat $TMPDIR
cd $TMPDIR
my_prog < foo.dat > foo.out
cp foo.out $PBS_O_WORKDIR

SuperLU

SuperLU is a library for the direct solution of large, sparse, nonsymmetric systems of linear equations on high performance machines. It comes in two different flavors: SuperLU_MT (multithreaded) for shared memory parallel machines and SuperLU_DIST for distributed memory parallel machines.

Availability and Restrictions

Versions

SuperLU is available on Oakley Cluster. It comes in two different flavors: SuperLU_MT (multithreaded) for shared memory parallel machines and SuperLU_DIST for distributed memory parallel machines. The versions currently available at OSC are:

Library	Version	Oakley	notes
SuperLU_MT	2.0	X	Default version on Oakley prior to 09/15/2015
SuperLU_MT	3.0	X
SuperLU_DIST	3.0	X	Default version on Oakley prior to 09/15/2015
SuperLU_DIST	3.1	X
SuperLU_DIST	4.1	X*

* Current Default Version

You can use module spider superlu to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

SuperLU is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

Initalizing the system for use of the SuperLU librarie is dependent on the system you are using and the compiler you are using. To use the SuperLU libraries in your compilation on Oakley, run the following command: module load superlu_mt. or module load superlu_dist. You can use module spider superlu to view available modules on Oakley.

Building With SuperLU on Oakley

Once loaded on Oakley, the SuperLU libraries can be linked in with your compilation. To do this, use the following environment variables:

Variable	Use
`$SUPERLU_MT_CFLAGS`	Include path for multithreaded libraries
`$SUPERLU_MT_LIBS`	Link flags for multithreaded libraries
`$SUPERLU_DIST_CFLAGS`	Include path for distributed-memory libraries
`$SUPERLU_DIST_LIBS`	Link flags for distributed-memory libraries

TAU Commander

TAU Commander is a user interface for the TAU Performance System, a set of tools for analyizing the performance of parallel programs.

Availability and Restrictions

Versions

TAU Commander is available on the Owens Cluster. The versions currently available at OSC are:

VERSION	OWENS
1.2.1	X
1.3.0	X*

* Current default version

You can use module spider taucmdr to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

TAU Commander is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of TAU Commander, use command module load taucmdr. This will load the default version.

Creating a project

The first step to use TAU Commander on your code is to create and configure a project. To create a project, use the command tau initalize. Additional options for compilers, MPI libraries, measurements. etc. are available.

For instance, to configure for Intel compilers use the command tau initialize --compilers Intel and to configure for MPI use tau initialize --mpi.

For more details about how to initialize your project use the command tau help initialize.

After running creating the project you should see a dashboard for your project with a target, application, and 3 default measurements. You can now create additional measurements or modify the application and target. See the TAU Commander user guide for more information about how to configure your project.

Compiling and Running your Code

To compile your code to run with TAU Commander, just add tau before the compiler. For instance, if you are compiling with gcc now compile with tau gcc. Similarly, when you run your code add tau before the run command. So, if you usually run with mpiexec -ppn 4 ./my_prog now run with tau mpiexec -ppn 4 ./my_prog. Each time the program is run with tau prepended, a new trial is created in the project with performance data for that run.

Post-processing

Once you have generated performance data with TAU Commander you have 3 options to view the trial performance data:

view the data in text format (not always available),
view the data in a GUI using an OnDemand VDI (Virtual Desktop Interface) or X11 forwarding enabled,
or export the data to your local machine (requires minimal installation of TAU commander on your local machine).

To view the data:

tau trial show trial_number

To export the data:

tau trial export trial_number

TensorFlow

"TensorFlow is an open source software library for numerical computation using data flow graphs. Nodes in the graph represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) that flow between them. This flexible architecture lets you deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device without rewriting code."

Quote from TensorFlow Github documentation.

Availability and Restrictions

Versions

The following version of TensorFlow is available on OSC clusters:

VERSION	OWENS
1.3 (using Python 3.6)	X

The current version of TensorFlow on Owens requires cuda/8.0.44 for GPU calculations. TensorFlow is a Python package and therefore requires loading module load python/3.6. The version of TensorFlow may actively change with updates to Anaconda Python on Owens so that you can check the latest version with conda list tensorflow.

Feel free to contact OSC Help if you need other versions for your work.

Access

TensorFlow is available to all OSC users without restriction.

Usage on Owens

Setup on Owens

TensorFlow package is installed using Anaconda Python 2. To configure the Owens cluster for the use of TensorFlow, use the following commands:

module load python/3.6 cuda/8.0.44

Batch Usage on Ruby or Owens

Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Owens, and Scheduling Policies and Limits for more info. In particular, TensorFlow should be run on a GPU-enabled compute node.

An Example of Using TensorFlow with MNIST model and Logistic Regression

Below is an example batch script (job.txt and logistic_regression_on_mnist.py) for using TensorFlow.

Contents of job.txt

#PBS -N TensorFlow
#PBS -l nodes=1:ppn=28:gpus=1:default
#PBS -l walltime=30:00
#PBS -j oe
#PBS -S /bin/bash 

cd $PBS_O_WORKDIR
module load python/3.6 cuda/8.0.44
python logistic_regression_on_mnist.py

Contents of logistic_regression_on_mnist.py

# logistic_regression_on_mnist.py Python script based on:
# https://github.com/aymericdamien/TensorFlow-Examples/blob/master/notebooks/0_Prerequisite/mnist_dataset_intro.ipynb
# https://github.com/aymericdamien/TensorFlow-Examples/blob/master/notebooks/2_BasicModels/logistic_regression.ipynb

import tensorflow as tf

# Import MNIST
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("data/", one_hot=True)

# Parameters
learning_rate = 0.01
training_epochs = 25
batch_size = 100
display_step = 1

# tf Graph Input
x = tf.placeholder(tf.float32, [None, 784]) # mnist data image of shape 28*28=784
y = tf.placeholder(tf.float32, [None, 10]) # 0-9 digits recognition => 10 classes

# Set model weights
W = tf.Variable(tf.zeros([784, 10]))
b = tf.Variable(tf.zeros([10]))

# Construct model
pred = tf.nn.softmax(tf.matmul(x, W) + b) # Softmax

# Minimize error using cross entropy
cost = tf.reduce_mean(-tf.reduce_sum(y*tf.log(pred), reduction_indices=1))
# Gradient Descent
optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)

# Initializing the variables
init = tf.global_variables_initializer()
# Launch the graph
with tf.Session() as sess:
    sess.run(init)

    # Training cycle
    for epoch in range(training_epochs):
        avg_cost = 0.
        total_batch = int(mnist.train.num_examples/batch_size)
        # Loop over all batches
        for i in range(total_batch):
            batch_xs, batch_ys = mnist.train.next_batch(batch_size)
            # Fit training using batch data
            _, c = sess.run([optimizer, cost], feed_dict={x: batch_xs,
                                                          y: batch_ys})
            # Compute average loss
            avg_cost += c / total_batch
        # Display logs per epoch step
        if (epoch+1) % display_step == 0:
            print ("Epoch:", '%04d' % (epoch+1), "cost=", "{:.9f}".format(avg_cost))

    print ("Optimization Finished!")

    # Test model
    correct_prediction = tf.equal(tf.argmax(pred, 1), tf.argmax(y, 1))
    # Calculate accuracy for 3000 examples
    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))
    print ("Accuracy:", accuracy.eval({x: mnist.test.images[:3000], y: mnist.test.labels[:3000]}))

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

TopHat

TopHat uses Bowtie, a high-throughput short read aligner, to analyze the mapping results for RNA-Seq reads and identify splice junctions.

Please note that tophat (and bowtie) cannot run in parallel, that is, on multiple nodes. Submitting multi-node jobs will only waste resources. In addition you must explicitly include the '-p' option to use multiple threads on a single node.

Availability and Restrictions

Versions

TopHat is available on the Owens Cluster. The versions currently available at OSC are:

VERSION	OWENS
2.1.1	X*

* Current default version

You can use module spider tophat to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

TopHat is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of TopHat, use command module load tophat. This will load the default version.

Torch

"Torch is a deep learning framework with wide support for machine learning algorithms. It's open-source, simple to use, and efficient, thanks to an easy and fast scripting language, LuaJIT, and an underlying C / CUDA implementation. Torch offers popular neural network and optimization libraries that are easy to use, yet provide maximum flexibility to build complex neural network topologies. It also runs up to 70% faster on the latest NVIDIA Pascal™ GPUs, so you can now train networks in hours, instead of days."

Quote from Torch documentation.

Availability and Restrictions

Versions

The following version of Torch is available on OSC cluster:

VERSION	OWENS
7	X

* Current default version

You can use module spider torch to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

The current version of Torch on Owens requires cuda/8.0.44 and CUDNN v5 for GPU calculations.

Access

Torch is available to all OSC users without restriction.

Usage

Usage on Owens

Setup on Owens

To configure the Owens cluster for the use of Torch, use the following commands:

module load torch

Batch Usage on Ruby or Owens

Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Owens, and Scheduling Policies and Limits for more info. In particular, Torch should be run on a GPU-enabled compute node.

An Example of Using Torch with CIFAR10 Training Data on Owens

Below is an example batch script (job.txt) for using Torch. Please see the reference https://github.com/szagoruyko/cifar.torch for more details.

#PBS -N Torch
#PBS -l nodes=1:ppn=28:gpus=1:default
#PBS -l walltime=30:00
#PBS -j oe
# Load module load for torch
module load torch
# Migrate to job temp directory 
cd $TMPDIR
# Clone sample data and scripts
git clone https://github.com/szagoruyko/cifar.torch.git .
# Run the image preprocessing (not necessary for subsequent runs, just re-use provider.t7)
OMP_NUM_THREADS=28 th -i provider.lua <<Input
provider = Provider()
provider:normalize()
torch.save('provider.t7',provider)
exit
y
Input
# Run the torch training
th train.lua --backend cudnn
# Copy results from job temp directory
cp -a * $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

TotalView Debugger

TotalView is a symbolic debugger which supports threads, MPI, OpenMP, C/C++, CUDA and Fortran, plus mixed-language codes. Advanced features include on-demand memory leak detection, heap allocation debugging and the Standard Template Library Viewer (STLView). Other features such as dive (which quickly shows more detail), a wide variety of breakpoints, the Message Queue Graph/Visualizer, powerful data analysis and control at the thread level enable debugging of parallel programs.

Availability and Restrictions

Versions

The versions currently available at OSC are:

Version	Oakley	Ruby	Owens	notes
8.9.2-1	X(CS)			Default version on Oakley prior to 09/15/2015
8.14.1-8	X	X
2016.04.08			X(S)
2017.0.12	X(S)*	X(CS)*	X(CS)*	New user Interface options available
2017.2.11			X(CS)

* Current default version; X(S) = standard version; X(CS) = both standard and CUDA versions are available; X = CUDA debugging included in standard module

You can use module spider totalview to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access and Licensing

TotalView is available for use by all OSC users. The installed license permits the use of 32 tokens (processes). This means that the maximum parallel tasks for a TotalView session is 32. If the token limit is exceeded an error with the text "Unable to acquire a license for X processes." will be presented in a message box. If you need to debug a program with more than 32 processes you can attach to a subset of the processes.

Usage

General Usage Notes

TotalView provides a graphical user interface (GUI) as well as a textual user interface. (In general, the command totalview, which invokes the GUI, can be swapped with the command totalviewcli, which invokes the Command-Line Interface.) For the GUI application, one should use an OnDemand VDI (Virtual Desktop Interface) or have X11 forwarding enabled (see Setting up X Windows). Note that X11 forwarding can be distractingly slow for interactive applications.

Debugging is typically done in an interactive batch environment. This is a requirement for MPI programs and is also important for multithreaded programs. See our documentation on Interactive Batch jobs.

Note: Debugging of CUDA 8.0 programs requires at least version 2017 of TotalView.

Usage on Oakley

Building your code for debugging with TotalView

You should build your code with the -g option for symbolic debugging. For example:

icc -g -o mycode mycode.c

or

mpicc -g -o mympicode mympicode.c

Set-up

To load the default standard version of Totalview module, use module load totalview. To select a particular software version, use module load totalview/version. For example, use module load totalview/8.9.2-1 to load Totalview version 8.9.2-1.

Using Totalview

To debug a serial or multithreaded program with TotalView, load the module and launch using the following commands. The -a argument is used only if you want to pass arguments to your program on the command line.

module load totalview
totalview mycode -a prog_args

Example code to debug can be found in $TOTALVIEW_HOME/linux-x86-64/examples. Refer to the file README.TXT in this directory for instructions for using these examples.

To debug an MPI program, use the following commands in an interactive job:

module load totalview
mpiexec -tv [other mpiexec arguments] mympiprog [program arguments]

Totalview will stop your program at the beginning to allow you to debug. For a parallel program, the initial breakpoint is in mpiexec. Press "Go" and you will be given the option of stopping when your job goes parallel. Normally it will stop inside MPI_Init. For a parallel job running on a single node, however, it can run past that point and stop later in your program. You may need to have your program sleep for 30 seconds at the beginning.

In C:

sleep(30);

In Fortran:

call sleep(30)

Usage on Ruby

Building your code for debugging with TotalView

You should build your code with the -g option for symbolic debugging. For example:

icc -g -o mycode mycode.c

or

mpicc -g -o mympicode mympicode.c

Set-up

To load the default standard version of Totalview module, use module load totalview. To select a particular software version, use module load totalview/version. For example, use module load totalview/8.14.1-8 to load Totalview version 8.14.1-8.

Using Totalview

To debug a serial or multithreaded program with TotalView, load the module and launch using the following commands. The -a argument is used only if you want to pass arguments to your program on the command line.

module load totalview
totalview ./mycode -a prog_args

Example code to debug can be found in $TOTALVIEW_HOME/linux-x86-64/examples. Refer to the file README.TXT in this directory for instructions for using these examples.

Use the following commands in an interactive job:

module load totalview
totalview -args mpiexec [mpiexec arguments] ./mympiprog [program arguments]

Totalview will stop your program at the beginning to allow you to debug. For a parallel program, the initial breakpoint is in mpiexec. Press "Go" and you will be given the option of stopping when your job goes parallel. Normally it will stop inside MPI_Init. For a parallel job running on a single node, however, it can run past that point and stop later in your program. You may need to have your program sleep for 30 seconds at the beginning.

In C:

sleep(30);

In Fortran:

call sleep(30)

Usage on Owens

Building your code for debugging with TotalView

You should build your code with the -g option for symbolic debugging. For example:

icc -g -o mycode mycode.c

or

mpicc -g -o mympicode mympicode.c

Set-up

To load the default standard version of Totalview module, use module load totalview. To select a particular software version, use module load totalview/version. For example, use module load totalview/2016.04.08 to load Totalview version 2016.04.08.

Using Totalview

To debug a serial or multithreaded program with TotalView, load the module and launch using the following commands. The -a argument is used only if you want to pass arguments to your program on the command line.

module load totalview
totalview ./mycode -a prog_args

Example code to debug can be found in $TOTALVIEW_HOME/linux-x86-64/examples. Refer to the file README.TXT in this directory for instructions for using these examples.

Use the following commands in an interactive job:

module load totalview
totalview -args mpiexec [mpiexec arguments] ./mympiprog [program arguments]

Totalview will stop your program at the beginning to allow you to debug. For a parallel program, the initial breakpoint is in mpiexec. Press "Go" and you will be given the option of stopping when your job goes parallel. Normally it will stop inside MPI_Init. For a parallel job running on a single node, however, it can run past that point and stop later in your program. You may need to have your program sleep for 30 seconds at the beginning.

In C:

sleep(30);

In Fortran:

call sleep(30)

TrimGalore

TrimGalore is a wrapper tool that automates quality and adapter trimming to FastQ files. It also provides functionality to RRBS sequence files.

Availability and Restrictions

Versions

TrimGalore is available on the Owens cluster. The versions currently available at OSC are:

VERSION	OWENS
0.4.5	X*

* Current default version

You can use module spider trimgalore to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

TrimGalore is available without restriction to all OSC users.

Usage

Usage on Owens

Set-up

To configure your enviorment for use of TrimGalore, use command module load trimgalore. This will load the default version.

Trimmomatic

Trimmomatic performs a variety of useful trimming tasks for illumina paired-end and single ended data.The selection of trimming steps and their associated parameters are supplied on the command line.

Availability and Restrictions

Versions

The following versions of Trimmomatic are available on OSC clusters:

Version	Oakley	Owens
0.36	X*	X*

* Current default version

You can use module spider trimmomatic to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

Trimmomatic is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of Trimmomatic, run the following command: module load trimmomati. The default version will be loaded. To select a particular Trimmomatic version, use module load trimmomatic/version. For example, use module load trimmomatic/0.36 to load Trimmomatic 0.36.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load trimmomatic, a new environment variable, TRIMMOMATIC, will be set.

Thus, users can use the software by running the following command: java -jar $TRIMMOMATIC {other options}.

Usage on Owens

Set-up

To configure your environment for use of Trimmomatic, run the following command: module load trimmomati. The default version will be loaded. To select a particular Trimmomatic version, use module load trimmomatic/version. For example, use module load trimmomatic/0.36 to load Trimmomatic 0.36.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load trimmomatic, a new environment variable, TRIMMOMATIC, will be set.

Thus, users can use the software by running the following command: java -jar $TRIMMOMATIC {other options}.

TurboVNC

TurboVNC is an implementation of VNC optimized for 3D graphics rendering. Like other VNC software, TurboVNC can be used to create a virtual desktop on a remote machine, which can be useful for visualizing CPU-intensive graphics produced remotely.

Availability and Restrictions

Versions

The versions currently available at OSC are:

VERSION	OAKLEY	Ruby	Owens	notes
1.0
1.1	X ^[1]	X		Default version on Oakley prior to 09/15/2015
1.2	X ^[2]
2.0	X	X
2.0.91	X	X	X
2.1.1	X*	X*	X*	Must load intel compiler, version 16.0.3 for Owens, 15.0.3 for Oakley and Ruby.

* Current default version

NOTE:

[1] -- TurboVNC 1.1's version of vncviewer does not work on Oakley. Use the 1.2 module's version.
[2] -- TurboVNC 1.2's version of vncserver does not work on Oakley. Use the 1.1 module's version.
To simplify vncviewer and vncserver incompatibility with prior 1.X versions, a new version of TurboVNC 2.0 is available as of 10/30/2015 on Oakley and Ruby clusters.
A version of TurboVNC 2.0.91 was installed in September 2016 to be uniformly available on all OSC clusters.

You can use module spider turbovnc to view available modules for a given cluster. Feel free to contact OSC Help if you need other versions for your work.

Access

TurboVNC is available for use by all OSC users.

Usage

Usage on Oakley

Setup on Oakley

To load the default version of TurboVNC module, use module load turbovnc. To select a particular software version, use module load turbovnc/version. For example, use module load turbovnc/2.0 to load TurboVNC version 2.0 on Oakley.

Please do not SSH directly to compute nodes and start VNC sessions! This will negatively impact other users (even if you have been assigned a node via the batch scheduler), and we will consider repeated occurances an abuse of the resources. If you need to use VNC on a compute node, please see our HOWTO for instructions.

Using TurboVNC

To start a VNC server on your current host, use the following command:

vncserver

After starting the VNC server you should see output similar to the following:

New 'X' desktop is hostname:display
Starting applications specified in /nfs/nn/yourusername/.vnc/xstartup.turbovnc
Log file is /nfs/nn/yourusername/.vnc/hotsname:display.log

Make a note of the hostname and display number ("hostname:display"), because you will need this information later in order to connect to the running VNC server.

To establish a standard unencrypted connection to an already running VNC server, X11 forwarding must first be enabled in your SSH connection. This can usually either be done by changing the preferences or settings in your SSH client software application, or by using the -X or -Y option on your ssh command.

Once you are certain that X11 forwarding is enabled, create your VNC desktop using the vncviewer command in a new shell.

vncviewer

You will be prompted by a dialogue box asking for the VNC server you wish to connect to. Enter "hostname:display".

You may then be prompted for your HPC password. Once the password has been entered your VNC desktop should appear, where you should see all of your home directory contents.

When you are finished with your work on the VNC desktop, you should make sure to close the desktop and kill the VNC server that was originally started. The VNC server can be killed using the following command in the shell where the VNC server was originally started:

vncserver -kill :[display]

For a full explanation of each of the previous commands, type man vncserver or man vncviewer at the command line to view the online manual.

Usage on Ruby

Setup on Ruby

To load the default version of TurboVNC module, use module load turbovnc. To select a particular software version, use module load turbovnc/version. For example, use module load turbovnc/2.0 to load TurboVNC version 2.0 on Oakley.

Please do not SSH directly to compute nodes and start VNC sessions! This will negatively impact other users (even if you have been assigned a node via the batch scheduler), and we will consider repeated occurances an abuse of the resources. If you need to use VNC on a compute node, please see our HOWTO for instructions.

Using TurboVNC

To start a VNC server on your current host, use the following command:

vncserver

After starting the VNC server you should see output similar to the following:

New 'X' desktop is hostname:display
Starting applications specified in /nfs/nn/yourusername/.vnc/xstartup.turbovnc
Log file is /nfs/nn/yourusername/.vnc/hotsname:display.log

Make a note of the hostname and display number ("hostname:display"), because you will need this information later in order to connect to the running VNC server.

To establish a standard unencrypted connection to an already running VNC server, X11 forwarding must first be enabled in your SSH connection. This can usually either be done by changing the preferences or settings in your SSH client software application, or by using the -X or -Y option on your ssh command.

Once you are certain that X11 forwarding is enabled, create your VNC desktop using the vncviewer command in a new shell.

vncviewer

You will be prompted by a dialogue box asking for the VNC server you wish to connect to. Enter "hostname:display".

You may then be prompted for your HPC password. Once the password has been entered your VNC desktop should appear, where you should see all of your home directory contents.

When you are finished with your work on the VNC desktop, you should make sure to close the desktop and kill the VNC server that was originally started. The VNC server can be killed using the following command in the shell where the VNC server was originally started:

vncserver -kill :[display]

For a full explanation of each of the previous commands, type man vncserver or man vncviewer at the command line to view the online manual.

Usage on Owens

Setup on Owens

To load the default version of TurboVNC module, use module load turbovnc. To select a particular software version, use module load turbovnc/version. For example, use module load turbovnc/2.0 to load TurboVNC version 2.0 on Oakley.

Please do not SSH directly to compute nodes and start VNC sessions! This will negatively impact other users (even if you have been assigned a node via the batch scheduler), and we will consider repeated occurances an abuse of the resources. If you need to use VNC on a compute node, please see our HOWTO for instructions.

Using TurboVNC

To start a VNC server on your current host, use the following command:

vncserver

After starting the VNC server you should see output similar to the following:

New 'X' desktop is hostname:display
Starting applications specified in /nfs/nn/yourusername/.vnc/xstartup.turbovnc
Log file is /nfs/nn/yourusername/.vnc/hotsname:display.log

Make a note of the hostname and display number ("hostname:display"), because you will need this information later in order to connect to the running VNC server.

To establish a standard unencrypted connection to an already running VNC server, X11 forwarding must first be enabled in your SSH connection. This can usually either be done by changing the preferences or settings in your SSH client software application, or by using the -X or -Y option on your ssh command.

Once you are certain that X11 forwarding is enabled, create your VNC desktop using the vncviewer command in a new shell.

vncviewer

You will be prompted by a dialogue box asking for the VNC server you wish to connect to. Enter "hostname:display".

You may then be prompted for your HPC password. Once the password has been entered your VNC desktop should appear, where you should see all of your home directory contents.

When you are finished with your work on the VNC desktop, you should make sure to close the desktop and kill the VNC server that was originally started. The VNC server can be killed using the following command in the shell where the VNC server was originally started:

vncserver -kill :[display]

For a full explanation of each of the previous commands, type man vncserver or man vncviewer at the command line to view the online manual.

Turbomole

TURBOMOLE is an ab initio computational chemistry program that implements various quantum chemistry algorithms. It is developed at the group of Prof. Reinhart Ahlrichs at the University of Karlsruhe.

Availability and Restrictions

Versions

The versions currently available at OSC are:

VERSION		OAKLEY	Owens	notes
6.3.1	serial	X		Default version on Oakley prior to 09/15/2015
6.3.1	parallel	X		Default version on Oakley prior to 09/15/2015
6.4	serial	X
6.4	parallel	X
6.5	serial	X
6.5	parallel	X
7.0.1	serial	X*
7.0.1	parallel	X*
7.1	serial		X*
	parallel		X*

* Current default version

You can use module spider turbomole to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access for Academic Users

Use of Turbomole for academic purposes requires validation. In order to obtain validation, please contact OSC Help for further instruction.

Usage

Usage on Oakley

Set-up on Oakley

To load the default version of Turbomole module on Oakley, use module load turbomole for both serial and parallel programs. To select a particular software version, use module load turbomole/version. For example, use module load turbomole/6.4 to load Turbomole version 6.4 for both serial and parallel programs on Oakley.

Using Turbomole on Oakley

To execute a turbomole program:

module load turbomole
<turbomole command>

Batch Usage on Oakley

When you log into oakley.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=12 -l walltime=00:20:00

which requests one core (-l nodes=1:ppn=12), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

Usage on Owens

Set-up on Owens

To load the default version of Turbomole module on Owens, use module load turbomole for both serial and parallel programs. To select a particular software version, use module load turbomole/version. For example, use module load turbomole/7.1 to load Turbomole version 7.1 for both serial and parallel programs on Owens.

Using Turbomole on Owens

To execute a turbomole program:

module load turbomole
<turbomole command>

Batch Usage on Owens

When you log into owens.osc.edu you are actually logged into a linux box referred to as the login node. To gain access to the mutiple processors in the computing environment, you must submit your job to the batch system for execution. Batch jobs can request mutiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations and Batch Limit Rules for more info. Batch jobs run on the compute nodes of the system and not on the login node. It is desirable for big problems since more resources can be used.

Interactive Batch Session

For an interactive batch session one can run the following command:

qsub -I -l nodes=1:ppn=28 -l walltime=00:20:00

which requests 28 cores (-l nodes=1:ppn=28), for a walltime of 20 minutes (-l walltime=00:20:00). You may adjust the numbers per your need.

VASP

The Vienna Ab initio Simulation Package, VASP, is a program package for ab-initio quantum-mechanical molecular dynamics (MD) simulations and electronic structure calculations, from first principles.

Availability and Restrictions

Access

Due to licensing considerations, OSC does not provide general access to this software.

However, we are available to assist with the configuration of individual research-group installations on all our systems. See the The VASP FAQ page or contact OSC Help for information regrading licensing.

Usage

Using VASP

See the VASP documentation page for tutorial materials.

VCFtools

VCFtools is a program package designed for working with VCF files, such as those generated by the 1000 Genomes Project. The aim of VCFtools is to provide easily accessible methods for working with complex genetic variation data in the form of VCF files.

Availability and Restrictions

The following versions of VCFtools are available on OSC clusters:

Version	Oakley	Owens
0.1.14	X*	X*

* Current default version

You can use module spider vcftools to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

VCFtools is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of VCFtools, run the following command: module load vcftools. The default version will be loaded. To select a particular VCFtools version, use module load vcftools/version . For example, use module load vcftools/0.1.14 to load VCFtools 0.1.14.

Usage on Owens

Set-up

To configure your environment for use of VCFtools, run the following command: module load vcftools. The default version will be loaded. To select a particular VCFtools version, use module load vcftools/version . For example, use module load vcftools/0.1.14 to load VCFtools 0.1.14.

VMD

VMD is a visulaization program for the display and analysis of molecular systems.

Availability and Restrictions

Versions

The following versions of VMD are available on OSC clusters:

VERSION	OWENS
1.9.4	X*

* Current default version

Access

VMD is for academic purpose only. Please review the license agreement before you use this software.

Usage

Usage on Owens

Using VMD with OSC OnDemand

It is recommended to use VMD with OSC OnDemand. On the OnDemand page launch the VMD GUI from the interactive apps dropdown menu. This will open the VMD Main, OpenGL Display, and terminal windows. End a session through the VMD Main window by selecting File → Quit.

See VMD Tutorials for basic VMD usage instructions.

VarScan

VarScan is a platform-independent software tool developed at the Genome Institute at Washington University to detect variants in NGS data.

Availability and Restrictions

Versions

The following versions of VarScan are available on OSC clusters:

Version	Oakley	Owens
2.4.1	X*	X*

* Current default version

You can use module spider varscan to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

VarScan is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of VarScan, run the following command: module load varscan. The default version will be loaded. To select a particular VarScan version, use module load varscan/version. For example, use module load varscan/2.4.1 to load VarScan 2.4.1.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load varscan, a new environment variable, VARSCAN, will be set.

Thus, users can use the software by running the following command: java -jar $VARSCAN {other options}.

Usage on Owens

Set-up

To configure your environment for use of VarScan, run the following command: module load varscan. The default version will be loaded. To select a particular VarScan version, use module load varscan/version. For example, use module load varscan/2.4.1 to load VarScan 2.4.1.

Usage

This software is a Java executable .jar file; thus, it is not possible to add to the PATH environment variable.

From module load varscan, a new environment variable, VARSCAN, will be set.

Thus, users can use the software by running the following command: java -jar $VARSCAN {other options}.

VirtualGL

VirtualGL allows OpenGL applications to run with 3D hardware accerlation.

Availability & Restrictions

Versions

The following versions of VirtualGL are available on OSC clusters:

VERSION	OAKLEY	RUBY	OWENS	NOTes
2.3	X*	X		Intel 15.0.3 must be loaded on Ruby
2.5			X*	Intel 16.0.3 must be loaded Owens
2.5.2		X*

* Current default version

Access

OSC provides VirtualGL to all OSC users.

Usage

Usage on Oakley

Set-up

Configure your environment for use of VirtualGL with module load virtualgl. This will load the default version.

Usage on Ruby

Set-up

Configure your environment for use of VirtualGL with module load virtualgl. This will load the default version. To load a different verison, use module load virtualgl/version. For example, load VirtualGL 2.3 with module laod virtualgl/2.3. Note that VirtualGL 2.3 requires the intel compiler, version 15.0.3. Before loading VirtualGL 2.3, load intel 15.0.3 with module load intel/15.0.3.

Usage on Owens

Set-up

To configure your environment for use of VirtualGL the intel compiler, verison 16.0.3, must be loaded. Load intel 16.0.3 with module load intel/16.0.3. Then use module load virtualgl to load VirtualGL. The default version will be loaded.

WARP3D

From WARP3D's webpage:

WARP3D is under continuing development as a research code for the solution of large-scale, 3-Dsolid models subjected to static and dynamic loads. The capabilities of the code focus on 
fatigue & fracture analyses primarily in metals. WARP3D runs on laptops-to-supercomputers and can analyze models with several million nodes and elements.

Availability and Restrictions

Versions

The following versions of WARP3D are available on OSC clusters:

version

Owens

17.7.1

X

17.7.4

X

17.8.0

17.8.7

X

* Default version depends on the compiler and MPI version loaded

You can use module spider warp3d to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

WARP3D is available to all OSC users without restriction.

Usage

Usage on Owens

Setup on Owens

To configure the Owens cluster for the use of WARP3D, use the following commands:

module load intel
module load intelmpi
module load warp3d

Batch Usage on Owens

Batch jobs can request multiple nodes/cores and compute time up to the limits of the OSC systems. Refer to Queues and Reservations for Oakley, Queues and Reservations for Ruby, and Scheduling Policies and Limits for more info.

Running WARP3D

Below is an example batch script (job.txt) for using WARP3D:

#PBS -N WARP3D
#PBS -l nodes=1:ppn=28
#PBS -l walltime=30:00
#PBS -j oe 
#PBS -S /bin/bash 
# Load the modules for WARP3D
module load intel
module load intelmpi
module load warp3d
# Copy files to $TMPDIR and move there to execute the program
cp $WARP3D_HOME/example_problems_for_READMEs/mt_cohes_*.inp $TMPDIR
cd $TMPDIR
# Run the solver using 4 MPI tasks and 6 threads per MPI task 
$WARP3D_HOME/warp3d_script_linux_hybrid 4 6 < mt_cohes_4_cpu.inp
# Finally, copy files back to your home directory 
cp -r * $PBS_O_WORKDIR

In order to run it via the batch system, submit the job.txt file with the following command:

qsub job.txt

bam2fastq

bam2fastq is used to extract raw sequences (with qualities) from programs like SAMtools, Picard, and Bamtools.

Availability and Restrictions

Versions

The following versions of bam2fastq are available on OSC clusters:

Version	Oakley	owens
1.1.0	X*	X*

* Current default version

You can use module spider bam2fastq to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

bam2fastq is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of bam2fastq, run the following command: module load bam2fastq. The default version will be loaded. To select a particular bam2fastq version, use module load bam2fastq/version. For example, use module load bam2fastq/1.1.0 to load bam2fastq 1.1.0.

Usage on Owens

Set-up

To configure your environment for use of bam2fastq, run the following command: module load bam2fastq. The default version will be loaded. To select a particular bam2fastq version, use module load bam2fastq/version. For example, use module load bam2fastq/1.1.0 to load bam2fastq 1.1.0.

bcftools

bcftools is a set of utilities that manipulate variant calls in the Variant Call Format (VCF) and its binary counterpart BCF.

Availability and Restrictions

Versions

The following versions of bcftools are available on OSC clusters:

Version	Oakley	OWEns
1.3.1	X*	X*

* Current default version

You can use module spider bcftools to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

bcftools is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of bcftools, run the following command: module load bcftools. The default version will be loaded. To select a particular bcftools version, use module load bcftools/version. For example, use module load bcftools/1.3.1 to load bcftools 1.3.1.

Usage on Owens

Set-up

To configure your environment for use of bcftools, run the following command: module load bcftools. The default version will be loaded. To select a particular bcftools version, use module load bcftools/version. For example, use module load bcftools/1.3.1 to load bcftools 1.3.1.

bedtools

Collectively, the bedtools utilities are a swiss-army knife of tools for a wide-range of genomics analysis tasks. The most widely-used tools enable genome arithmetic: that is, set theory on the genome. While each individual tool is designed to do a relatively simple task, quite sophisticated analyses can be conducted by combining multiple bedtools operations on the UNIX command line.

Availability and Restrictions

Versions

The following versions of bedtools are available on OSC clusters:

Version	Oakley	Owens
2.25.0	X*	X*

* Current default version

You can use module spider bedtools to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

bedtools is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of bedtools, run the following command: module load bedtools. The default version will be loaded. To select a particular bedtools version, use module load bedtools/version. For example, use module load bedtools/2.25.0 to load bedtools 2.25.0.

Usage on Owens

Set-up

To configure your environment for use of bedtools, run the following command: module load bedtools. The default version will be loaded. To select a particular bedtools version, use module load bedtools/version. For example, use module load bedtools/2.25.0 to load bedtools 2.25.0.

eXpress

eXpress is a streaming tool for quantifying the abundances of a set of target sequences from sampled subsequences.

Availability and Restrictions

Versions

The following versions of eXpress are available on OSC clusters:

Version	Oakley	Owens
1.5.1	X*	X*

* Current default version

You can use module spider express to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

eXpress is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of eXpress, run the following command: module load express. The default version will be loaded. To select a particular eXpress version, use module load express/version. For example, use module load express/1.5.1 to load eXpress 1.5.1.

Usage on Owens

Set-up

To configure your environment for use of eXpress, run the following command: module load express. The default version will be loaded. To select a particular eXpress version, use module load express/version. For example, use module load express/1.5.1 to load eXpress 1.5.1.

miRDeep2

miRDeep2 is a completely overhauled tool which discovers microRNA genes by analyzing sequenced RNAs. The tool reports known and hundreds of novel microRNAs with high accuracy in seven species representing the major animal clades. The low consumption of time and memory combined with user-friendly interactive graphic output makes miRDeep2 accessible for straightforward application in current reasearch.

Availability and Restrictions

Versions

The following versions of miRDeep2 are available on OSC clusters:

Version	Oakley	Owens
2.0.0.8	X*	X*

* Current default version

You can use module spider mirdeep2 to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

miRDeep2 is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up

To configure your environment for use of miRDeep2, run the following command: module load mirdeep2. The default version will be loaded. To select a particular miRDeep2 version, use module load mirdeep2/version. For example, use module load mirdeep2/2.0.0.8 to load miRDeep2 2.0.0.8.

Usage on Owens

Set-up

To configure your environment for use of miRDeep2, run the following command: module load mirdeep2. The default version will be loaded. To select a particular miRDeep2 version, use module load mirdeep2/version. For example, use module load mirdeep2/2.0.0.8 to load miRDeep2 2.0.0.8.

mpiBLAST

mpiBLAST is a freely available, open-source, parallel implementation of NCBI BLAST. mpiBLAST takes advantage of distributed computational resources, i.e., a cluster, through explicit MPI communication and thereby utilizes all available resources unlike standard NCBI BLAST which can only take advantage of shared-memory multi-processors (SMPs).

Availability and Restrictions

Versions

The following version of mpiBLAST are available on OSC systems:

Version	Oakley	OWENS
1.6.0	X*	X*

* Current default version

You can use module spider mpiblast to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

mpiBLAST is available without restriction to all OSC users.

Usage

Set-up

To load the mpiBLAST software, use the following command:

module load mpiblast

Using mpiBLAST

Once mpiblast module is loaded, the commands are available for your use.

mpiblast 
mpiblast_cleanup 
mpiformatdb

Formatting a database

Before processing blast queries the sequence database must be formatted with mpiformatdb. The command line syntax looks like this: mpiformatdb -N 16 -i nt -o T

The above command would format the nt database into 16 fragments. Note that currently mpiformatdb does not support multiple input files.

mpiformatdb places the formatted database fragments in the same directory as the FASTA database. To specify a different target location, use the "-n" option as what is available in the NCBI formatdb.

Querying the database

mpiblast command line syntax is nearly identical to NCBI's blastall program. Running a query on 18 nodes would look like: mpiexec -n 18 mpiblast -p blastn -d nt -i blast_query.fas -o blast_results.txt

The above command would query the sequences in blast_query.fas against the nt database and write out results to the blast_results.txt file in the current working directory. By default, mpiBLAST reads configuration information from ~/.ncbirc. Furthermore, mpiBLAST needs at least 3 processes to perform a search: two processes dedicated for scheduling tasks and coordinating file output, while any additional processes actually perform search tasks.

Extra options to mpiblast

--partition-size=[integer]
Enable hierarchical scheduling with multiple masters. The partition size equals the number of workers in a partition plus 1 (the master process). For example, a partition size of 17 creates partitions consisting of 16 workers and 1 master. An individual output file will be generated for each partition. By default, mpiBLAST uses one partition. This option is only available for version 1.6 or above.
--replica-group-size=[integer]
Specify how database fragments are replicated within a partition. Suppose the total number of database fragments is F, the number of MPI processes in a partition is N, and the replica-group-size is G, then in total (N-1)/G database replicas will be distributed in the partition (the master process does not host any database fragments), and each worker process will host F/G fragments. In other words, a database replica will be distributed to every G MPI processes.
--query-segment-size=[integer]
The default value is 5. Specify the number of query sequences that will be fetched from the supermaster to the master at a time. This parameter controls the granularity of load balancing between different partitions. This option is only available for version 1.6 or above.
--use-parallel-write
Enable the high-performance parallel output solution. Note the current implementation of parallel-write does not require a parallel file system.
--use-virtual-frags
Enable workers to cache database fragments in memory instead of local storage. This is recommended on diskless platforms where there is no local storage attaching to each processor. Default to be enabled on Blue Gene systems.
--predistribute-db
Distribute database fragments to workers before the search begins. Especially useful in reducing data input time when multiple database replicas need to be distributed to workers.
--output-search-stats
Enable output of the search statistics in the pairwise and XML output format. This could cause performance degradation on some diskless systems such as Blue Gene.
--removedb
Removes the local copy of the database from each node before terminating execution.
--copy-via=[cp|rcp|scp|mpi|none]
Sets the method of copying files that each worker will use. Default = "cp"
- cp : use standard file system "cp" command. Additional option is --concurrent.
- rcp : use rsh "rcp" command. Additonal option is --concurrent.
- scp : use ssh "scp" command. Additional option is --concurrent.
- mpi : use MPI_Send/MPI_Recv to copy files. Additional option is --mpi-size.
- none : do not copy files, instead use shared storage as local storage.
--debug[=filename]
Produces verbose debugging output for each node, optionally logs the output to a file.
--time-profile=[filename]
Reports execution time profile.
--version
Print the mpiBLAST version.

Please refer to the README file in the mpiBLAST package for performance tuning guide.

Removing a database

The --removedb command line option will cause mpiBLAST to do all work in a temporary directory that will get removed from each node's local storage directory upon successful termination. For example:
mpiexec -n 18 mpiblast -p blastx -d yeast.aa -i ech_10k.fas -o results.txt --removedb

The above command would perform a 18 node (16 worker) search of the yeast.aa database, writing the output to results.txt. Upon completion, worker nodes would delete the yeast.aa database fragments from their local storage.

Databases can also be removed without performing a search in the following manner:
mpiexec -n 18 mpiblast_cleanup

Batch Usage

Below is a sample batch script for running mpiBLAST job. It asks for 24 processors and 4 hours of walltime.

#PBS -l walltime=4:00:00
#PBS -l nodes=2:ppn=12
#PBS -N mpiBLAST
#PBS -j oe

# Set up the variable MPIBLAST_SHARED. This is where mpiblast looks for database.
# Leave the variable MPIBLAST_LOCAL unset. It will be default to $TMPDIR/mpiBLAST_local_db/.
export MPIBLAST_SHARED=/nfs/gpfs/PZS0530/yzhang/db/mpiblast

# Load the software environment for mpiblast.
module load mpiblast

# Copy input file over to $TMPDIR on the compute nodes.
cd $PBS_O_WORKDIR
pbsdcp query.fasta $TMPDIR

# Break the database into 20 pieces 
# Note: you need to work in the directory $TMPDIR for this step. 
# This step fails if you try it in your home directory or lustre file system.
cd $TMPDIR
cp $MPIBLAST_SHARED/nr ./
/usr/bin/time mpiformatdb -N 20 -i nr -o T -p T
cp nr.* $MPIBLAST_SHARED/

# Run mpiblast. 
cp /usr/local/mpiblast/1.6.0/.ncbirc ./
/usr/bin/time mpiexec -n 22 mpiblast -p blastp -d nr -i query.fasta -o blast_results.txt

# Copy output file back to the working directory. 
# The result only exists on the head node. So we are using "cp" command here.
mkdir $PBS_O_WORKDIR/$PBS_JOBID
cp blast_results.txt $PBS_O_WORKDIR/$PBS_JOBID

NOTE: ppn should = 28 for Owens.

parallel-command-processor

There are many instances where it is necessary to run the same serial program many times with slightly different input. Parametric runs such as these either end up running in a sequential fashion in a single batch job, or a batch job is submitted for each parameter that is varied (or somewhere in between.) One alternative to this is to allocate a number of nodes/processors to running a large number of serial processes for some period of time. The command parallel-command-processor allows the execution of large number of independent serial processes in parallel. parallel-command-processor works as follows: In a parallel job with N processors allocated, the PCP manager process will read the first N-1 commands in the command stream and distribute them to the other N-1 processors. As processes complete, the PCP manager will read the next one in the stream and send it out to an idle processor core. Once the PCP manager runs out of commands to run, it will wait on the remaining running processes to complete before shutting itself down.

Availability and Restrictions

Parallel-Command-Processor is available for Oakley, Ruby, and Owens for all users.

Usage

Here is an interactive batch session that demonstrates the use of parallel-command-processor with a config file, pconf. pconf contains several lines of simple commands, one per line. The output of the commands were redirected to individual files.

-bash-3.2$ qsub -I -l nodes=2:ppn=4 
-bash-3.2$ cd $PBS_O_WORKDIR 
-bash-3.2$ cp pconf $TMPDIR
-bash-3.2$ cd $TMPDIR
-bash-3.2$ cat pconf
ls / > 1 
ls $TMPDIR > 2 
ls $HOME > 3 
ls /usr/local/ > 4 
ls /tmp > 5 
ls /usr/src > 6 
ls /usr/local/src > 7
ls /usr/local/etc > 8 
hostname > 9 
uname -a > 10 
df > 11
-bash-3.2$ module load pcp
-bash-3.2$ mpiexec parallel-command-processor pconf
-bash-3.2$ pwd
/tmp/pbstmp.1371894 
-bash-3.2$ mpiexec -ppn=1 ls -l $TMPDIR 
854 total 16 
-rw------- 1 yzhang G-3040 1082 Feb 18 16:26 11
-rw------- 1 yzhang G-3040 1770 Feb 18 16:26 4 
-rw------- 1 yzhang G-3040 67 Feb 18 16:26 5
-rw------- 1 yzhang G-3040 32 Feb 18 16:26 6 
-rw------- 1 yzhang G-3040 0 Feb 18 16:26 7 
855 total 28
-rw------- 1 yzhang G-3040 199 Feb 18 16:26 1
-rw------- 1 yzhang G-3040 111 Feb 18 16:26 10
-rw------- 1 yzhang G-3040 12 Feb 18 16:26 2
-rw------- 1 yzhang G-3040 87 Feb 18 16:26 3 
-rw------- 1 yzhang G-3040 38 Feb 18 16:26 8
-rw------- 1 yzhang G-3040 20 Feb 18 16:26 9
-rw------- 1 yzhang G-3040 163 Feb 18 16:25 pconf 
-bash-3.2$ exit

As the command "mpiexec -ppn=1 ls -l $TMPDIR" shows, the output files are distributed on the two nodes. In a batch file, pbsdcp can be used to distribute-copy the files to $TMPDIR on all nodes of the job and gather output files once execution has completed. This step is important due to the load that executing many processes in parallel can place on the user home directories.

Here is a slightly more complex example showing the usage of parallel-command-processor and pbsdcp:

#PBS -l nodes=13:ppn=4 
#PBS -l walltime=1:00:00 
#PBS -S /bin/bash 
#PBS -N blast-PCP 
#PBS -j oe 
date

module load biosoftw 
module load blast

set -x

cd $PBS_O_WORKDIR 
pbsdcp query/query.fsa.* $TMPDIR 
pbsdcp db/rice.* $TMPDIR 
cd $TMPDIR

for i in $(seq 1 49)

do 
      cmd="blastall -p blastn -d rice -i query.fsa.$i -o out.$i" 
      echo ${cmd} >> runblast 
done

module load pcp
mpiexec parallel-command-processor runblast

mkdir $PBS_O_WORKDIR/output 
pbsdcp -g out.* $PBS_O_WORKDIR/output

date

pkg-config

pkg-config is a helper tool used when compiling applications and libraries. It helps you insert the correct compiler options on the command line. This metadata is stored in pkg-config files. These files have the suffix .pc and reside in specific locations known to the pkg-config tool.

Availability and Restrictions

Versions

The following versions of pkg-config are available on OSC clusters:

VERSION	OAKLEY
0.28	X*

* Current default version

You can use module spider pkg-config to view available modules for a given machine. Feel free to contact OSC Help if you need other versions for your work.

Access

pkg-config is available to all OSC users without restriction.

Usage

Usage on Oakley

Set-up on Oakley

To configure your environment for use of pkg-config on Oakley, run the following command: module load pkg-config. The default version will be loaded. To select a particular pkg-config version, use module load pkg-config/version. For example, use module load pkg-config/0.28 to load pkg-config 0.28 on Oakley.

software	new default	notes
boost	1.63.0	Intel and gnu compilers only, no mpi
cairo	1.14.2
fftw3	3.3.5
hdf5	1.8.17	serial & parallel
metis	5.1.0
mkl	11.3.3	Interfaces not built. Contact oschelp@osc.edu if you need them.
netcdf	4.3.3.1	serial & parallel, with Fortran & C++ interfaces
parmetis	4.0.3
scalapack	2.0.2
suitesparse	4.5.3

Software	Old default	New default
MPP_DYNA	971_d_R7.1.1	9.0.1
NAMD	2.11	2.12
OPENFORM	2.3.0	3.0.0
WARP3D	17.5.3	17.7.4
CMAKE	2.8.10.2	3.7.2
PARAVIEW		4.4.0
JAVA	1.7.0_55	1.8.0_60
BLAST	2.2.26	2.6.0+
TURBOMOLE	6.5	7.0.1
QCHEM	4.3	4.4.1
SCHRODINGER	14	15
ABAQUS	6.14	2016
FLUENT	15.0.7	16.0
LS-DYNA	7.1.1	9.0.1
COMSOL	51	52
CUDA	6.5.14	7.5.18 (oakley) 8.0.44(ruby)
STARCCM	10.04.009	11.06.011
TURBOVNC	1.2	2.0.91
MATLAB	r2014b	r2016b
GAUSSIAN	g09d01	g09e01

Software		Current default	New default	To be removed	Notes
ABAQUS		6.11-2	6.14	6.8-4, 6.8-4-test, 6.11-1, 6.11-1-test
AMBER		11	14
ANSYS		14.5.7	16.0	13, 15.0	ansyspar licenses moving to Oakley
COMSOL		43a-4.3.1.161	51	42a, 50
CUDA		5.0.35	6.5.14	4.1.28, 4.2.9
FLUENT		15.0.7	15.0.7	13, 13-ndem, 13-test	no change to default; ansyspar licenses moving to Oakley
Gaussian		g09c01	g09d01
GROMACS		4.5.5	4.6.3
LAMMPS		12Feb12	5Sep14
LS-DYNA	LS-DYNA (smp solvers)	971_d_5.1.1	971_d_7.1.1	R5.0
LS-DYNA	MPP-DYNA (mpp solvers)	971_s_R5.1.1_ndem	971_d_7.1.1	R4.2.1, R5.0	Dependent on IntelMPI
MATLAB		r2013a	r2014b	R2011b, R2012a, R2012b
Python		2.7.1	3.4.2	3.4.1	Module only (misnamed)
Q-Chem		4.0.1	4.3
STAR-CCM		7.06.012	10.4.009	7.04
TotalView		8.9.2-1	8.14.1-8
Turbomole		6.3.1	6.5
TurboVNC		1.1	1.2

Available Software

Available software by OSC System and Application

Access Forms

Statewide Software

Software Refresh

License Server Status

Browse Software

License Server Status

Current Issues

Previous Issues

Scientific Database List

Cambridge Structural Database

Availability and Restrictions

Usage

Further Reading

Software Forms

Software List

ABAQUS

Availability and Restrictions

Versions

Access for Academic Users

Access for Commerical Users

Usage

Token Usage

Usage on Ruby

Set-up on Ruby

Using ABAQUS

Batch Usage on Ruby

Interactive Batch Session

Non-interactive Batch Job (Serial Run)

Non-interactive Batch Job (Parallel Run)

Usage on Owens

Set-up on Owens

Using ABAQUS

Batch Usage on Owens

Interactive Batch Session

Non-interactive Batch Job (Serial Run)

Non-interactive Batch Job (Parallel Run)

Further Reading

ABINIT

Availability and Restrictions

Versions

Access

Usage on Oakley

Set-up

Further Reading

AMBER

Availability and Restrictions

Versions

Access for Academic Users

Access for Commerical Users

Usage

Usage on Oakley

Set-up

Using AMBER

Batch Usage

Interactive Batch Session

Non-interactive Batch Job (Serial Run)

Usage on Ruby

Set-up

Using AMBER

Batch Usage

Interactive Batch Session

Non-interactive Batch Job (Serial Run)

Usage on Owens

Set-up

Using AMBER

Batch Usage

Interactive Batch Session

Non-interactive Batch Job (Serial Run)

Troubleshooting

Further Reading

ANSYS

Availability and Restrictions

Versions

Access for Academic Users

Access for Commerical Users

Usage

Note

Further Reading