The following are technical specifications for Pitzer.  

This document is obsoleted and kept as a reference to previous Pitzer programming environment. Please refer to here for the latest version.

Compilers

C, C++ and Fortran are supported on the Pitzer cluster. Intel, PGI and GNU compiler suites are available. The Intel development tool chain is loaded by default. Compiler commands and recommended options for serial programs are listed in the table below. See also our compilation guide.

The Skylake processors that make up Pitzer support the Advanced Vector Extensions (AVX512) instruction set, but you must set the correct compiler flags to take advantage of it. AVX512 has the potential to speed up your code by a factor of 8 or more, depending on the compiler and options you would otherwise use. However, bare in mind that clock speeds decrease as the level of the instruction set increases. So, if your code does not benefit from vectorization it may be beneficial to use a lower instruction set.

In our experience, the Intel and PGI compilers do a much better job than the GNU compilers at optimizing HPC code.

With the Intel compilers, use -xHost and -O2 or higher. With the GNU compilers, use -march=native and -O3. The PGI compilers by default use the highest available instruction set, so no additional flags are necessary.

This advice assumes that you are building and running your code on Owens. The executables will not be portable.  Of course, any highly optimized builds, such as those employing the options above, should be thoroughly validated for correctness.

Parallel Programming

MPI

OSC systems use the MVAPICH2 implementation of the Message Passing Interface (MPI), optimized for the high-speed Infiniband interconnect. MPI is a standard library for performing parallel processing using a distributed-memory model. For more information on building your MPI codes, please visit the MPI Library documentation.

Parallel programs are started with the mpiexec command. For example,

mpiexec ./myprog

The mpiexec command will normally spawn one MPI process per CPU core requested in a batch job. Use the -n and/or -ppn option to change that behavior.

The table below shows some commonly used options. Use mpiexec -help for more information.

OpenMP

The Intel, PGI and GNU compilers understand the OpenMP set of directives, which support multithreaded programming. For more information on building OpenMP codes on OSC systems, please visit the OpenMP documentation.

Process/Thread placement

Processes and threads are placed differently depending on the compiler and MPI implementation used to compile your code. This section summarizes the default behavior and how to modify placement.

For all three compilers (Intel, GNU, PGI), purely threaded codes do not bind to particular cores by default.

For MPI-only codes, Intel MPI first binds the first half of processes to one socket, and then second half to the second socket so that consecutive tasks are located near each other. MVAPICH2 first binds as many processes as possible on one socket, then allocates the remaining processes on the second socket so that consecutive tasks are near each other. OpenMPI alternately binds processes on socket 1, socket 2, socket 1, socket 2, etc, with no particular order for the core id.

For Hybrid codes, Intel MPI first binds the first half of processes to one socket, and then second half to the second socket so that consecutive tasks are located near each other. Each process is allocated ${OMP_NUM_THREADS} cores and the threads of each process are bound to those cores. MVAPICH2 allocates ${OMP_NUM_THREADS} cores for each process and each thread of a process is placed on a separate core. By default, OpenMPI  behaves the same for hybrid codes as it does for MPI-only codes, allocating a single core for each process and all threads of that process.

The following tables describe how to modify the default placements for each type of code.

OpenMP options:

MPI options:

Hybrid MPI+OpenMP options (combine with options from OpenMP table for thread affinity within cores allocated to each process):

The above tables list the most commonly used settings for process/thread placement. Some compilers and Intel libraries may have additional options for process and thread placement beyond those mentioned on this page. For more information on a specific compiler/library, check the more detailed documentation for that library.

GPU Programming

64 Nvidia V100 GPUs are available on Pitzer.  Please visit our GPU documentation.

Compilers

C, C++ and Fortran are supported on the Pitzer cluster. Intel, PGI and GNU compiler suites are available. The Intel development tool chain is loaded by default. Compiler commands and recommended options for serial programs are listed in the table below. See also our compilation guide.

The Skylake and Cascade Lake processors that make up Pitzer support the Advanced Vector Extensions (AVX512) instruction set, but you must set the correct compiler flags to take advantage of it. AVX512 has the potential to speed up your code by a factor of 8 or more, depending on the compiler and options you would otherwise use. However, bare in mind that clock speeds decrease as the level of the instruction set increases. So, if your code does not benefit from vectorization it may be beneficial to use a lower instruction set.

In our experience, the Intel compiler usually does the best job of optimizing numerical codes and we recommend that you give it a try if you’ve been using another compiler.

With the Intel compilers, use -xHost and -O2 or higher. With the GNU compilers, use -march=native and -O3. The PGI compilers by default use the highest available instruction set, so no additional flags are necessary.

This advice assumes that you are building and running your code on Owens. The executables will not be portable.  Of course, any highly optimized builds, such as those employing the options above, should be thoroughly validated for correctness.

Parallel Programming

MPI

OSC systems use the MVAPICH2 implementation of the Message Passing Interface (MPI), optimized for the high-speed Infiniband interconnect. MPI is a standard library for performing parallel processing using a distributed-memory model. For more information on building your MPI codes, please visit the MPI Library documentation.

MPI programs are started with the srun command. For example,

#!/bin/bash
#SBATCH --nodes=2

srun [ options ] mpi_prog

The srun command will normally spawn one MPI process per task requested in a Slurm batch job. Use the -n ntasks and/or --ntasks-per-node=n option to change that behavior. For example,

#!/bin/bash
#SBATCH --nodes=2

# Use the maximum number of CPUs of two nodes
srun ./mpi_prog

# Run 8 processes per node
srun -n 16 --ntasks-per-node=8  ./mpi_prog

The table below shows some commonly used options. Use srun -help for more information.

OpenMP

The Intel, GNU and PGI compilers understand the OpenMP set of directives, which support multithreaded programming. For more information on building OpenMP codes on OSC systems, please visit the OpenMP documentation.

An OpenMP program by default will use a number of threads equal to the number of CPUs requested in a Slurm batch job. To use a different number of threads, set the environment variable OMP_NUM_THREADS. For example,

#!/bin/bash
#SBATCH --ntasks=8

# Run 8 threads
./omp_prog

# Run 4 threads
export OMP_NUM_THREADS=4
./omp_prog

To run a OpenMP job on an exclusive node:

#!/bin/bash
#SBATCH --nodes=1
#SBATCH --exclusive

export OMP_NUM_THREADS=$SLURM_CPUS_ON_NODE
./omp_prog

Interactive job only

Please use -c, --cpus-per-task=X instead of -n, --ntasks=X to request an interactive job. Both result in an interactive job with X CPUs available but only the former option automatically assigns the correct number of threads to the OpenMP program. If  the option --ntasks is used only, the OpenMP program will use one thread or all threads will be bound to one CPU core. 

Hybrid (MPI + OpenMP)

An example of running a job for hybrid code:

#!/bin/bash
#SBATCH --nodes=2
#SBATCH --constraint=48core

# Run 4 MPI processes on each node and 12 OpenMP threads spawned from a MPI process
export OMP_NUM_THREADS=12
srun -n 8 -c 12 --ntasks-per-node=4 ./hybrid_prog

To run a job across either 40-core or 48-core nodes exclusively:

#!/bin/bash
#SBATCH --nodes=2

# Run 4 MPI processes on each node and the maximum available OpenMP threads spawned from a MPI process 
export OMP_NUM_THREADS=$(($SLURM_CPUS_ON_NODE/4))
srun -n 8 -c $OMP_NUM_THREADS --ntasks-per-node=4 ./hybrid_prog

Tuning Parallel Program Performance: Process/Thread Placement

To get the maximum performance, it is important to make sure that processes/threads are located as close as possible to their data, and as close as possible to each other if they need to work on the same piece of data, with given the arrangement of node, sockets, and cores, with different access to RAM and caches. 

While cache and memory contention between threads/processes are an issue, it is best to use scatter distribution for code. 

Processes and threads are placed differently depending on the computing resources you requste and the compiler and MPI implementation used to compile your code. For the former, see the above examples to learn how to run a job on exclusive nodes. For the latter, this section summarizes the default behavior and how to modify placement.

OpenMP only

For all three compilers (Intel, GNU, PGI), purely threaded codes do not bind to particular CPU cores by default. In other words, it is possible that multiple threads are bound to the same CPU core. 

The following table describes how to modify the default placements for pure threaded code:

1. Threads in the same socket might be bound to the same CPU core.
2. PGI LLVM-backend (version 19.1 and later) does not support thread/processors affinity on NUMA architecture. To enable this feature, compile threaded code with --Mnollvm to use proprietary backend.

MPI Only

For MPI-only codes, MVAPICH2 first binds as many processes as possible on one socket, then allocates the remaining processes on the second socket so that consecutive tasks are near each other.  Intel MPI and OpenMPI alternately bind processes on socket 1, socket 2, socket 1, socket 2 etc, as cyclic distribution.

For process distribution across nodes, all MPIs first bind as many processes as possible on one node, then allocates the remaining processes on the second node. 

The following table describe how to modify the default placements on a single node for MPI-only code with the command srun:

1. MV2_CPU_BINDING_POLICY will not work if MV2_ENABLE_AFFINITY=0 is set.

To distribute processes evenly across nodes, please set SLURM_DISTRIBUTION=cyclic.

Hybrid (MPI + OpenMP)

For Hybrid codes, each MPI process is allocated  OMP_NUM_THREADS cores and the threads of each process are bound to those cores. All MPI processes (as well as the threads bound to the process) behave as we describe in the previous sections. It means the threads spawned from a MPI process might be bound to the same core. To change the default process/thread placmements, please refer to the tables above. 

Summary

The above tables list the most commonly used settings for process/thread placement. Some compilers and Intel libraries may have additional options for process and thread placement beyond those mentioned on this page. For more information on a specific compiler/library, check the more detailed documentation for that library.

GPU Programming

164 Nvidia V100 GPUs are available on Pitzer.  Please visit our GPU documentation.

Reference

• Slurm: Support for Multi-core/Multi-thread Architecture
• TACC amask
• Thread Placement and Thread Affinity

Memory limit

A small portion of the total physical memory on each node is reserved for distributed processes.  The actual physical memory available to user jobs is tabulated below.

Summary

A job may request more than the max memory per core, but the job will be allocated more cores to satisfy the memory request instead of just more memory.
e.g. The following slurm directives will actually grant this job 3 cores, with 10 GB of memory
(since 2 cores * 4.5 GB = 9 GB doesn't satisfy the memory request).
#SBATCH --ntask=2
 #SBATCH --mem=10g

It is recommended to let the default memory apply unless more control over memory is needed.
Note that if an entire node is requested, then the job is automatically granted the entire node's main memory. On the other hand, if a partial node is requested, then memory is granted based on the default memory per core.

See a more detailed explanation below.

Regular Compute Node

• For the regular 'Skylake' processor-based node, it has 40 cores/node. The physical memory equates to 4.8 GB/core or 192 GB/node; while the usable memory equates to 4,556 MB/core or 182,240 MB/node (177.96 GB/node).
• For the regular 'Cascade Lake' processor-based node, it has 48 cores/node. The physical memory equates to 4.0 GB/core or 192 GB/node; while the usable memory equates to 3,797 MB/core or 182,256 MB/node (177.98 GB/node). 

Jobs requesting no more than 1 node

If your job requests less than a full node, it may be scheduled on a node with other running jobs. In this case, your job is entitled to a memory allocation proportional to the number of cores requested (4,556 MB/core or 3,797 MB/core depending on which type of node your job lands on).  For example, without any memory request ( --mem=XX ):

• A job that requests --ntasks=1 and lands on a 'Skylake' node will be assigned one core and should use no more than 4556 MB of RAM; a job that requests --ntasks=1 and lands on a 'Cascade Lake' node will be assigned one core and should use no more than 3797 MB of RAM
• A job that requests --ntasks=3 and lands on a 'Skylake' node will be assigned 3 cores and should use no more than 3*4556 MB of RAM; a job that requests --ntasks=3 and lands on a 'Cascade Lake' node will be assigned 3 cores and should use no more than 3*3797 MB of RAM
• A job that requests  --ntasks=40 and lands on a 'Skylake' node will be assigned the whole node (40 cores) with 178 GB of RAM; a job that requests --ntasks=40 and lands on a 'Cascade Lake' node will be assigned 40 cores (partial node) and should use no more than 40* 3797 MB of RAM
• A job that requests  --exclusive and lands on a 'Skylake' node will be assigned the whole node (40 cores) with 178 GB of RAM; a job that requests --exclusive and lands on a 'Cascade Lake' node will be assigned the whole node (48 cores) with 178 GB of RAM
• A job that requests  --exclusive --constraint=40core will land on a 'Skylake' node and will be assigned the whole node (40 cores) with 178 GB of RAM. 
  
  For example, with a memory request:
• A job that requests --ntasks=1 --mem=16000MB  and lands on 'Skylake' node will be assigned 4 cores and have access to 16000 MB of RAM, and charged for 4 cores worth of usage; a job that requests --ntasks=1 --mem=16000MB  and lands on 'Cascade Lake' node will be assigned 5 cores and have access to 16000 MB of RAM, and charged for 5 cores worth of usage
• A job that requests --ntasks=8 --mem=16000MB  and lands on 'Skylake' node will be assigned 8 cores but have access to only 16000 MB of RAM , and charged for 8 cores worth of usage; a job that requests --ntasks=8 --mem=16000MB  and lands on 'Cascade Lake' node will be assigned 8 cores but have access to only 16000 MB of RAM , and charged for 8 cores worth of usage

Jobs requesting more than 1 node

A multi-node job ( --nodes > 1 ) will be assigned the entire nodes and charged for the entire nodes regardless of --ntasks or --ntasks-per-node request. For example, a job that requests --nodes=10 --ntasks-per-node=1  and lands on 'Skylake' node will be charged for 10 whole nodes (40 cores/node*10 nodes, which is 400 cores worth of usage); a job that requests --nodes=10 --ntasks-per-node=1  and lands on 'Cascade Lake' node will be charged for 10 whole nodes (48 cores/node*10 nodes, which is 480 cores worth of usage). We usually suggest not including --ntasks-per-node and using --ntasks if needed.   

Large Memory Node

On Pitzer, it has 48 cores per node. The physical memory equates to 16.0 GB/core or 768 GB/node; while the usable memory equates to 15,872 MB/core or 761,856 MB/node (744 GB/node).

For any job that requests no less than 363 GB/node but less than 744 GB/node, the job will be scheduled on the large memory node.To request no more than a full large memory node, you need to specify the memory request between 363 GB and 744 GB, i.e.,  363GB <= mem <744GB. --mem is the total memory per node allocated to the job. You can request a partial large memory node, so consider your request more carefully when you plan to use a large memory node, and specify the memory based on what you will use. 

Huge Memory Node

On Pitzer, it has 80 cores per node. The physical memory equates to 37.5 GB/core or 3 TB/node; while the usable memory equates to 38,259 MB/core or  3,060,720 MB/node (2988.98 GB/node).

To request no more than a full huge memory node, you have two options:

• The first is to specify the memory request between 744 GB and 2988 GB, i.e., 744GB <= mem <=2988GB).
• The other option is to use the combination of --ntasks-per-node and --partition, like --ntasks-per-node=4 --partition=hugemem . When no memory is specified for the huge memory node, your job is entitled to a memory allocation proportional to the number of cores requested (38,259 MB/core). Note, --ntasks-per-node should be no less than 20 and no more than 80 

Summary

In summary, for serial jobs, we will allocate the resources considering both the # of cores and the memory request. For parallel jobs (nodes>1), we will allocate the entire nodes with the whole memory regardless of other requests. Check requesting resources on pitzer for information about the usable memory of different types of nodes on Pitzer. To manage and monitor your memory usage, please refer to Out-of-Memory (OOM) or Excessive Memory Usage.

GPU Jobs

Dual GPU Node

• For the dual GPU node with 'Skylake' processor, it has 40 cores/node. The physical memory equates to 9.6 GB/core or 384 GB/node; while the usable memory equates to 9292 MB/core or 363 GB/node. Each node has 2 NVIDIA Volta V100 w/ 16 GB GPU memory. 
• For the dual GPU node with 'Cascade Lake' processor, it has 48 cores/node. The physical memory equates to 8.0 GB/core or 384 GB/node; while the usable memory equates to 7744 MB/core or 363 GB/node. Each node has 2 NVIDIA Volta V100 w/32GB GPU memory.  

For serial jobs, we will allow node sharing on GPU nodes so a job may request either 1 or 2 GPUs (--ntasks=XX --gpus-per-node=1 or --ntasks=XX --gpus-per-node=2)

For parallel jobs (nodes>1), we will not allow node sharing. A job may request 1 or 2 GPUs ( gpus-per-node=1 or gpus-per-node=2 ) but both GPUs will be allocated to the job.

Quad GPU Node

For quad GPU node, it has 48 cores/node. The physical memory equates to 16.0 GB/core or 768 GB/node; while the usable memory equates to 15,872 MB/core or 744 GB/node.. Each node has 4 NVIDIA Volta V100s w/32 GB GPU memory and NVLink.

For serial jobs, we will allow node sharing on GPU nodes, so a job can land on a quad GPU node if it requests 3-4 GPUs per node (--ntasks=XX --gpus-per-node=3 or --ntasks=XX --gpus-per-node=4), or requests quad GPU node explicitly with using --gpus-per-node=v100-quad:4, or gets backfilled with requesting 1-2 GPUs per node with less than 4 hours long. 

For parallel jobs (nodes>1), only up to 2 quad GPU nodes can be requested in a single job. We will not allow node sharing and all GPUs will be allocated to the job.

Partition time and job size limits

Here is the walltime and node limits per job for different queues/partitions available on Pitzer:

To specify a partition for a job, either add the flag --partition=<partition-name> to the sbatch command at submission time or add this line to the job script:
#SBATCH --paritition=<partition-name>

To access one of the restricted queues, please contact OSC Help. Generally, access will only be granted to these queues if the performance of the job cannot be improved, and job size cannot be reduced by splitting or checkpointing the job.

Job/Core Limits

An individual user can have up to the max concurrently running jobs and/or up to the max processors/cores in use. However, among all the users in a particular group/project, they can have up to the max concurrently running jobs and/or up to the max processors/cores in use.

For more information about citations of OSC, visit https://www.osc.edu/citation.

To cite Pitzer, please use the following Archival Resource Key:

ark:/19495/hpc56htp

Please adjust this citation to fit the citation style guidelines required.

Ohio Supercomputer Center. 2018. Pitzer Supercomputer. Columbus, OH: Ohio Supercomputer Center. http://osc.edu/ark:19495/hpc56htp

Here is the citation in BibTeX format:

@misc{Pitzer2018,
ark = {ark:/19495/hpc56htp},
url = {http://osc.edu/ark:/19495/hpc56htp},
year  = {2018},
author = {Ohio Supercomputer Center},
title = {Pitzer Supercomputer}
}

And in EndNote format:

%0 Generic
%T Pitzer Supercomputer
%A Ohio Supercomputer Center
%R ark:/19495/hpc56htp
%U http://osc.edu/ark:/19495/hpc56htp
%D 2018

Here is an .ris file to better suit your needs. Please change the import option to .ris.

These are the public key fingerprints for Pitzer:
pitzer: ssh_host_rsa_key.pub = 8c:8a:1f:67:a0:e8:77:d5:4e:3b:79:5e:e8:43:49:0e 
pitzer: ssh_host_ed25519_key.pub = 6d:19:73:8e:b4:61:09:a9:e6:0f:e5:0d:e5:cb:59:0b 
pitzer: ssh_host_ecdsa_key.pub = 6f:c7:d0:f9:08:78:97:b8:23:2e:0d:e2:63:e7:ac:93 

These are the SHA256 hashes:​
pitzer: ssh_host_rsa_key.pub = SHA256:oWBf+YmIzwIp+DsyuvB4loGrpi2ecow9fnZKNZgEVHc 
pitzer: ssh_host_ed25519_key.pub = SHA256:zUgn1K3+FK+25JtG6oFI9hVZjVxty1xEqw/K7DEwZdc 
pitzer: ssh_host_ecdsa_key.pub = SHA256:8XAn/GbQ0nbGONUmlNQJenMuY5r3x7ynjnzLt+k+W1M 

This page includes a summary of differences to keep in mind when migrating jobs from other clusters to Pitzer. 

Guidance for Oakley Users

Guidance for Owens Users

Hardware Specifications

File Systems

Pitzer accesses the same OSC mass storage environment as our other clusters. Therefore, users have the same home directory, project space, and scratch space as on the Owens cluster.

Software Environment

Pitzer uses the same module system as Owens.

Use   module load <package>   to add a software package to your environment. Use   module list   to see what modules are currently loaded and  module avail   to see the modules that are available to load. To search for modules that may not be visible due to dependencies or conflicts, use   module spider  . 

You can keep up to on the software packages that have been made available on Pitzer by viewing the Software by System page and selecting the Pitzer system.

Programming Environment

Like Owens, Pitzer supports three compilers: Intel, PGI, and gnu. The default is Intel. To switch to a different compiler, use  module swap intel gnu  or  module swap intel pgi . 

Pitzer also use the MVAPICH2 implementation of the Message Passing Interface (MPI), optimized for the high-speed Infiniband interconnect and support the Advanced Vector Extensions (AVX2) instruction set.

See the Pitzer Programming Environment page for details. 

Accounting

Below is a comparison of job limits between Pitzer and Owens:

Please see Queues and Reservations for Pitzer and Batch Limit Rules for more details.

Guidance for Ruby Users

In late 2018, OSC installed 260 Intel® Xeon® 'Skylake' processor-based nodes as the original Pitzer cluster. In September 2020, OSC installed additional 398 Intel® Xeon® 'Cascade Lake' processor-based nodes as part of a Pitzer Expansion cluster. This expansion makes Pitzer a heterogeneous cluster, which means that the jobs may land on different types of CPU and behaves differently if the user submits the same job script repeatedly to Pitzer but does not request the resources properly. This document provides you some general guidance on how to request resources on Pitzer due to this heterogeneous nature. 

Step 1: Identify your job type

According to this table,

• If your job requests standard compute node(s) or dual GPU node(s), it can potentially land on different types of nodes and may result in different job performance. Please follow the steps below to determine whether you would like to restrain your job to a certain type of node(s). 
• If your job requests quad GPU node(s), large memory node(s), or huge memory node(s), please check pitzer batch limit rules on how to request these special types of resources properly. 

Step 2: Perform test

This step is to submit your jobs requesting the same resources to different types of nodes on Pitzer. For your job script is prepared with either PBS syntax or Slurm syntax:

Request 40 or 48 core nodes

#SBATCH --constraint=40core
#SBATCH --constraint=48core

Request 16gb, 32gb gpu

#SBATCH --constraint=v100
#SBATCH --constraint=v100-32g --partition=gpuserial-48core

Once the script is ready, submit your jobs to Pitzer and wait till the jobs are completed. 

Step 3: Compare the results

Once the jobs are completed, you can compare the job performances in terms of core-hours, gpu-hours, walltime, etc. to determine how your job is sensitive to the type of the nodes. If you would like to restrain your job to land on a certain type of nodes based on the testing, you can add  #SBATCH --constraint=. The disadvantage of this is that you may have a longer queue wait time on the system. If you would like to have your jobs scheduled as fast as possible and do not care which type of nodes your job will land on, do not include the constraint in the job request.