This workshop will present techniques for improving the performance of scientific software on High Performance Computing (HPC) systems such as those available at OSC. The focus will be on serial performance, including vectorization and cache utilization, with a brief mention of parallel computing.
Topics will include:
- Hardware overview
- Important factors for good performance
- Compiler optimizations
- Profiling tools
Who Should Attend
Current or prospective OSC users who are involved in the development of software for scientific computing.
Some familiarity with Linux/UNIX and either Fortran 90 or C/C++ is preferred.