Back to cluster home page FAQ Open MP
Open MP implements shared memory parallel processing. To use Open MP in the CMU Physics cluster, you may compile programs written in fortran (fortran-77 or fortran-90) using the PGI compilers. For example, you can compile your program (with pgf77, pgf90 or pghpf) using the "-mp" flag, then enter setenv NCPUS 2, before running your program.
Cluster CPU usage guidelines apply, and you must be sure that all processes are "niced" at level +15. Do not run parallel programs without careful performance tuning and proof through benchmarking that high parallel efficiency is achieved.