-
Notifications
You must be signed in to change notification settings - Fork 314
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Replace pgi tests with new Nvidia compiler #1610
Comments
Since, PGI is not working with NUOPC (it builds, but fails at runtime), this becomes more important. We don't have nvhpc on izumi still, so I think we should add a few nvhpc tests for cheyenne, and have cheyenne run intel, and gnu, nvhpc test suites. So this will change run_sys_tests adding nvhpc for cheyenne and removing pgi for izumi. There's a limited amount of testing with nvhpc for CESM tags. I'm trying some tests in cesm2_3_alpha08d for CTSM, doing that should show us how feasible using nvhpc will be as a replacement for PGI. For portability and robustness showing we can run on four compilers: intel, gnu, nag, and nvhpc is really good. Removing a compiler from the list might be OK, but worthwhile to keep four active compilers. Since, we also test on two machines that also tests a few different versions which is also helpful as well. |
Because of #1733 the first test list for nvhpc should probably be small and only include one MPI test with DEBUG on (that will be an expected fail). But, tests with DEBUG off can be done as well as DEBUG on for mpi-serial cases. |
The PGI compiler was bought by Nvidia and as such the name of the compiler changes. The new compiler is available on Cheyenne, when it's available on izumi we'll want to change our izumi pgi tests to nvidia.
The text was updated successfully, but these errors were encountered: