Skip to content
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.

Sign up for GitHub

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

Enable vega gpu #215

Merged
merged 2 commits into from
Dec 3, 2024
Merged

Enable vega gpu #215

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
0436452
Add gpu config back into reframe for vega
Dec 3, 2024
4e0d2d7
Update vega config
Dec 3, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
89 changes: 45 additions & 44 deletions config/izum_vega.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,50 +68,51 @@
},
'descr': 'CPU partition Standard, see https://en-doc.vega.izum.si/architecture/'
},
# {
# 'name': 'gpu',
# 'scheduler': 'slurm',
# 'prepare_cmds': [
# common_eessi_init(),
# # Pass job environment variables like $PATH, etc., into job steps
# 'export SLURM_EXPORT_ENV=ALL',
# # Needed when using srun launcher
# # 'export SLURM_MPI_TYPE=pmix', # WARNING: this broke the GROMACS on Vega
# # Avoid https://github.com/EESSI/software-layer/issues/136
# # Can be taken out once we don't care about old OpenMPI versions anymore (pre-4.1.1)
# 'export OMPI_MCA_pml=ucx',
# ],
# 'launcher': 'mpirun',
# # Use --export=None to avoid that login environment is passed down to submitted jobs
# 'access': ['-p gpu', '--export=None'],
# 'environs': ['default'],
# 'max_jobs': 60,
# 'devices': [
# {
# 'type': DEVICE_TYPES[GPU],
# 'num_devices': 4,
# }
# ],
# 'resources': [
# {
# 'name': '_rfm_gpu',
# 'options': ['--gpus-per-node={num_gpus_per_node}'],
# },
# {
# 'name': 'memory',
# 'options': ['--mem={size}'],
# }
# ],
# 'features': [
# FEATURES[GPU],
# ] + list(SCALES.keys()),
# 'extras': {
# # Make sure to round down, otherwise a job might ask for more mem than is available
# # per node
# 'mem_per_node': 476.837 * 1024 # in MiB (should be checked, its unclear from slurm.conf)
# },
# 'descr': 'GPU partition, see https://en-doc.vega.izum.si/architecture/'
# },
{
'name': 'gpu',
'scheduler': 'slurm',
'prepare_cmds': [
common_eessi_init(),
# Pass job environment variables like $PATH, etc., into job steps
'export SLURM_EXPORT_ENV=ALL',
# Needed when using srun launcher
# 'export SLURM_MPI_TYPE=pmix', # WARNING: this broke the GROMACS on Vega
# Avoid https://github.com/EESSI/software-layer/issues/136
# Can be taken out once we don't care about old OpenMPI versions anymore (pre-4.1.1)
'export OMPI_MCA_pml=ucx',
],
'launcher': 'mpirun',
# Use --export=None to avoid that login environment is passed down to submitted jobs
'access': ['-p gpu', '--export=None'],
'environs': ['default'],
'max_jobs': 60,
'devices': [
{
'type': DEVICE_TYPES[GPU],
'num_devices': 4,
}
],
'resources': [
{
'name': '_rfm_gpu',
'options': ['--gpus-per-node={num_gpus_per_node}'],
},
{
'name': 'memory',
'options': ['--mem={size}'],
}
],
'features': [
FEATURES[GPU],
] + list(SCALES.keys()),
'extras': {
GPU_VENDOR: GPU_VENDORS[NVIDIA],
# Make sure to round down, otherwise a job might ask for more mem than is available
# per node
'mem_per_node': 476.837 * 1024 # in MiB (should be checked, its unclear from slurm.conf)
},
'descr': 'GPU partition, see https://en-doc.vega.izum.si/architecture/'
},
]
},
],
Expand Down
Loading