Skip to content

Commit

Permalink
support for gpu's untested
Browse files Browse the repository at this point in the history
  • Loading branch information
Joon-Klaps committed Nov 18, 2024
1 parent 3dbc69e commit b498e4e
Showing 1 changed file with 34 additions and 13 deletions.
47 changes: 34 additions & 13 deletions conf/vsc_kul_uhasselt.config
Original file line number Diff line number Diff line change
Expand Up @@ -39,7 +39,7 @@ process {
stageOutMode = "rsync"
errorStrategy = { sleep(Math.pow(2, task.attempt) * 200 as long); return 'retry' }
maxRetries = 5
// array = 50
array = 50
}

// Specify that singularity should be used and where the cache dir will be for the images
Expand All @@ -66,14 +66,11 @@ profiles {
config_profile_description = 'HPC_GENIUS profile for use on the genius cluster of the VSC HPC.'
config_profile_contact = 'GitHub: @Joon-Klaps - Email: [email protected]'
config_profile_url = 'https://docs.vscentrum.be/en/latest/index.html'
max_memory = 703.GB // 768 - 65 so 65GB for overhead, max is 720000MB
max_time = 168.h
max_cpus = 36
}

process {
resourceLimits = [
memory: 703.GB,
memory: 703.GB, // 768 - 65 so 65GB for overhead, max is 720000MB
cpus: 136,
time: 168.h
]
Expand All @@ -96,6 +93,17 @@ profiles {
}
}
}

withLabel: gpu {
beforeScript = 'cuDNN/8.4.1.50-CUDA-11.7.0'
queue = { task.time >= 72.h ? 'gpu_p100_long,gpu_v100_long,amd_long' : 'gpu_p100,gpu_v100' }
containerOptions = {
workflow.containerEngine == "singularity"
? '--containall --cleanenv --nv'
: (workflow.containerEngine == "docker" ? '--gpus all' : null)
}
}

clusterOptions = { "--clusters=genius --account=$tier1_project" }
scratch = "$scratch_dir"
}
Expand All @@ -107,14 +115,11 @@ profiles {
config_profile_description = 'HPC_WICE profile for use on the Wice cluster of the VSC HPC.'
config_profile_contact = 'GitHub: @Joon-Klaps - Email: [email protected]'
config_profile_url = 'https://docs.vscentrum.be/en/latest/index.html'
max_memory = 1968.GB // max is 2016000
max_cpus = 72
max_time = 168.h
}

process {
resourceLimits = [
memory: 1968.GB,
memory: 1968.GB, // max is 2016000
cpus: 72,
time: 168.h
]
Expand All @@ -137,6 +142,25 @@ profiles {
}
}
}

withLabel: gpu {
beforeScript = 'cuDNN/8.4.1.50-CUDA-11.7.0'
queue = { task.time >= 72.h ?
['dedicated_big_gpu',
'dedicated_big_gpu_h100',
].join(',')
:
['gpu',
'gpu_h100',
].join(',')
}
containerOptions = {
workflow.containerEngine == "singularity"
? '--containall --cleanenv --nv'
: (workflow.containerEngine == "docker" ? '--gpus all' : null)
}
}

clusterOptions = { "--clusters=wice --account=$tier1_project"}
scratch = "$scratch_dir"
}
Expand All @@ -147,14 +171,11 @@ profiles {
config_profile_description = 'HPC_SUPERDOME profile for use on the genius cluster of the VSC HPC.'
config_profile_contact = 'GitHub: @Joon-Klaps - Email: [email protected]'
config_profile_url = 'https://docs.vscentrum.be/en/latest/index.html'
max_memory = 5772.GB // 6000 - 228 so 228GB for overhead, max is 5910888MB
max_cpus = 14
max_time = 168.h
}

process {
resourceLimits = [
memory: 5772.GB,
memory: 5772.GB, // 6000 - 228 so 228GB for overhead, max is 5910888MB
cpus: 14,
time: 168.h
]
Expand Down

0 comments on commit b498e4e

Please sign in to comment.