Page tree
Skip to end of metadata
Go to start of metadata

Job cost = ncpus_request x max[ 1, (ncpus_per_node/mem_per_node)x(mem_request/ncpus_request)] x walltime_usage x queue_charge_rate

Queue

CPUs request

GPUs request

Memory Request (GB)

Walltime Usage

Cost

Comments

normal4n.a.16 GB5 hours4 x 5 x 2 = 40 SU
normal8n.a.16 GB5 hours8 x 5 x 2 = 80 SUCPU request dominates.
normal8n.a.128 GB5 hours48 x (128/192) x 5 x 2 = 320 SUMemory request dominates. 32=48*(128/192) cpus is proportion of node resources.
normal8n.a.192 GB5 hours48 x 5 x 2 = 480 SUMemory request dominates. 
express8n.a.16 GB5 hours8 x 5 x 6 = 240 SUExpress charge rate is 6.0
gpuvolta121*90GB5 hours12 x 5 x 3 = 180 SUCPU request dominates. *Because of the `ncpus = 12 x ngpus` constraint in GPU job request, ngpus determines the cost through ncpus in the equation. 
gpuvolta121*380GB5 hours48 x (380/382) x 5 x 3 = 716.23 SUMemory request dominates. 
  • No labels