Daily Bulletin

Run your high-throughput jobs on the Casper cluster

May 31, 2022

CISL recommends running small jobs that use only CPUs on the Casper cluster’s high-throughput computing (HTC) nodes. Casper has 64 HTC nodes specifically for running small batch jobs. It also has:

  • More shared resources than the Cheyenne share queue.
  • A far higher concurrent-use limit for CPU cores than Cheyenne: 468 vs. 36.
  • More available memory, plus NVMe swap for overflow.

See Starting Casper jobs with PBS for an example of how to run a job on the HTC nodes. You can even use peer scheduling to run your data-analysis HTC workflows on Casper after successful completion of your HPC model runs on Cheyenne. See Peer scheduling between Cheyenne and Casper for how to do that.