Skip to content
Contact Support

Job Limits

These are open for review if you find any of them unreasonable or inefficient.

Info

This page is about hard limits applied to jobs and users. For more information about per project usage limitations, see our page on Fairshare.

Per Job

job limits

  • 10 nodes
  • 21 days walltime
  • 21 node-days (walltime x nodes)

Jobs using the debug QoS have much smaller limits.

Per User

  • 2688 CPU cores occupied,
  • 3528 core-days booked by running jobs.
  • 6 TB of memory occupied
  • 31.5 TB-days booked by running jobs.
  • 6 GPUs occupied, 14 GPU-days booked by running jobs.
  • No user can have more than 1,000 jobs in the queue at a time.