Partition Definitions

There are eight (8) Slurm partitions on Shamu that users are able to utilize. Below are their definitions.
  • defq - This is the main queue and all users are able to submit to it. By default if you do not specify a queue name with your qsub or qlogin session, your job will automatically go into this queue.
  • bigmem - This queue is protected by an Access Control List. This queue is specifically for large memory or large multi-threaded jobs.
  • ids - This queue is protected by an Access Control List. Only users of a specific lab are able to use this queue
  • gpu - This queue is protected by an Access Control List. This is the queue used to access the Tesla K80 GPU nodes.
  • gpu-v100 - This queue is proteceted by an Access Control List. This is the queue used to access the Volta V100 GPU nodes.
  • millwater - This queue is protected by an Access Control List. Only users of a specific lab are able to use this queue.
  • plasmon - This queue is protected by an Access Control list. This queue is specific for Dr. Large.
  • gpu-v100-shared - This queue is for shared access to one of the V100 nodes. It allows up to 4 GPU jobs to be executed at the same time. Users of the "gpu-v100" partition are able to access this queue.
-- AdminUser - 14 Sep 2016
Topic revision: r8 - 24 Jul 2020, AdminUser
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback