Slurm Workload Manager: Difference between revisions

Revision as of 17:53, 27 April 2012

Revision as of 20:28, 9 May 2012

Line 3:

SLURM is the batch system on many of the [[TOP500]] supercomputers, including the second fastest one in the world, China's [[Tianhe-1]]. SLURM is designed to handle thousands of nodes in a single cluster and can sustain throughput of 120,000 jobs per hour.

SLURM uses a best fit algorithm based on [[Hilbert curve scheduling]] in order to optimize locality of task assignments on parallel computers.<ref name=Eitan>''Job Scheduling Strategies for Parallel Processing:'' by Eitan Frachtenberg and Uwe Schwiegelshohn 2010 ISBN 36420463203-642-04632-0 pages 138-144</ref>

==History==