Announcements, 14th October 2015

Extended maintenance sessions in October.

We would like to remind our users of the extended outage on Shaheen II in October for necessary maintenance:

25th October for 3 days

There will be no access to the system during these periods.

Tip of the week: Queues on Shaheen II Cray XC40

Two different queue are available on Shaheen II:

  • workq: wall clock time max 24 hours with a maximum of 6174 nodes. The workq_high and workq_low have been merged with workq. This is effective with the deployment of SLURM dynamic power management, which balance effectively the total power available across the nodes and jobs depending on the applications power usage.
  • 72hours: wall clock time max 72 hours for a maximum of 32 nodes at a given time on the whole system. This queue is mainly intended for pre/post processing and running 1 node jobs that cannot be check pointed to finish within stipulated 24 hours wall clock time.

Follow us on Twitter

Follow all the latest news on HPC within the Supercomputing Lab and at KAUST, on Twitter @KAUST_HPC.

Previous Announcements

Previous Tips