Additional SMP nodes: short-term fix for busy cluster

Dear Users,

To help address the recent high utilization on the SMP cluster, we have added 20 legacy Intel Sandy Bridge nodes to SMP.  Each node has 16 cores and 64 GB of RAM.  To access these nodes, you will need to modify your job submission script to target the legacy partition:

#SBATCH --cluster=smp

#SBATCH --partition=legacy


You can use the sinfo command to gauge the availability of nodes within various partitions:

[ ~]$sinfo

PARTITION AVAIL TIMELIMIT NODES STATE NODELIST smp* up infinite 42 mix smp-n[0,24,28-31,35,38-39,48,50,52-56,58-61,68-69,71,73,81-83,86,92,97,102,104,108,112-113,119-120,128,130,132,148,152]

smp* up infinite 95 alloc smp-n[6,8,12,15-16,19,22,25-27,32-34,36-37,40-47,49,51,57,62-67,70,72,74-80,84-85,87-91,93-96,98-101,103,105-107,109-111,114-118,121-122,126-127,129,131,133-147,149-151,153-155]

smp* up infinite 14 idle smp-n[1-5,7,9-11,13-14,20-21,23]

high-mem up infinite 2 mix smp-256-n2,smp-512-n2

high-mem up infinite 1 alloc smp-512-n1

high-mem up infinite 2 idle smp-256-n1,smp-nvme-n1

legacy up infinite 3 alloc legacy-n[0-2]

legacy up infinite 17 idle legacy-n[3-19]

[ ~]$



Please submit an online ticket if you have questions or need help,