We are very pleased to report that we have added 40 new compute nodes to the HPC cluster. These nodes were ordered in October 2021, but due to supply chain issues and vendor delays, were received at the beginning of this month.
The new nodes were added to the owner Slurm accounts/queues today (September 14), and are automatically available for jobs submitted to dedicated queues/partitions. There is no need to change any parameters; if you purchased resources on our cluster, Slurm will automatically include these new resources when scheduling jobs.
The new nodes have the following specs:
- Two 32-core AMD EPYC 7543 2.80Ghz Processors
- 512 GB RAM
- 512GB solid-state disks
- InfiniBand 100Gbps networking
As part of our annual upgrade procedure, we are also rotating more of our prior owner nodes into the general access portion of the cluster. This will happen in the coming weeks, and will be transparent and not have any impact on running jobs.