The Hewlett-Packard cluster was funded by a New York State Empire State Development grant to provide the WNY and NY State industrial community with access to state-of-the-art high performance computing resources (hardware, software and consulting services) to help foster economic development. This cluster consists of 216 HP SL230 Gen8 servers of which 144 are “Parallel” compute nodes with a FDR InfiniBand interconnect. The remaining 72 nodes are “Serial” nodes with GigE connections. The nodes consist of two Intel “Ivy Bridge” Xeon 2.6GHz (E5-2650V2) 8-core processors, 64GB of memory and 500GB of local scratch.
|Type of Node||# of Nodes||# Cores/Node||Clock Rate||RAM||Network||SLURM TAGS||Local /scratch|
|Compute ||144||16||2.60GHz||64GB||FDR Infiniband||IB CPU-E5-2650v2 ||500GB|
|Sample SLURM Directives |
|To request Infiniband networked nodes only: |
|--nodes=2 --ntasks-per-node=8 --constraint=IB |
The industrial cluster is broken up into several partitions (formerly known as "queues") that users can request to have their jobs run on. The "compute" partition is only available to industrial partners. UB faculty and students have the option of running their jobs in the "scavenger" partition. This allows jobs to run when there are no other pending jobs in the compute partition. Once an industrial user submits a job requesting resources, jobs in the scavenger partition are stopped and requeued. Please contact us if you'd like to test your jobs using this preemption feature of the job scheduler. Note that your jobs MUST be able to checkpoint. Please see our documentation about checkpointing here.
|Partition Name||Time Limit||Default Number CPUs||Notes |
|industry ||72 hours||1|| available only to industrial partners |
|scavenger||72 hour||1||--requeue flag required for jobs to be restarted |