HCC is pleased to announce a major addition of GPU resources to the Crane cluster, funded by a recent Nebraska EPSCoR award. This expansion consists of 21 GPU-enabled nodes, each with 2 x Nvidia Tesla V100 cards with 32GB RAM. This expansion more than doubles the amount of GPU-enabled nodes in Crane.

These nodes are part of the existing gpu partition. If you are already submitting GPU jobs to Crane, no additional changes are needed to utilize this hardware. The GPU resources on Crane consist of a variety of Tesla models. If you would like to request a V100 node specifically, you can do so by adding the --constraint=gpu_v100 option to your SLURM scripts. Additionally, some nodes are reserved for priority access use. HCC supports opportunistic use of these nodes via a separate guest_gpu partition. For more details on opportunistic use, please see the link below.

Full details on GPU resources are available at https://hcc.unl.edu/docs/guides/submitting_jobs/submitting_cuda_or_openacc_jobs/.