New Research Cluster GPU Nodes

November 4, 2022
Categories: News, URC

We have acquired new GPU nodes with a total of 32 GPUs, which have all been added to the Research Cluster. Within these new systems are 24 NVIDIA A40 Data Center GPUs which are best for single precision floating point operations and AI workloads and 8 NVIDIA A100 Tensor Core GPUs which are best for double precision operations required for many types of simulation. Sixteen older GeForce GTX-1080Ti GPUs will be transferred to the Educational Cluster. Sixteen obsolete NVIDIA Tesla K80 GPUs in the Educational Cluster will be retired from service at the end of this semester. The detailed configurations are listed below.

New Research Cluster GPU Nodes

  • 2 “A100” GPU nodes with
    • Dual 16-Core Intel Xeon Gold 6326 CPU @ 2.90GHz (32 cores total)
    • 256GB RAM (8GB / core)
    • 4 x NVIDIA A100 Tensor Core GPUs (80GB HBM2e RAM per GPU)
    • 100Gbit HDR Infiniband Interconnect
  • 6 “A40” GPU nodes with
    • Dual 16-Core Intel Xeon Gold 6326 CPU @ 2.90GHz (32 cores total)
    • 256GB RAM (8GB / core)
    • 4 x NVIDIA A40 Data Center GPUs (48GB GDDR6 ECC RAM per GPU)
    • 100Gbit HDR Infiniband Interconnect

The need for more GPU capability was identified by a number of users in a recent survey and we will continue to evaluate demand in different ways. Your responses are helping us to continue to improve the Research Computing services.