High Performance Computing System

*WVCTSI to launch fee and prioritization structure for research data services on January 5*

WVCTSI's High Performance Computing (HPC) system provides researchers at West Virginia University a secure computing solution for large datasets that is faster and more powerful than any individual computer can offer. This platform allows users to model and predict outbreaks, conduct genomic analysis, analyze patient outcomes that include continuous sensor data and large dataset analysis, analyze medical images, and run artificial intelligence and machine learning algorithms.

Frequently Asked Questions

• Who can access the HPC?
  • WVCTSI members at West Virginia University
• What types of projects should use the HPC?
  • The HPC should be used to projects involving large-scale datasets containing Protected Health Information (PHI)
• Who can I talk to if I need assistance with the HPC?
  • For project questions, assistance with setup, or training, please submit an HPC Request Form in Piestar. For more immediate assistance, please contact BMIR staff directly.
• What are the technical specifications of the HPC?
  • The cluster consists of head nodes (2) and a login node, as well as 1 large memory node, 2 GPU compute nodes, and 5 Central Processing Unit’s (CPU) compute nodes, all of which are connected via a 200 GB HDR Infiniband interconnect. The large memory node is a dual-socket Intel 6248 CPU (20 cores each at 2.5 GHz) with 1.5 Terabyte (TB) Random Access Memory (RAM) and 480 GB Solid State Drive (SSD). GPU compute nodes consist of dual-socket Intel 6248 CPU (20 cores each at 2.5 GHz) with 384 GB RAM, a 1.6 TB SSD, and (2) Nvidia V100 GPUs. CPU compute nodes consist of dual-socket Intel 6248R CPU (24 cores each at 3.0 GHz) with 384 GB RAM and two 480 GB SSDs. GPU login node is a dual-socket Intel Gold 6242 (16 cores each at 2.8 GHz) with 192 GB RAM, four 480 GB SSDs, and one Nvidia V100S GPU.

Contact