NCSA and PIER Group Collaborate on Delta Supercomputer

This case study outlines the collaboration between the National Center for Supercomputing Applications (NCSA) and PIER Group in developing the Delta supercomputer. The NCSA sought funding from the National Science Foundation (NSF) to create a supercomputing resource that could handle both traditional CPU workloads and advanced GPU jobs. Due to challenges posed by COVID-19 and tight deadlines, PIER Group was engaged to assist in the orchestration, development, and deployment of the supercomputer. The implementation involved close collaboration with NVIDIA and HPE to meet NSF grant criteria and ensure system performance. The Delta supercomputer was successfully deployed, becoming the largest NSF cluster in terms of GPUs, enabling advanced research computing. In its first year, Delta utilized significant GPU and CPU resources, supporting nearly 2,000 active users and facilitating over 2 million compute jobs. This case study highlights the effective partnership and technical achievements in advancing scientific research capabilities.