Deploying NVIDIA Omniverse on HPC Resources at Clemson University

Digital twins allow designers to create virtual worlds that represent real things, people, or processes. NVIDIA Omniverse is a platform that allows for design, use, and training of digital twin systems. This paper discusses deploying the NVIDIA Omniverse platform on a datacenter class NVIDIA OVX system managed by Clemson University Research Computing and Data (RCD). We discuss integrating the Omniverse platform into the existing HPC infrastructure. We discuss lessons learned from this pilot program and items for future work on the project. We hope this information will be useful to other institutions considering an on-premises implementation of NVIDIA's Omniverse platform.