Slicing Supercomputers for Trusted Research Environments

Learn how to slice supercomputers for trusted research environments in this 32-minute conference talk from the OpenInfra Foundation. Discover StackHPC's specialized approach to creating multi-tenant cloud environments that provide isolation between competing workloads and scientific communities using OpenStack, Slurm, and Kubernetes technologies. Explore how Azimuth enables self-service catalogs of science platforms for target communities by leveraging local cloud resources and external Kubernetes clusters. Understand the integration with existing Resource Allocation processes, including the use of OpenStack Blazar for resource management. Examine the critical requirements for relationships between infrastructure providers and Trusted Research Environment providers, and gain insights from early implementation results on the DAWN AI supercomputer at the University of Cambridge. Master the techniques for connecting scientists to the infrastructure they need while maintaining security and isolation in high-performance computing environments.