Advanced Research Computing (ARC) provides research services and support for computational and storage technology. ARC provides HPC services, a variety of storage services, both active and archive, and a private cloud, connected by a high-performance network. The University seeks to refresh its general-purpose High-Performance Computing (HPC) cluster to support a broad range of research workloads. The current cluster consists of approximately 21,000 CPU cores and 504 GPUs, many of which are faculty-owned. The existing storage infrastructure is based on GPFS with a capacity of roughly 3 PB. The refreshed cluster should scale this storage capacity to at least 8 PB and will continue to use Slurm as the primary job scheduler.
