HPC clusters#
We operate several distinct clusters. These pages contain specific information for each, including their individual hardware details and technical specifications.
Because these facilities were funded and built at different times for different research purposes, their architectures and access models vary. We try where possible to create a consistent user experiences across all clusters, however, this is not always possible.
GPU focused cluster specialising in AI and ML and HTC workloads. No MPI.
Our most modern general access HPC cluster, suitable for traditional HPC and MPI workloads.
A dedicated national facility supporting the creative industries, virtual production, and AI media.
A private HPC cluster for the Thermo-Fluid Systems University Technology Centre
Information on Performance benchmarking exercises conducted on Surrey HPC clusters.