Cirrus provides a CPU-based high performance computing (HPC) service within the EIDF infrastructure. The Cirrus service supports high performance, parallel workloads for a wide variety of different data analysis and modelling and simulation use cases. Availability Cirrus is available to all potential EIDF users. Service profile Full details of the Cirrus service can be found at: Dedicated Cirrus website - https://www.cirrus.ac.ukDedicated Cirrus documentation - https://docs.cirrus.ac.uk The Cirrus hardware is provided by an HPE Cray EX4000 system with 73,728 CPU cores in 256 compute nodes. The compute nodes each have 288 cores. They are dual socket nodes with two 144-core AMD EPYC(tm) 9825 processors. There are 192 standard memory nodes and 64 high memory nodes. Standard memory nodes have 768 GB DDR5 memory, and high memory nodes have 1,536 GB DDR5 memory. Compute nodes are connected together by a HPE Cray Slingshot 11 interconnect. Users on the Cirrus service can access the following EIDF storage options: CephFS with 3-copies redundancye1000 with backupS3 without backup This article was published on 2026-01-07
