04 Aug 2026
In March 2026 UKRI announced the first phase of UK National Compute Resources (NCR) that will provide high performance computing facilities (HPC) to support research across the country.
Cirrus is one of the initial set of six NCRs and is the only one that is already up and running and supporting researchers. As part of the UKRI investment, Cirrus has more than doubled in size.
The expansion added 384 standard memory nodes to the existing 256 nodes (192 standard memory and 64 high memory nodes) taking the system up to 640 compute nodes in total and 184,320 CPU cores all connected by HPE Cray Slingshot 11 interconnect. The new nodes are already available to Cirrus users.
Full details of the hardware that make up Cirrus are available on the Hardware page in the Cirrus Documentation.
The expansion has also led to some changes in the scheduler setup to allow for more, larger jobs and an updated priority configuration to support larger jobs. For details of the updated scheduler configuration, see the Scheduler page in the Cirrus Documentation.
Cirrus compliments the announced next national supercomputer which, like Cirrus, will also be operated by EPCC, the first UK National Supercomputing Centre.