The FLACS-Cloud service runs on the Cirrus system at EPCC. The Cirrus compute provision consists of 280 compute nodes connected by a single Infiniband fabric. Cirrus compute nodes each contain two 2.1 GHz, 18-core Intel Xeon E5-2695 (Broadwell) series processors. Each of the cores in these processors support 2 hardware threads (Hyperthreads), which are enabled by default.

The compute nodes on Cirrus have 256 GB of memory shared between the two processors. The memory is arranged in a non-uniform access (NUMA) form: each 18-core processor is a single NUMA region with local memory of 128 GB.

FLACS-Cloud typically uses one core per FLACS simulation and rather runs many jobs in parallel.

Full information about the Cirrus hardware is available here.