The SciML group manages the PEARL service for the Alan Turing Institute and its partners working on AI for Science. The service relies on GPU and IPU clusters (Graphcore), and complement other compute resources managed by the Scientific Computing Department. The full details of the systems are given below.
PEARL / GRAPHCORE
| GPU / IPU Cluster
| The PEARL service, provided through a collaboration between The Alan Turing Institute and Science and Technology Facilities Council, is a state-of-the-art GPU-based system for AI and machine learning research. Recently, a massively parallel Intelligence Processing Unit (IPU) by Graphcore has been integrated into PEARL's login node.
|
SCARF
| CPU & GPU Cluster
| SCARF is an HPC (High Performance Computing) cluster open to all STFC departments and Diamond, which is run by Scientific Computing's Research Infrastructures Group to support scientists and researchers.
|
STFC CLOUD
| Cloud virtual machines with multiple GPUs | The STFC Cloud is a dedicated cloud infrastructure which provides access to compute resources for users across STFC and partner organisations.
|
ECHO
| Cloud storage
| Echo is the Ceph-backed storage service
developed at STFC to meet the UK’s data
storage commitments to the Large Hadron
Collider (LHC) experiments at CERN.
|
