We provide specialist computing services for research projects needing high capacity systems.
In 2010 we introduced a new high performance computing environment into which research teams can add compute nodes and storage dedicated to their needs (eg the ATLAS group in physics).
The University wide service consists of 324 cores and a parallel file system providing 30TB of Lustre scratch storage with Infiniband interconnects. A further 400 cores and 115TB of storage are dedicated to specific research groups.
The HPC environment provides a centrally-managed front-end system with local file storage which can be accessed from any node in the attached clusters.
The service is supported by a dedicated team within ITS who maintain the system and can advise research teams on best practice, simplifying the process to deploy new research facilities.
Research teams fund their own clusters within the environment and make a contribution towards the costs of the centralised facilities.
Updated on 06 July 2012