Skip to content

Lochan HPC Cluster

Information Services run a High Performance Compute (HPC) cluster called Lochan, that is available to use by all staff, students and affiliates of the University of Glasgow.

The platform and team around it have been supporting researchers in all colleges since 2014. Managed by Research Computing as a Service (RCaaS) and motivated by the needs of the community, Lochan is here for people with large upscaling needs but also users who are just starting their HPC journey.

Technology

Lochan offers the following compute resources:

  • 856 virtual CPUs
  • 3 NVIDIA Tesla V100 GPUs
  • 2 Nvidia L40S GPUs
  • 8 Nvidia GeForce GTX 1080 GPUs
  • 14+TB GB of memory with up to 2TB per node

All servers of the cluster run on Oracle Linux 9. Oracle Linux (OL) is a Red Hat based open-source operating system.

The cluster's scheduler is the Slurm Workload Manager, developed by SchedMD. This software is crucial for HPC and helps achieve fair usage of all available compute resources.

The cluster and storage mounted on to the system are located in Saughfield House on the University of Glasgow Campus next to the Library.

Acknowlegement

Where Lochan is used in the development of research outputs the following attribution should be used:

This research utilised the University of Glasgow’s Lochan HPC, supported by University of Glasgow Research Computing as a Service. (https://www.gla.ac.uk/myglasgow/it/hpcc/lochan/)