Scientific computing : Technologies

Data servers

Technologies

We provide high-performance computing systems essential for capturing, storing and analysing scientific data. 

CAMP

Our high-performance computing and data analytics platform (CAMP: Crick data Analysis and Management Platform) is available to all Crick researchers.

The following compute clusters are currently available:

  • CPU cluster with over 6000 cores and 56 TB RAM, divided into standard and high memory nodes (replacement due 2023)
  • Main GPU cluster with 160 Nvidia V100 GPUs, 1600 CPU cores and 30 TB RAM
  • Structural Biology GPU cluster with 44 Nvidia RTX 5000 GPUs and 4 TB RAM

A number of CPU and GPU compute nodes are also available for interactive use.

All compute nodes run Linux and are controlled by the SLURM batch job scheduler. Jobs are submitted either from the Linux command line or a dedicated web portal (Open OnDemand).

Up to 10 petabytes (increasing to 15 petabytes) of high-performance research data storage is available covering all of the institute’s research data needs, accessible from the CAMP compute cluster and via workstations across the institute.

A high performance, low latency InfiniBand network is used to connect compute and storage nodes.

Our Research Computing Platforms team offers training and support to ensure that scientists understand and make the best use of the systems on offer.