Compute Cluster

The goal of the AI for Oncology Lab is to use innovations in artificial intelligence for the benefit of cancer patients. Our researchers and developers require powerful graphical processing units (GPUs) to create high-quality and performant deep learning-based algorithms.

We have access to two GPU clusters, one hosted at the Netherlands Cancer Institute and the other one provided by the Snellius cluster at Surf, where we have dedicated access to three nodes with 4 x A100 (40GB).

Our on-prem cluster, named Kosmos, hosts various compute nodes and provides compute capabilities for AI researchers at the Netherlands Cancer Institute.

Some basic statistics:

Type Value
Active users/month ~75
Compute Nodes 15
Number of cores ~1250
Number of GPUs 82
Total memory ~12TB

Our cluster consists of the following GPUs: H100 (80GB), A100 (80GB), A6000, P6000, RTX8000 and RTX2080Ti.