Oncode GPU infrastructure to enable Artificial Intelligence and Deep Learning applications (NKI)

Keywords: large and complex datasets, using AI

Location: NKI
Contact person: Anastassis Perrakis

For more information about access, please contact Anastassis Perrakis through the Oncode community platform.

How it can help you

A distributed GPU infrastructure to enable Artificial Intelligence and Deep Learning applications in Cancer Genomics, Proteomics, and Structural Cell Biology. The revolution in AI technology is based on parallelizable algorithms and is thus largely driven by Graphical Processor Units (GPU). Unlike Central Processing Units (CPUs), which are rather limited in the number of concurrent executions of software commands (threads), GPUs process thousands of threads in parallel, opening opportunities for many novel AI applications.

Oncode therefore supports a GPU cluster, distributed amongst three Oncode institutes. This distributed infrastructure with common login and job execution procedures, ensures immediate local availability to enable new science, provides a testbed for distributed (cloud, grid, federated) computing, allows to evaluate solutions for data security/privacy and storage, ensures local extensibility, and offers the possibility to leverage local resources for Oncode use.

For further information, please contact Jacqueline Staring, programme manager at Oncode.

Jacqueline Staring

Jacqueline Staring

Programme Manager

After her study of Biomedical Sciences at Utrecht University, Jacqueline finished her PhD in virology and functional genetics at the Netherlands cancer institute. She then moved to Denmark to do a postdoc in ancient pathogens and genomics at the University of Copenhagen. At Oncode, Jacqueline is working on the technology focussed programmes such as the Drug Repurposing, Equipment & Infrastructure, and Technology Access programmes. She also likes to engage the younger Oncode community and is therefore for instance co-organizing the Oncode postdoc retreats.


Programme Manager