Case Study: California Institute of Technology achieves scalable cancer research breakthroughs with Oracle Cloud Infrastructure

Caltech taps Oracle Cloud in the race to treat, cure cancer

The California Institute of Technology’s Thomson Lab faced a major big data and high-performance computing challenge in its cancer research, as Matt Thomson and his team needed to train massive machine learning and protein engineering models—some with up to 100 billion parameters—using resources far beyond what local HPC clusters could provide. Caltech turned to Oracle and Oracle Cloud Infrastructure (OCI) GPU instances to gain the scalable computing power needed for this work.

With Oracle’s help, Caltech ran a proof of concept that enabled the lab to create and test 1,000 models, compared with only 10 at a time previously, dramatically improving its research throughput. Oracle also collaborated closely with the team on future data storage and management needs, helping position Caltech to make larger datasets more accessible for model training and advance its cancer research more effectively.

California Institute of Technology

Matt Thomson

Assistant Professor of Computational Biology