Nvidia backs projects to tap GPUs for cancer research

Stanford professor Vijay Pande

The Nvidia Foundation has awarded $400,000 to computational cancer projects at the Dana-Farber Cancer Institute and Stanford University. Nvidia backed the two projects--which took an equal split of the funding--as part of an initiative to finance programs that use parallel computing to advance cancer research.

Staffers at the National Cancer Institute (NCI) helped Nvidia--a company best known for its graphics processing units (GPUs)--to pick the projects from the more than 20 applications it received. Both projects are underpinned by GPUs, but go about accessing the technology in very different ways. The funding received by Stanford will go toward its long-running [email protected] project, which asks the public to contribute their computing power to research into protein folding.

When someone downloads the [email protected] software, their computer becomes part of a network of more than 10,000 other home PCs. Stanford has used the resulting distributed supercomputer to generate data for more than 100 research papers. The plan now is to use the $200,000 awarded by Nvidia to incorporate molecular simulation and machine learning into the project. Home PCs will continue to power these new features.

The Dana-Farber project is focused on making sense of the tumor genome data the institute gathers from 10,000 patients each year. Dana-Farber professor John Quackenbush plans to use GPUs to look at the data, an approach he hopes will reveal genetic similarities between different types of cancer. If the same mutation is found in colon and breast cancer, for example, a drug designed to treat one form could be repurposed for use against the other. 

- read the Nvidia blog