NVIDIA Accelerates the Search for New Treatments

Using the combined computing power of many GeForce GPUs, the Folding@home protein-folding simulation runs up to 140 times faster than on a traditional CPU. It uses a Folding@home client for NVIDIA GeForce GPUs.

Archive notice: This article was originally published on July 28, 2008. Links and embedded videos are preserved as part of the historical record.

Folding@home, a distributed-computing project at Stanford University in California, uses GPU power from millions of private computers to simulate the structure of proteins—the process of protein folding. The project aims to provide scientific insights that could help treat diseases such as cancer, BSE, cystic fibrosis and Parkinson’s. Folding@home is the latest example in a steadily growing list of applications using GPUs for non-graphics computing.

“The positive effects of using GeForce GPUs to simulate protein folding were quickly apparent,” says Vijay Pande, Assistant Professor of Chemistry at Stanford University and Director of the Folding@home project. “Computational output is increasing exponentially among teams using GeForce GPUs for folding simulations. GPU computing power gives Folding@home additional momentum and could significantly reduce the time required to conduct our biomedical research.”

The research project has prompted an enormous number of computer enthusiasts to organize into teams to process as many data units as possible. Unofficial statistics can be viewed at http://folding.extremeoverclocking.com.

  • More information