Computer Science | News

Stanford Folding@home Prof Builds New Cluster

• By Dian Schaffhauser
• 03/30/11

One of Stanford University's researchers involved in the Folding@home research project will be getting some new computing power. Vijay Pande, an associate professor of chemistry, structural biology, and computer science, has worked with a number of companies to design a new 50-node cluster to support a variety of research projects. Nicknamed "numbercruncher," the new system will handle processing for initiatives in computational biology, molecular dynamics, structural biology, and chemical biology simulation, among others.

Pande is one of the scientists behind the distributed computing program, Folding@home, which taps the unused processing power of consumer computers around the world--about 5 million CPUs since 2000--to study protein "folding." Folding is the assembly process the protein goes through before carrying out its specific function--turning into an enzyme or antibody, for example.

The latest cluster of computers uses Xtreme Compute Technologies' a-BriX Tesla graphics processing unit (GPU) servers; Scalable Informatics' host servers with AMD Opterton processors and Delta-V storage; Bright Computing's Cluster Manager; and AccelerEyes' Jacket for MATLAB, a GPU accelerator for high-end computation processing.

"We believe 'numbercruncher' will contribute to our research efforts across all related disciplines and deliver class leading time to discovery and insight resulting in unprecedented scientific breakthroughs," said Pande.