Visual-computing solutions provider NVIDIA sees computational biology as "one of the lowest hanging fruits" that can benefit from application acceleration via graphics processing units. Performance increases on the order of 10 to 100 times that of a central processing unit are "fairly typical," the company says in an industry magazine.
Another GPU plus: Cutting the cost of high-performance computing, which is especially problematic for those who need it most--namely small and mid-sized biotech researchers. Graphics hardware and software yield processing performance increases in standard PCs.
NVIDIA recently unveiled the Tesla Bio Workbench, which comprises GPU-optimized bioscience applications for use by researchers studying molecular dynamics and quantum chemistry. The Workbench offering includes access to a community site for downloading applications. The site also provides benchmark data, academic papers and tutorials, and discussion forums.
The GPU approach allows small-scale simulations on GPU workstations and large-scale simulations via GPU workstation clusters and code scaling. Such a technique lets researchers simulate large molecules without the need for supercomputer time.