Microsoft is funding a Microsoft Institute for High-Performance Computing at Cornell with annual funding of $400,000, renewable each year for an indefinite period. The new institute will greatly expand the ability of researchers to work with huge databases of DNA sequences and protein composition and shapes, and explore new software and applications for the analysis of biological information.
The institute will exist within the Computational Biology Service Unit (CBSU) at Cornell Theory Center (CTC). Its work will focus on the development of new software and applications in the Microsoft high-performance computing environment.
Ron Elber, professor of computer science and director of CBSU, also will be director of the new institute, and it will be managed by CTC senior research associate Jaroslaw Pillardy, who manages CBSU. "The support of Microsoft is very much appreciated," Elber said. "They have supported us before, but this increased support makes it possible for the CBSU to expand and to do more interesting things and larger things based on new software technology from Microsoft."
Computational biology embraces the use of computers to simulate biological processes, such as the folding of proteins or the identification of gene function, as well as the management of large databases of biological information, an area known as bioinformatics. Biologists have created huge databases of DNA sequences and protein composition and shapes. To compare a sample with everything in these databases on an ordinary computer could take about a month, according to Pillardy, but a high-performance computer cluster like those operated by CTC can do it in minutes.
CBSU has created Web interfaces that allow researchers anywhere in the world to perform complex tasks remotely and also offers consulting services. "If you don't have experience in computational biology, talk to us and we will help you," Pillardy says.
At CBSU he works with three senior researchers, a postdoctoral research assistant and a programmer. The Microsoft support will enable the hiring of another senior researcher and expansion of the cluster-computing facilities. CBSU currently operates two dedicated clusters with a total of 512 parallel processors. About 60 more dual-processor nodes will be added.
Microsoft and CTC have a long history of collaboration. Beginning in 2000, Microsoft partnered with CTC to develop Microsoft's capabilities in high-performance computing; eventually CTC developed a Windows-based cluster computing environment consisting of more than 1,500 processors. CTC consultants have helped develop cluster computing facilities for several business and academic users.
CBSU was created as part of the Tri-Institutional collaboration between biomedical and clinical researchers at Cornell, Rockefeller University and Memorial Sloan-Kettering Cancer Center in New York City, and has become a core resource for biology research throughout the Ithaca and New York campuses. It currently supports about 40 Cornell researchers. "We can give advanced computational services to biologists at significantly reduced cost, compared to what a single group can do," Pillardy said. "No single group would be able to afford these computational resources."