U.S. National Cancer Institute purchases Starbridge Hypercomputer to accelerate genome data research

Salt Lake City 10 May 2004The National Cancer Institute's Advanced Biomedical Computing Center (ABCC) in Frederick, Maryland, recently purchased a Starbridge Hypercomputer, the first high-end reconfigurable computer acquired for bioscience research. The Starbridge Hypercomputer will serve as an important scientific computational resource and represents a revolutionary leap in high performance computing capabilities.

Advertisement

Starbridge consulted with Dr. Jack Collins of the National Cancer Institute to assess NCI's research needs, and produced a massively parallel and scalable implementation of the Smith Waterman algorithm, which is used to compare sets of genetic data, on the Starbridge Hypercomputer. Starbridge's implementation of the Smith Waterman algorithm enables NCI to compare genetic data sets of virtually any size in record-breaking time. As part of the final acceptance criteria, Starbridge demonstrated a Smith Waterman comparison of the human X to Y chromosomes. This comparison, which typically requires months of computation time on a cluster, was completed in approximately five days on the Hypercomputer.

"NCI's purchase illustrates the potential for Hypercomputing to accelerate research in the Life Sciences", stated Daniel Oswald, President and CEO of Starbridge Systems. "We expect this to be the first of many grand challenge problems solved by Starbridge Hypercomputers."

Utah-based Starbridge Systems was founded in 1998 to advance reconfigurable computing technologies, creating the powerful, high-performance Hypercomputer platforms. Compared with conventional processors, Starbridge's scalable Hypercomputers offer significant advantages in speed, flexibility, versatility and efficiencies of size, weight and power consumption. Starbridge's Viva development environment turns months into minutes by supporting highly parallel expression of algorithms and optimally configuring the Hypercomputer hardware for peak execution of each algorithm.


Leslie Versweyveld

[Medical IT News][Calendar][Virtual Medical Worlds Community][News on Advanced IT]