The world's most powerful supercomputers are frequently housed in the leading universities and national laboratories that are extending the frontiers of scientific research. Acceleration is one of the new areas of technology that is being actively pursued by the architects of these ground breaking facilities in their quest to provide ever more performance for their users. With energy consumption at today's major supercomputing sites typically exceeding a megawatt and rising towards five megawatts or more, these same visionaries are exploring innovative ways to reduce the economic and environmental impact of their facilities.
Through the vision of Professor Satoshi Matsuoka, the Tokyo Institute of Technology now has the first accelerated industry standard cluster ranked among world's fastest supercomputers.
The TSUBAME cluster has achieved a 47 TeraFLOP Linpack score by adding 360 ClearSpeed Advance boards to its 655 Sun x4600 nodes, a 24% boost over its previous performance. Plugging the Advance boards into the existing nodes increased system performance by 9 TeraFLOPS without the need for additional facilities space and added only one percent to the power requirements.
Tokyo Tech. have estimated that their innovative and record setting architecture will save them over approximately $1 million per year in facilities and energy costs, amounting to a reduction of between 10 and 20 percent of the system's total cost of ownership. Life sciences research using the molecular modeling application AMBER is one of the major consumers of compute resources at Tokyo Tech and one of the first applications planned for ClearSpeed acceleration.
At Bristol University in the United Kingdom, under the direction of Dr. Richard Sessions and Dr. Anthony Clarke, the Bristol Protein Folding Group has developed a generic molecular docking application. Known as Bude, (Bristol University Docking Engine), it is a scientific application that has also benefited from ClearSpeed acceleration.