The National Science Foundation (NSF) Friday announced a US$53 million project to connect a series of remotely located powerful computers into a high-speed Linux supercomputer grid that could open vast new opportunities for scientific and medical breakthroughs.
The project, to be funded by a three-year grant from the NSF, will be built by the middle of next year, giving scientists and researchers access to massive combined supercomputer power they have until now only dreamed about.
Called the Distributed Terascale Facility, the project will link powerful servers running Linux into a high-speed grid that will allow researchers to use all the computing resources they need, regardless of where the servers are located. At their disposal will be computing power of huge proportions, with a total of 8.1 TFLOPS and the ability to perform 13.6-trillion calculations per second. The grid will have storage of more than 450TB of data through a high-speed optical network called a TeraGrid, which will link computers and data at four academic research facilities in the U.S.
Armonk, N.Y.-based IBM Corp. will provide more than 1,000 IBM eServer Linux clusters that will be running more than 3,300 of Intel Corp.'s upcoming McKinley Itanium processors for the system, as well as IBM data storage products and support services. Qwest Communications International Inc. in Denver will provide a 40-gigabit high-speed network for the grid system, which will be 16 times faster than what is available today.
The supercomputer grid will link the National Center for Supercomputing Applications in Illinois, the San Diego Supercomputer Center in California, the Argonne National Laboratory in suburban Chicago and the California Institute of Technology in Pasadena into a cohesive group of computers with tremendous research potential, according to project organizers.
"This is the first salvo in transforming how science and engineering research is done in the world," said Dan Reed, the director of the National Center for Supercomputing Applications.
The facility is expected to reach peak performance of 13.6 TFLOPS by April 2003, and will be used for a wide range of projects, including research related to storm, climate and earthquake predictions; development of more-efficient combustion engines; chemical and molecular factors in biology; and physical, chemical and electrical properties of materials.
"This facility will stretch the boundaries of high-performance computing and give U.S. computer scientists and other researchers in all science and engineering disciplines access to a critical new resource," said National Science Board Chairman Eamon Kelly.
Eventually, similar grid computing systems are seen as having many uses for business computing, according to proponents.
The announcement is the second related to grid computing this week. On Monday, IBM announced that it's building a worldwide grid computing network to tie together systems at its various data centers to combine their computing power for customers. Users would pay for processing time on an as-needed basis, similar to any other utility. IBM also said it's been chosen by the British government to build a national grid for various universities for collaborative scientific research.
The grid will be run using middleware being built under the open-source Globus Project, a research initiative funded by various U.S. government agencies. Globus software will allow servers and computers to be connected into seamless networks that can be used together to conduct research and other work.
Also involved in the project are cluster computing vendor Myricom Inc., software vendor Oracle Corp. and Sun Microsystems Inc.