U.S. making supercomputing push
- By Randall Edwards
- Aug 25, 2003
An interagency task force established to maintain U.S. leadership in supercomputing is expected to unveil a road map for improving research and development and guiding future investments in high-end computing systems later this month.
The High End Computing Revitalization Task Force, managed by the National Coordination Office for Information Technology Research and Development, has been working since March to develop a plan for improving high-end computing technology crucial for advanced scientific research and national security.
Supercomputers are used for specialized applications that require massive amounts of mathematical calculations such as weather forecasting, nuclear energy research and petroleum exploration.
The U.S. supercomputing community received a wake-up call in April 2002 when Japan unveiled the powerful Earth Simulator supercomputer. It has motivated the U.S. community to reclaim high-end computing superiority.
The shockingly fast Earth Simulator, capable of sustaining an operating speed of 35 trillion calculations per second, gives Japan an edge in a traditionally American-dominated field. The top-rated U.S. supercomputer, the ASCI Q, located at Los Alamos National Laboratory in New Mexico, operates at roughly half that speed, according to a House Science Committee report published July 16.
After observing Japan's accomplishment, several government officials and computing experts have pushed for a joint effort among federal agencies.
"Maintaining U.S. leadership requires a coordinated, concerted effort by the federal government," said House Science Committee Chairman Sherwood Boehlert (R-N.Y.) at the July 16 hearing. "Let me stress that: a coordinated, concerted effort by the federal government."
"If we really want to get the lead, it's going to require a diverse set of architectures targeted to a set of certain problems," said Ed Oliver, associate director for Advanced Scientific Computing Research in the Energy Department's Office of Science. "Maybe our approach will be encouraging diversity and making wise decisions about how we buy the machines."
Many in the high-end computing field, however, feel that a U.S. push for superiority cannot be made without increased funding.
"There is a great opportunity ahead of us, and to be able to exploit those opportunities will take a lot more funding," said Peter Freeman, associate director for computer and information science and engineering at the National Science Foundation.
Despite the urge for more funding, several experts feel the United States is not far behind Japan in terms of what can be accomplished through high-end computing and that the technology and power of U.S. machines will soon be equal to the Earth Simulator's.