ENGINEERING

U.S. needs for supercomputing to strengthen defense and national security cannot be satisfied with current policies and levels of spending, says a report from the National Academies' National Research Council. -- The federal government should provide stable, long-term funding and support multiple supercomputing hardware and software vendors in order to give scientists and policy-makers better tools to solve problems in areas such as intelligence, nuclear stockpile stewardship, and climate change, said the committee that wrote the report. "For the past decade, insufficient government funding for supercomputing, little long-term planning, and inadequate coordination among relevant federal agencies have reduced opportunities to make the most of this technology to improve national security and promote scientific discovery," said Susan Graham, committee co-chair and professor of computer science, University of California, Berkeley. "U.S. leadership in supercomputing technologies is eroding, and with it the broader scientific and technological progress underlying a strong and robust U.S. economy." Investment decisions regarding supercomputing research and development should not be based on whether the United States possesses the fastest supercomputer in the world, the committee said. Instead, the government should make long-term plans to secure U.S. leadership in the hardware, software, and other technologies that are essential to national defense and scientific research. "Whether the most powerful and most expensive supercomputer is located in the United States or elsewhere does not indicate loss or gain of leadership in supercomputing technology," said Marc Snir, committee co-chair and professor of computer science, University of Illinois, Urbana-Champaign. "Rather, our concern is that current investments and plans are not sufficient to provide the capabilities that our country needs." By providing stable, sustained investments in supercomputing research and development, the government would ensure stability in the marketplace and allow companies in supercomputing to stay viable and grow. Such stability is critical because many of the machines needed for the most demanding supercomputing applications -- particularly for national security -- must be custom-made, the committee said. The U.S. government also should support the creation and maintenance of supercomputing software, the committee added. The supercomputing community should create a road map defining future goals in supercomputing research and development, the report says. To meet these goals, the relevant government agencies should develop a single, integrated plan that would allow them to identify common needs at an early stage, leverage resources, and minimize duplicative efforts, the report adds. The agencies and Congress would then use the plan to coordinate annual investments and track progress. "Steady investments over time and clear milestones for the future should provide returns on investment that greatly exceed the original costs," said Graham. While the committee emphasized the need for greater U.S. leadership in the development of supercomputing technologies, it also called for research collaboration between agencies here and international organizations. Foreign partners should have access to U.S. supercomputing systems, the committee said. The study was sponsored by the U.S. Department of Energy. The National Research Council is the principal operating arm of the National Academy of Sciences and the National Academy of Engineering. It is a private, nonprofit institution that provides science and technology advice under a congressional charter. Copies of Getting Up to Speed: The Future of Supercomputing will be available from the National Academies Press; tel. 202-334-3313 or 1-800-624-6242 or order on the Internet at www.nap.edu .