Hearing Summary - House Science Committee Holds Hearing on U.S. Leadership in Supercomputing
May 13, 2004
The House Committee on Science held a hearing on May 13, 2004, to examine federal high-performance computing (HPC) research and development activities and to consider H.R. 4218, the High-Performance Computing Revitalization Act of 2004, which would amend the High-Performance Computing Act of 1991.
In particular, Committee Members expressed concern regarding the effects of HPC on U.S. global competitiveness; current efforts of federal civilian science agencies [e.g., National Science Foundation (NSF) and Department of Energy] to ensure U.S. global leadership; and future targets (e.g., industrial sectors) of HPC resources.
Members of both parties of the House Science Committee agreed that HPC is an essential component of U.S. scientific, industrial, and military competitiveness. Chairman Sherwood Boehlert (R-NY), in his opening remarks said, "the Japanese Earth Simulator was a wake-up call that our leadership is being challenged and that we perhaps had put too many eggs in pursuing computer architectures with commercial applications."
Witnesses addressing this and the aforementioned concerns before the Committee included:
- Dr. John H. Marburger III - Director, White House Office of Science and Technology Policy.
- Dr. Irving Wladawsky-Berger - Vice President, Technology and Strategy for IBM Corporation.
- Dr. Rick Stevens - Director, Mathematics and Computer Science Division, Argonne National Laboratory; Director, NSF TeraGrid project.
- Dr. Daniel Reed - William R. Kenan, Jr., Eminent Professor, University of North Carolina.
The witnesses agreed that an integrated, adequately funded supercomputing effort among federal agencies is paramount for the U.S. to maintain international leadership and to catalyze new discoveries in many research areas, especially biomedical research. The witnesses reiterated the need to integrate HPC efforts across federal agencies, including a more active role by NIH. In addition, the witnesses concurred that supercomputers should be treated as natural resources, with access open and available to a greater portion of the research community. Changing HPC measurement criteria from benchmarking or peak performance to productivity based on real applications and solving problem will help reshape federal investments as well as invigorate discovery based on novel application-specific systems.
Dr. Marburger concurred with the need for an integrated and strategic HPC research and education plan for the U.S. and advised of the Administration's full support for H.R. 4218 in its current form. Marburger highlighted a May 10, 2004, report entitled, "Federal Plan for High-End Computing," issued by the High-End Computing Revitalization Task Force. The report lays out detailed roadmaps for investments in key research and development areas, including hardware, software, and systems. Marburger stated that streamlining federal investments as well as providing broad accessibility would mandate that relevant agencies be engaged at the highest levels. This level of engagement would require that all agencies be at the same level of awareness and that barriers of resource flow across agencies be reduced. As part of the strategic vision, the integration of NSF and DOE HPC facilities, infrastructure, and activities will be explored further. A renewed focus on HPC applications and solving real-life problems will also invigorate the federal HPC effort.