|
A forthcoming report from the Computer Science and Technology Board of the National Research Council titled “The Future of Supercomputing” concludes that the needs of some high-end defense and government users are diverging from the "cluster" approaches to high performance computing dominant in most scientific and commercial computing applications. The widening gap between government and civilian uses for supercomputers raises important questions for U.S. national security policy, which has become increasingly dependent on advances in supercomputer technology. How rapidly are supercomputing needs for specialized national security applications like nuclear weapons design and testing, cryptography, and aeronautics diverging from commercial and scientific applications in applied physics, biotechnology and genomics, climatological modeling, and engineering? What are the economic and national security consequences of the increasingly divergent needs of government and mainstream commercial customers for the fastest computers? Would such an increasing separation work against the technological advantages of keeping our national security computing platforms harnessed to the rapid technological innovation in commercial computing systems? Is the widespread availability of "commodity" supercomputers applicable to military use in emerging powers such as China and India likely to alter the balance of qualitative military power? “Of Crays and Clusters: The Future of U.S. Supercomputing,” sponsored by the University of Texas Global Challenges Initiative in cooperation with the LBJ School of Public Affairs and the Department of Computer Science, will examine the security, economic, and scientific implications of the emerging separation between government/defense and civilian supercomputer applications. |
January 27, 2005 Schedule: Moderator: Panelists: Kenneth S. Flamm, bio
Stephen W. Keckler, bio
Marc Snir, bio
|