Horst D. Simon

Director, NERSC Center and Computational Research Divisions
Lawrence Berkeley National Laboratory

February 24, 4:15pm


The divergence problem is the recognition of the fact that in
recent years scientific computing in America has been handicapped by its
dependence on hardware that is designed and optimized for commercial
applications. The performance of the recently completed Earth Simulator in
Japan, which is five times faster than the fastest American supercomputer,
dramatically exposed the seriousness of this problem. Typical scientific
applications are now able to extract only 5 to 10 percent of the power of
American supercomputers built from commercial web and data servers. By
contrast, the design of the Earth Simulator makes 30 to 50 percent of its
power accessible to the majority of types of scientific calculations.
Reliance on commercial webservers as building blocks for scientific
supercomputers has served the scientific community well for the last eight
years. But events in 2002 demonstrated that the needs of commercial and
scientific applications are diverging.

In this talk I will describe several steps that should be taken in the
U.S. to overcome the reliance on purely commercial approaches to
supercomputing, and to develop science-driven computer architectures.
These steps include new partnerships with vendors, and the development of
projects such as Red Storm (Cray/AMD) and Blue Planet (IBM), which are
based on commodity components, with customization in some key areas.

Reference: "Creating Science-Driven Computer Architecture: A New Path to
Scientific Leadership"  at

Return to Applied Math Colloquium home page