Monday, July 14

10:30 AM-12:30 PM
Law School, Room 290

LINPACK and Its Impact on High Performance Computing

(Sponsored by SIAM Activity Group on Supercomputing)

The intent of this minisymposium is to give some historical perspective on the impact LINPACK had on the development of high-performance computing. Fifteen years ago, at the 30th SIAM Annual Meeting, also at Stanford, SIAM introduced the concept of "minisymposia" at its conferences. One of the first such minisymposia featured a discussion of the performance of linear algebra algorithms on early vector computers such as the Cray-1S. Many of the key developments of the last 15 years in high-performance computing can be traced back directly to this minisymposium: the development of fast linear algebra algorithms for high-performance computing architectures; the use of LINPACK as a benchmark for new computer architectures; the development of high-quality mathematical software, which tracked the architectural evolution of high-performance architectures (BLAS2 and BLAS3, LAPACK, SCALAPACK). In the current minisymposium we will provide a perspective on these developments, and present survey talks on the current state of the art in these areas.

Organizer: Horst D. Simon
Lawrence Berkeley National Laboratory

10:30 LINPACK and Its Impact on High-Performance Computing
Horst D. Simon, Organizer
11:00 Linear Algebra Libraries for High-Performance Computers
Jack Dongarra, University of Tennessee, Knoxville and Oak Ridge National Laboratory
11:30 Benchmarking for High-Performance Computing - Current State of the Art
David H. Bailey, NASA Ames Research Center
12:00 Linear Algebra Algorithms for High-Performance Computing - Current State of the Art
James Demmel, University of California, Berkeley

AN97 Homepage | Program Updates|
Registration | Hotel and Dormitory Information | Transportation | Program-at-a-Glance | Program Overview

MMD, 4/4/97 tjf, 5/27/97