Dongarra's TOP500 List of World's Fastest Supercomputers Released at Mannheim Conference

CRPC researcher Jack Dongarra of the University of Tennessee and Oak Ridge National Laboratory is one of three renowned computer scientists who assemble the legendary TOP500 List of the world's fastest supercomputers. Released twice a year since 1993, the list features sites with the most powerful computer systems, determined with information from a questionnaire sent to high-performance computer (HPC) experts, computational scientists, manufacturers, and the Internet community at large. Dongarra, Hans Meuer of the University of Mannheim, and Erich Strohmaier of the University of Tennessee released their June 1999 TOP500 List at the 14th Mannheim Supercomputing Conference and Seminar, held June 10-12 in Mannheim, Germany.

CRPC's Jack Dongarra
According to Dongarra, the TOP500 List shows some remarkable changes for the first half of 1999. "The list continues to change rapidly in all ranges of performance," he says. "Seventy systems exceed the mark of 100 gigaflops per second [Gflop)/s] this time. The last entry of the TOP500 list shows a performance rate of 24.73 GFlop/s compared to 17.12 GFlop/s in November 1998. Nearly 200 systems fell off the list in the last six months because of this increase in entry-level performance. The total accumulated performance of all 500 systems grew from 29.6 teraflops per second [TFlop/s] in November 1998 to 39.1 TFlop/s in June 1999."
The LINPACK benchmark is used as a performance measure to rank the computers. Developed by Dongarra, LINPACK was chosen because it is widely used and performance numbers are available for almost all relevant systems. For the TOP500, the researchers use a version of the benchmark that allows the user to scale the size of the problem and to optimize the software to achieve the best performance for a given machine.

The top 10 positions of the TOP500 List are taken by three Department of Energy ASCI machines, six Silicon Graphics Inc. (SGI) T3E systems, and one new Hitachi SR8000 system at the University of Tokyo. The performance level necessary to be listed in one of the first 10 positions has shifted to more than 1/2 Tflop/s.

The upgraded Intel ASCI Red system at Sandia National Laboratory, with achieved performance of more than 2.1 Tflop/s using 9,472 Pentium processors, remains the number one computer. The ASCI Blue Mountain system at Los Alamos National Laboratory is number two, with 1.6 TFlop/s. These are the only two systems currently exceeding the 1 TFlop/s level on the LINPACK benchmark. However, the TOP500 already contains seven systems exceeding 1 TFlop/s peak performance.

SGI again had the most machines on the list, with 182 systems (36.4%) and 48% of the installed LINPACK performance. International Business Machines (IBM) is back in second position with 118 systems, followed by Sun Microsystems with 95 systems.

The number of industrial and commercial sites listed in the TOP500 continues to rise strongly. There now are 241 systems, almost half of the list. Sun and IBM (78 systems each), SGI (50 systems), and Hewlett-Packard (30 systems) account for almost all of these sites.

HPCwire, an online computer technology publication based in San Diego, California, offered a video-on-demand presentation from its Web site during the Mannheim conference. Two on-site cameras recorded the event, and feeds were digitized and transferred to the Mannheim computer center through a wireless connection to a Silicon Graphics Inc. (SGI) server. Footage was available on the Internet showing chief technology officers from nine major HPC manufacturers in the United States and Japan addressing the most current architectural, technological, and market trends for all types of HPC systems. The video is archived at the HPCwire Web site,

For the list of the 1999 TOP500 supercomputers and sites, as well as other information about TOP500, see For information on LINPACK, see

Other Issues of PCR Back to PCR CRPC Home Page