Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Ultracomputers: a teraflop before its time

Ultracomputers: a teraflop before its time Gordon Bell A Teraflop UTAO PTR L R C MU E S Before Its Time are specialized since only highly parallel, coarse-grained applications, requiring algorithm and p r o g r a m d e v e l o p m e n t , can exploit them. Government purchase of such computers would be foolish, since waiting three years will allow computers with a peak speed of a teraflop to be purchased at supercomputer prices ($30 million), due to advancements in semiconductors and the intense competition resulting in "commodity supercomputing." More important, substantially better computers will be available in 1995 in the supercomputer price range if the funding that would be wasted in buying such computers is instead spent on training and software to exploit their power. he quest for the Teraflops Super~puter to operate at ,eak speed of 10" ting- point operations per sec is almost a decade old, and only one three-year computer generation from being fulfilled. The acceleration of its development would require an ultracomputer. First-generation, ultracomputers are networked computers using switches that interconnect thousands of computers to form a multicomputer, and cost $50 to $300 million in 1992. These scalable computers are also http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Communications of the ACM Association for Computing Machinery

Ultracomputers: a teraflop before its time

Communications of the ACM , Volume 35 (8) – Aug 1, 1992

Loading next page...
 
/lp/association-for-computing-machinery/ultracomputers-a-teraflop-before-its-time-yBUqxm0obD

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Association for Computing Machinery
Copyright
Copyright © 1992 by ACM Inc.
ISSN
0001-0782
DOI
10.1145/135226.135227
Publisher site
See Article on Publisher Site

Abstract

Gordon Bell A Teraflop UTAO PTR L R C MU E S Before Its Time are specialized since only highly parallel, coarse-grained applications, requiring algorithm and p r o g r a m d e v e l o p m e n t , can exploit them. Government purchase of such computers would be foolish, since waiting three years will allow computers with a peak speed of a teraflop to be purchased at supercomputer prices ($30 million), due to advancements in semiconductors and the intense competition resulting in "commodity supercomputing." More important, substantially better computers will be available in 1995 in the supercomputer price range if the funding that would be wasted in buying such computers is instead spent on training and software to exploit their power. he quest for the Teraflops Super~puter to operate at ,eak speed of 10" ting- point operations per sec is almost a decade old, and only one three-year computer generation from being fulfilled. The acceleration of its development would require an ultracomputer. First-generation, ultracomputers are networked computers using switches that interconnect thousands of computers to form a multicomputer, and cost $50 to $300 million in 1992. These scalable computers are also

Journal

Communications of the ACMAssociation for Computing Machinery

Published: Aug 1, 1992

There are no references for this article.