Browse Prior Art Database

The Influence of the Los Alamos and Livermore National Laboratories on the Development of Supercomputing

IP.com Disclosure Number: IPCOM000129677D
Original Publication Date: 1991-Mar-31
Included in the Prior Art Database: 2005-Oct-06

Publishing Venue

Software Patent Institute

Related People

DONALD MACKENZIE: AUTHOR [+2]

Abstract

The Los Alamos and Lawrence Livermore National Laboratories have been important sponsors of, and customers for, supercomputers -- high- performance scientific computers. The laboratories played an important part in establishing speed of floating-point aritmetic (rather than, say, at logical operations) as the performance criterion defining supercomputing. But their more specific influence on the evolution of computer architecture has been limited by the diversity and classified nature of their central computational tasks, together with the expansion of supercomputer use elsewhere.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 3% of the total text.

Page 1 of 29

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

Copyright ©; 1991 by the American Federation of Information Processing Societies, Inc. Used with permission.

The Influence of the Los Alamos and Livermore National Laboratories on the Development of Supercomputing

DONALD MACKENZIE

(Image Omitted: Author's address: Department of Sociology, University of Edinburgh, 18 Buccleuch Place, Edinburgh EH8 9LN, Scotland.)

The Los Alamos and Lawrence Livermore National Laboratories have been important sponsors of, and customers for, supercomputers -- high- performance scientific computers. The laboratories played an important part in establishing speed of floating-point aritmetic (rather than, say, at logical operations) as the performance criterion defining supercomputing. But their more specific influence on the evolution of computer architecture has been limited by the diversity and classified nature of their central computational tasks, together with the expansion of supercomputer use elsewhere.

Categories and Subject Descriptors: K.2 [Computing Milieux]: History of Computing -- hardware, people, software, systems; C.5. 1 [Computer Systems Organization]: Computer System Implementation -- Large and medium computers. General Terms: Design, Documentation, Economics, Experimentation, Management, Performance, Reliability. Additional Terms: supercomputing, high performance computing, LARC, Stretch. Cray, Control Data, Nuclear Weapons Design, SOLOMON, STAR100.

Introduction

Two research concerns underlie this paper. The first is to understand the evolution of computer architectures, particularly parallel architectures. This paper discusses the field of computing in which parallelism has, to date, been most highly developed in practice: high-performance scientific computing, or, as it has come to be called, supercomputing. The second concern is with the role of the large, publicly-funded research and development organizations often called "national laboratories," particularly those whose primary mission lies within defense technology. What influence did these laboratories have on the overall development of science and technology?

The reason for fusing these two concerns is that supercomputing suggests itself as an area of technology in which the influence of national laboratories may have been large. National laboratories, especially the Los Alamos and Lawrence Livermore National Laboratories in the U.S., have been heavily involved in supercomputer development ever since the supercomputer as a distinct class of machine began to emerge in the latter part of the 1950s. What has their influence been? Have they actually shaped the technical evolution of supercomputer architectures?

We begin by reviewing briefly the history of high-performance computing and turn to the nature of the computational tasks involved in nuclear weapons design. The considerable influence of Los Alamos National Laboratory in the early years of digital computing...