Parallel Scientific Computation

• Published in: Science (American Association for the Advancement of Science) Vol. 261; no. 5123; pp. 856 - 863
• Main Authors: Hillis, W. Daniel, Boghosian, Bruce M.
• Format: Journal Article
• Language: English
• Published: Washington, DC American Society for the Advancement of Science 13.08.1993; American Association for the Advancement of Science; The American Association for the Advancement of Science
• Subjects: 990200 - Mathematics & Computers; Algorithmics. Computability. Computer arithmetics; ALGORITHMS; Analysis; Applied sciences; Arithmetic; Children; Computational complexity; Computational complexity (Machine theory); Computer memory; Computer programming; Computer science; control theory; systems; COMPUTERS; Connection machine; Connectivity; DIGITAL COMPUTERS; Exact sciences and technology; Fast Fourier transformations; Fortran; GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; Innovations; Libraries; MATHEMATICAL LOGIC; Mathematical optimization; Mathematical programming; Mathematics; Matrices; Optimization theory; Parallel computers; Parallel computing; PARALLEL PROCESSING; PROGRAMMING; Programming Languages; Science; Sequential Approach; SUPERCOMPUTERS; TECHNOLOGY ASSESSMENT; Theoretical computing; Writing Instruction; Parallelism; Parallel computation; Parallel programming
• ISSN: 0036-8075; 1095-9203
• DOI: 10.1126/science.261.5123.856

Massively parallel computers offer scientists a new tool for computation, with capabilities and limitations that are substantially different from those of traditional serial computers. Most categories of large-scale scientific computations have proven remarkably amenable to parallel computation, but often the algorithms involved are different from those used on sequential machines. By surveying a range of examples of parallel scientific computations, this article summarizes our current understanding of the issues of applicability and programming of parallel computers for scientific applications.