Parallel computer architectures
Document Type
Article
Publication Date
9-1986
Department
Department of Mathematical Sciences; Department of Computer Science
Abstract
Parallel processing is becoming a dominant way in which very high performance is being achieved in modern supercomputer systems. It is therefore becoming increasingly important that scientists and engineers know how supercomputers achieve parallelism in order to take advantage of the computers enormous problem solving capability. A computer solution to a problem must often be expressed so that the parallelism provided by the machine is reflected in the implementation of the program. Otherwise the system will not perform at even a fraction of its potential speed since the computing resources will not be used efficiently to solve the problem. In this paper, several ways in which computers are organized to achieve parallelism will be described. The descriptions are primarily conceptual and would be a useful introduction for someone wishing to make effective use of a machine that uses one or more of the parallel processing techniques.
Publication Title
Applied Mathematics and Computation
Recommended Citation
Poplawski, D.
(1986).
Parallel computer architectures.
Applied Mathematics and Computation,
20(1-2), 41-51.
http://doi.org/10.1016/0096-3003(86)90123-2
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/5567