Note to users. If you're seeing this message, it means that your browser cannot find this page's style/presentation instructions -- or possibly that you are using a browser that does not support current Web standards. Find out more about why this message is appearing, and what you can do to make your experience of our site the best it can be.


Science 28 February 1986:
Vol. 231. no. 4741, pp. 963 - 967
DOI: 10.1126/science.231.4741.963

Articles

Small Shared-Memory Multiprocessors

FOREST BASKETT 1 and JOHN L. HENNESSY 2

1 Director, Western Research Laboratory, Digital Equipment Corporation, 100 Hamilton Avenue, Palo Alto, CA 94301.
2 Associate professor, Departments of Electrical Engineering and Computer Science, Stanford University, Stanford, CA 94305.

Multiprocessors built from today's microprocessors are economically attractive. Although we can use these multiprocessors for time-sharing applications, it would be preferable to use them as true parallel processors. One key to achieving efficient parallel processing is to match the communications capabilities of the multiprocessor to the communications needs of the problem. The other key is improved parallel programming systems. If these are achieved, then efficient parallel processing can be approached from both ends by providing more communications capability in the hardware and restructuring the problem to reduce the communications requirements.





To Advertise     Find Products


Science. ISSN 0036-8075 (print), 1095-9203 (online)