Browse Prior Art Database

Exploiting Program Concurrency in Computing Systems

IP.com Disclosure Number: IPCOM000131367D
Original Publication Date: 1979-Jan-01
Included in the Prior Art Database: 2005-Nov-10
Document File: 13 page(s) / 47K

Publishing Venue

Software Patent Institute

Related People

Philip C. Treleaven: AUTHOR [+3]

Abstract

University of Newcastle upon Tyne [Figure containing following caption omitted: The solution of many large computational problems is an inherently concurrent process. But most existing computing systems are inherently sequential. This mismatch may be overcome by data-flow computing] Computing systems based on a data-flow organization of computation are presently under development~ 2 in laboratories in the US and Europe. These systems are a fundamentally new style of concurrent (tightly coupled, distributed) computer, which could eventually supersede the conventional general-purpose (von Neumann) computer. Conventional computers and programs, even those capable of concurrent activity, are implicitly sequential and require any desired concurrency to be indicated explicitly. They are control-flow computers, and they can be contrasted with data-flow computers, which allow concurrent operations to be activated as soon as their input data are available. The search for computing systems that can utilize the natural concurrency in problems is motivated by at least two interlinked requirements: the need to increase performance, and the need to better utilize system resources. In the case of performance, for example, there are still problems, such as wind-tunnel and weather modeling,3 that cannot be simulated effectively by conventional computers.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 8% of the total text.

Page 1 of 13

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

This record contains textual material that is copyright ©; 1979 by the Institute of Electrical and Electronics Engineers, Inc. All rights reserved. Contact the IEEE Computer Society http://www.computer.org/ (714-821-8380) for copies of the complete work that was the source of this textual material and for all use beyond that as a record from the SPI Database.

Exploiting Program Concurrency in Computing Systems

Philip C. Treleaven

University of Newcastle upon Tyne

(Image Omitted: The solution of many large computational problems is an inherently concurrent process. But most existing computing systems are inherently sequential. This mismatch may be overcome by data-flow computing)

Computing systems based on a data-flow organization of computation are presently under development~ 2 in laboratories in the US and Europe. These systems are a fundamentally new style of concurrent (tightly coupled, distributed) computer, which could eventually supersede the conventional general-purpose (von Neumann) computer. Conventional computers and programs, even those capable of concurrent activity, are implicitly sequential and require any desired concurrency to be indicated explicitly. They are control-flow computers, and they can be contrasted with data-flow computers, which allow concurrent operations to be activated as soon as their input data are available.

The search for computing systems that can utilize the natural concurrency in problems is motivated by at least two interlinked requirements: the need to increase performance, and the need to better utilize system resources. In the case of performance, for example, there are still problems, such as wind-tunnel and weather modeling,3 that cannot be simulated effectively by conventional computers.

In the past, progress in hardware technology has tended to be the driving force behind performance increases. The stage is now being reached, however, at which physical constraints, including heat dissipation and basic physical laws, limit performance improvement via progress in hardware.4 Therefore, organizational advances in the design of computers, which in the past have included "pipelining"5 the processing unit and designing hardware to support high-level languages,6 7 are likely to become increasingly important. However, while progress in hardware technology is limited by basic physical laws, so organizational variations are also limited if programs must continue to be executed in a manner that is basically sequential.

In principle, the concurrency that exists in problems can be exploited by either (1) executing a number of distinct programs in parallel or (2) utilizing the natural structure of individual, highly concurrent programs. Concurrent computation should overcome the inherent speed limitations of sequential computation by allowing a number of operations within a system to take place asynchronously, increasing performance b...