previous | contents | next

Chapter 37 A survey of problems and preliminary results concerning parallel processing and parallel processors 469

factor. Such a factor is intuitive and environment-sensitive, depending on the relative concern for speed and for costs of various sorts. For the present data we have chosen to display a function:

where K is a constant, throughput time a measure of the speed of computation, and total processor time a measure of the cost.

8. Conclusion

In this paper we have presented some thoughts on parallel processing. In particular we have chosen to survey the topic by including an extensive bibliography and some of the results of our work in this area. The discussion has had to be brief, but our intention has been to convey the picture of the potential that parallel processing systems offer for the future development of computing.

The key to successful exploitation lies in a new, unified, and scientific approach to the entire problem of the design and usage of computing systems. The development of large, integrated systems raises many problems, but there can be no doubt that economic solutions to these will be found. Their development should comprise a significant part of the computer system architectural design effort of the next few years.

Any ultimate evaluation of a parallel processing system within a working environment depends on actual operating experience. This in turn requires the existence of a system and the interest of users. Only when usable systems become available will the concept of parallel processing in integrated systems be accurately evaluated.

References

BlaaG64; BrigH64; ConwM63; CorbF65; DennJ66; DesmW64; DreyP58; FalkA64; GillS58; GregJ63; KatxJ66; LehmM65; LeinA59; McCuJ65; MiraW67; NievJ64; RoseJ65; SchlH??; ShedG66a, b; SlotD62; SmitR64; PL/I Language Specification, FormC28-6571

Bibliography

Al1eM63; AmdaG62; AndeJ62, 65; ArdeB66; BaldF62; BlaaG64; BrigH64; BuchW62; BussB63; CoddE62; ComfW65; ConwM63; CorbF62, 65; CritA63; DaleR65; DennJ65, 66; DesmW64; DijkE65; DreyP58; ErnsH63; EstrG60, 63; EwinR64; FalkA64; ForgJ65; FranJ57; GillS58; GlasE65; GregJ63; HellH61, 66; KatzJ66; KinsH64; KnutD66; LehmM63a, 63b, 65; LeinA59; LourN59; MarcM63; McCaJ62; McCuJ65; MeadR63; MillW63; MiraW67; NievJ64; OssaJ6S; PennJ62; RoseJ65; SchlH??; SeebR63; SenzD65; ShedG66a, 66b; SlotD62; SmitR64; SquiJ63; StraC59; VyssV65; WirtN66; IBM OS/360 PL/I Language Specification, Form C 28-6571; Proc. 1F1P1962, "Symposium on Multi-Programming" 1963.

previous | contents | next