Impact of self-scheduling order on performance of multiprocessor systems

Peiyi Tang; Pen Chung Yew; Chuan Qi Zhu

doi:10.1145/55364.55422

Impact of self-scheduling order on performance of multiprocessor systems

Peiyi Tang, Pen Chung Yew, Chuan Qi Zhu

Computer Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

13 Scopus citations

Abstract

Processor self-scheduling is an efficient dynamic scheduling for multiprocessors. This paper discusses the impact of the self-scheduling order on the performance of multiply-nested parallel loops. It is shown that, due to data synchronization for cross-iteration data dependences, the completion time of a multiply-nested loop is reduced when the nesting parallel loops with smaller delays are moved to the inside. The best performance is achieved when a shortest-delay scheduling order is used. The performance of the shortest-delay self-scheduling is compared to other self-scheduling orders and to compile-time static scheduling order proposed elsewhere. Program transformation needed to implement shortest-delay self-scheduling is also included.

Original language	English (US)
Title of host publication	Proceedings of the 2nd International Conference on Supercomputing, ICS 1988
Editors	J. Lenfant
Publisher	Association for Computing Machinery
Pages	593-603
Number of pages	11
ISBN (Electronic)	0897912721
DOIs	https://doi.org/10.1145/55364.55422
State	Published - Jun 1 1988
Event	2nd International Conference on Supercomputing, ICS 1988 - St. Malo, France Duration: Jul 4 1988 → Jul 8 1988

Publication series

Name	Proceedings of the International Conference on Supercomputing
Volume	Part F130184

Other

Other	2nd International Conference on Supercomputing, ICS 1988
Country/Territory	France
City	St. Malo
Period	7/4/88 → 7/8/88

Bibliographical note

Publisher Copyright:
© 1988 ACM.

Publisher link

10.1145/55364.55422

Find It

Find It at U of M Libraries

Cite this

Tang, P., Yew, P. C., & Zhu, C. Q. (1988). Impact of self-scheduling order on performance of multiprocessor systems. In J. Lenfant (Ed.), Proceedings of the 2nd International Conference on Supercomputing, ICS 1988 (pp. 593-603). (Proceedings of the International Conference on Supercomputing; Vol. Part F130184). Association for Computing Machinery. https://doi.org/10.1145/55364.55422

Impact of self-scheduling order on performance of multiprocessor systems. / Tang, Peiyi; Yew, Pen Chung; Zhu, Chuan Qi.
Proceedings of the 2nd International Conference on Supercomputing, ICS 1988. ed. / J. Lenfant. Association for Computing Machinery, 1988. p. 593-603 (Proceedings of the International Conference on Supercomputing; Vol. Part F130184).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Tang, P, Yew, PC & Zhu, CQ 1988, Impact of self-scheduling order on performance of multiprocessor systems. in J Lenfant (ed.), Proceedings of the 2nd International Conference on Supercomputing, ICS 1988. Proceedings of the International Conference on Supercomputing, vol. Part F130184, Association for Computing Machinery, pp. 593-603, 2nd International Conference on Supercomputing, ICS 1988, St. Malo, France, 7/4/88. https://doi.org/10.1145/55364.55422

@inproceedings{c2e2e9ba1d474f5cb105a844f1f1b6f0,

title = "Impact of self-scheduling order on performance of multiprocessor systems",

abstract = "Processor self-scheduling is an efficient dynamic scheduling for multiprocessors. This paper discusses the impact of the self-scheduling order on the performance of multiply-nested parallel loops. It is shown that, due to data synchronization for cross-iteration data dependences, the completion time of a multiply-nested loop is reduced when the nesting parallel loops with smaller delays are moved to the inside. The best performance is achieved when a shortest-delay scheduling order is used. The performance of the shortest-delay self-scheduling is compared to other self-scheduling orders and to compile-time static scheduling order proposed elsewhere. Program transformation needed to implement shortest-delay self-scheduling is also included.",

author = "Peiyi Tang and Yew, {Pen Chung} and Zhu, {Chuan Qi}",

note = "Publisher Copyright: {\textcopyright} 1988 ACM.; 2nd International Conference on Supercomputing, ICS 1988 ; Conference date: 04-07-1988 Through 08-07-1988",

year = "1988",

month = jun,

day = "1",

doi = "10.1145/55364.55422",

language = "English (US)",

series = "Proceedings of the International Conference on Supercomputing",

publisher = "Association for Computing Machinery",

pages = "593--603",

editor = "J. Lenfant",

booktitle = "Proceedings of the 2nd International Conference on Supercomputing, ICS 1988",

}

TY - GEN

T1 - Impact of self-scheduling order on performance of multiprocessor systems

AU - Tang, Peiyi

AU - Yew, Pen Chung

AU - Zhu, Chuan Qi

PY - 1988/6/1

Y1 - 1988/6/1

N2 - Processor self-scheduling is an efficient dynamic scheduling for multiprocessors. This paper discusses the impact of the self-scheduling order on the performance of multiply-nested parallel loops. It is shown that, due to data synchronization for cross-iteration data dependences, the completion time of a multiply-nested loop is reduced when the nesting parallel loops with smaller delays are moved to the inside. The best performance is achieved when a shortest-delay scheduling order is used. The performance of the shortest-delay self-scheduling is compared to other self-scheduling orders and to compile-time static scheduling order proposed elsewhere. Program transformation needed to implement shortest-delay self-scheduling is also included.

AB - Processor self-scheduling is an efficient dynamic scheduling for multiprocessors. This paper discusses the impact of the self-scheduling order on the performance of multiply-nested parallel loops. It is shown that, due to data synchronization for cross-iteration data dependences, the completion time of a multiply-nested loop is reduced when the nesting parallel loops with smaller delays are moved to the inside. The best performance is achieved when a shortest-delay scheduling order is used. The performance of the shortest-delay self-scheduling is compared to other self-scheduling orders and to compile-time static scheduling order proposed elsewhere. Program transformation needed to implement shortest-delay self-scheduling is also included.

UR - http://www.scopus.com/inward/record.url?scp=84943734639&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84943734639&partnerID=8YFLogxK

U2 - 10.1145/55364.55422

DO - 10.1145/55364.55422

M3 - Conference contribution

AN - SCOPUS:84943734639

T3 - Proceedings of the International Conference on Supercomputing

SP - 593

EP - 603

BT - Proceedings of the 2nd International Conference on Supercomputing, ICS 1988

A2 - Lenfant, J.

PB - Association for Computing Machinery

T2 - 2nd International Conference on Supercomputing, ICS 1988

Y2 - 4 July 1988 through 8 July 1988

ER -

Impact of self-scheduling order on performance of multiprocessor systems

Abstract

Publication series

Other

Bibliographical note

Publisher link

Find It

Other files and links

Fingerprint

Cite this