Analysis of randomized work stealing with false sharing

Richard Cole, Vijaya Ramachandran

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper analyzes the overhead due to false sharing when parallel tasks are scheduled using randomized work stealing (RWS). We obtain high-probability bounds on the cache miss overhead, including the overhead due to false sharing, for several parallel cache-efficient algorithms when scheduled using RWS. These include algorithms for fundamental problems, such as matrix computations, FFT, sorting, basic dynamic programming, list ranking and graph connected components. Our main technical contribution, from which these results follow, is the derivation of nontrivial high-probability bounds on the number of steals incurred by these algorithms in the presence of false sharing, when using RWS.

Original languageEnglish (US)
Title of host publicationProceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013
Pages985-998
Number of pages14
DOIs
StatePublished - 2013
Event27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013 - Boston, MA, United States
Duration: May 20 2013May 24 2013

Other

Other27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013
CountryUnited States
CityBoston, MA
Period5/20/135/24/13

Fingerprint

Dynamic programming
Sorting
Fast Fourier transforms

Keywords

  • false sharing
  • performance analysis
  • Randomized work stealing

ASJC Scopus subject areas

  • Software

Cite this

Cole, R., & Ramachandran, V. (2013). Analysis of randomized work stealing with false sharing. In Proceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013 (pp. 985-998). [6569879] https://doi.org/10.1109/IPDPS.2013.86

Analysis of randomized work stealing with false sharing. / Cole, Richard; Ramachandran, Vijaya.

Proceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013. 2013. p. 985-998 6569879.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Cole, R & Ramachandran, V 2013, Analysis of randomized work stealing with false sharing. in Proceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013., 6569879, pp. 985-998, 27th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2013, Boston, MA, United States, 5/20/13. https://doi.org/10.1109/IPDPS.2013.86
Cole R, Ramachandran V. Analysis of randomized work stealing with false sharing. In Proceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013. 2013. p. 985-998. 6569879 https://doi.org/10.1109/IPDPS.2013.86
Cole, Richard ; Ramachandran, Vijaya. / Analysis of randomized work stealing with false sharing. Proceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013. 2013. pp. 985-998
@inproceedings{3dd2d3e7e19d419fa0ec7e4c6b0296ca,
title = "Analysis of randomized work stealing with false sharing",
abstract = "This paper analyzes the overhead due to false sharing when parallel tasks are scheduled using randomized work stealing (RWS). We obtain high-probability bounds on the cache miss overhead, including the overhead due to false sharing, for several parallel cache-efficient algorithms when scheduled using RWS. These include algorithms for fundamental problems, such as matrix computations, FFT, sorting, basic dynamic programming, list ranking and graph connected components. Our main technical contribution, from which these results follow, is the derivation of nontrivial high-probability bounds on the number of steals incurred by these algorithms in the presence of false sharing, when using RWS.",
keywords = "false sharing, performance analysis, Randomized work stealing",
author = "Richard Cole and Vijaya Ramachandran",
year = "2013",
doi = "10.1109/IPDPS.2013.86",
language = "English (US)",
pages = "985--998",
booktitle = "Proceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013",

}

TY - GEN

T1 - Analysis of randomized work stealing with false sharing

AU - Cole, Richard

AU - Ramachandran, Vijaya

PY - 2013

Y1 - 2013

N2 - This paper analyzes the overhead due to false sharing when parallel tasks are scheduled using randomized work stealing (RWS). We obtain high-probability bounds on the cache miss overhead, including the overhead due to false sharing, for several parallel cache-efficient algorithms when scheduled using RWS. These include algorithms for fundamental problems, such as matrix computations, FFT, sorting, basic dynamic programming, list ranking and graph connected components. Our main technical contribution, from which these results follow, is the derivation of nontrivial high-probability bounds on the number of steals incurred by these algorithms in the presence of false sharing, when using RWS.

AB - This paper analyzes the overhead due to false sharing when parallel tasks are scheduled using randomized work stealing (RWS). We obtain high-probability bounds on the cache miss overhead, including the overhead due to false sharing, for several parallel cache-efficient algorithms when scheduled using RWS. These include algorithms for fundamental problems, such as matrix computations, FFT, sorting, basic dynamic programming, list ranking and graph connected components. Our main technical contribution, from which these results follow, is the derivation of nontrivial high-probability bounds on the number of steals incurred by these algorithms in the presence of false sharing, when using RWS.

KW - false sharing

KW - performance analysis

KW - Randomized work stealing

UR - http://www.scopus.com/inward/record.url?scp=84884873307&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84884873307&partnerID=8YFLogxK

U2 - 10.1109/IPDPS.2013.86

DO - 10.1109/IPDPS.2013.86

M3 - Conference contribution

SP - 985

EP - 998

BT - Proceedings - IEEE 27th International Parallel and Distributed Processing Symposium, IPDPS 2013

ER -