Memory coherence activity prediction in commercial workloads

Stephen Somogyi*, Thomas F. Wenisch, Nikolaos Hardavellas, Jangwoo Kim, Anastassia Ailamaki, Babak Falsafi

*Corresponding author for this work

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

17 Citations (Scopus)

Abstract

Recent research indicates that prediction-based coherence optimizations offer substantial performance improvements for scientific applications in distributed shared memory multiprocessors. Important commercial applications also show sensitivity to coherence latency, which will become more acute in the future as technology scales. Therefore it is important to investigate prediction of memory coherence activity in the context of commercial workloads.This paper studies a trace-based Downgrade Predictor (DGP) for predicting last stores to shared cache blocks, and a pattern-based Consumer Set Predictor (CSP) for predicting subsequent readers. We evaluate this class of predictors for the first time on commercial applications and demonstrate that our DGP correctly predicts 47%-76% of last stores. Memory sharing patterns in commercial workloads are inherently non-repetitive; hence CSP cannot attain high coverage. We perform an opportunity study of a DGP enhanced through competitive underlying predictors, and in commercial and scientific applications, demonstrate potential to increase coverage up to 14%.

Original languageEnglish (US)
Title of host publicationProceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture
Pages37-45
Number of pages9
DOIs
StatePublished - Dec 1 2004
Event3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture - Munich, Germany
Duration: Jun 20 2004Jun 20 2004

Publication series

NameACM International Conference Proceeding Series
Volume68

Other

Other3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture
CountryGermany
CityMunich
Period6/20/046/20/04

Fingerprint

Data storage equipment

Keywords

  • coherence misses
  • coherence prediction
  • commercial workloads
  • sharing patterns
  • trace-based prediction

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Cite this

Somogyi, S., Wenisch, T. F., Hardavellas, N., Kim, J., Ailamaki, A., & Falsafi, B. (2004). Memory coherence activity prediction in commercial workloads. In Proceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture (pp. 37-45). (ACM International Conference Proceeding Series; Vol. 68). https://doi.org/10.1145/1054943.1054949
Somogyi, Stephen ; Wenisch, Thomas F. ; Hardavellas, Nikolaos ; Kim, Jangwoo ; Ailamaki, Anastassia ; Falsafi, Babak. / Memory coherence activity prediction in commercial workloads. Proceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture. 2004. pp. 37-45 (ACM International Conference Proceeding Series).
@inproceedings{32ba7431a3f84837a042323df3ff3dd2,
title = "Memory coherence activity prediction in commercial workloads",
abstract = "Recent research indicates that prediction-based coherence optimizations offer substantial performance improvements for scientific applications in distributed shared memory multiprocessors. Important commercial applications also show sensitivity to coherence latency, which will become more acute in the future as technology scales. Therefore it is important to investigate prediction of memory coherence activity in the context of commercial workloads.This paper studies a trace-based Downgrade Predictor (DGP) for predicting last stores to shared cache blocks, and a pattern-based Consumer Set Predictor (CSP) for predicting subsequent readers. We evaluate this class of predictors for the first time on commercial applications and demonstrate that our DGP correctly predicts 47{\%}-76{\%} of last stores. Memory sharing patterns in commercial workloads are inherently non-repetitive; hence CSP cannot attain high coverage. We perform an opportunity study of a DGP enhanced through competitive underlying predictors, and in commercial and scientific applications, demonstrate potential to increase coverage up to 14{\%}.",
keywords = "coherence misses, coherence prediction, commercial workloads, sharing patterns, trace-based prediction",
author = "Stephen Somogyi and Wenisch, {Thomas F.} and Nikolaos Hardavellas and Jangwoo Kim and Anastassia Ailamaki and Babak Falsafi",
year = "2004",
month = "12",
day = "1",
doi = "10.1145/1054943.1054949",
language = "English (US)",
isbn = "159593040X",
series = "ACM International Conference Proceeding Series",
pages = "37--45",
booktitle = "Proceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture",

}

Somogyi, S, Wenisch, TF, Hardavellas, N, Kim, J, Ailamaki, A & Falsafi, B 2004, Memory coherence activity prediction in commercial workloads. in Proceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture. ACM International Conference Proceeding Series, vol. 68, pp. 37-45, 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture, Munich, Germany, 6/20/04. https://doi.org/10.1145/1054943.1054949

Memory coherence activity prediction in commercial workloads. / Somogyi, Stephen; Wenisch, Thomas F.; Hardavellas, Nikolaos; Kim, Jangwoo; Ailamaki, Anastassia; Falsafi, Babak.

Proceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture. 2004. p. 37-45 (ACM International Conference Proceeding Series; Vol. 68).

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

TY - GEN

T1 - Memory coherence activity prediction in commercial workloads

AU - Somogyi, Stephen

AU - Wenisch, Thomas F.

AU - Hardavellas, Nikolaos

AU - Kim, Jangwoo

AU - Ailamaki, Anastassia

AU - Falsafi, Babak

PY - 2004/12/1

Y1 - 2004/12/1

N2 - Recent research indicates that prediction-based coherence optimizations offer substantial performance improvements for scientific applications in distributed shared memory multiprocessors. Important commercial applications also show sensitivity to coherence latency, which will become more acute in the future as technology scales. Therefore it is important to investigate prediction of memory coherence activity in the context of commercial workloads.This paper studies a trace-based Downgrade Predictor (DGP) for predicting last stores to shared cache blocks, and a pattern-based Consumer Set Predictor (CSP) for predicting subsequent readers. We evaluate this class of predictors for the first time on commercial applications and demonstrate that our DGP correctly predicts 47%-76% of last stores. Memory sharing patterns in commercial workloads are inherently non-repetitive; hence CSP cannot attain high coverage. We perform an opportunity study of a DGP enhanced through competitive underlying predictors, and in commercial and scientific applications, demonstrate potential to increase coverage up to 14%.

AB - Recent research indicates that prediction-based coherence optimizations offer substantial performance improvements for scientific applications in distributed shared memory multiprocessors. Important commercial applications also show sensitivity to coherence latency, which will become more acute in the future as technology scales. Therefore it is important to investigate prediction of memory coherence activity in the context of commercial workloads.This paper studies a trace-based Downgrade Predictor (DGP) for predicting last stores to shared cache blocks, and a pattern-based Consumer Set Predictor (CSP) for predicting subsequent readers. We evaluate this class of predictors for the first time on commercial applications and demonstrate that our DGP correctly predicts 47%-76% of last stores. Memory sharing patterns in commercial workloads are inherently non-repetitive; hence CSP cannot attain high coverage. We perform an opportunity study of a DGP enhanced through competitive underlying predictors, and in commercial and scientific applications, demonstrate potential to increase coverage up to 14%.

KW - coherence misses

KW - coherence prediction

KW - commercial workloads

KW - sharing patterns

KW - trace-based prediction

UR - http://www.scopus.com/inward/record.url?scp=77954446302&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954446302&partnerID=8YFLogxK

U2 - 10.1145/1054943.1054949

DO - 10.1145/1054943.1054949

M3 - Conference contribution

AN - SCOPUS:77954446302

SN - 159593040X

SN - 9781595930408

T3 - ACM International Conference Proceeding Series

SP - 37

EP - 45

BT - Proceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture

ER -

Somogyi S, Wenisch TF, Hardavellas N, Kim J, Ailamaki A, Falsafi B. Memory coherence activity prediction in commercial workloads. In Proceedings of the 3rd Workshop on Memory Performance Issues, WMPI '04, in Conjunction with the 31st International Symposium on Computer Architecture. 2004. p. 37-45. (ACM International Conference Proceeding Series). https://doi.org/10.1145/1054943.1054949