Frequent instruction sequential pattern mining in hardware sample data

Jia Zou, Jing Xiao, Rui Hou, Yanqi Wang

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

9 Citations (Scopus)

Abstract

When parallelism and heterogeneity has become the trend for computer system design, both the size and the complexity of the hardware sample data generated by Performance Monitoring Unit (PMU) increase fast, thus automatic analysis methods, i.e. data mining methods, are urgently needed to accelerate hardware sample data analysis. We are the first to study instruction sequential pattern mining for hardware sample data. It is a challenging task due to the implicit sequential relationship contained in the data and due to the importance of low frequency patterns. As a solution, we i) provide a novel algorithm Prof Span; ii) adapt two existing algorithms, which are based on candidate generation and projected database generation, to hardware sample data. Our evaluation results show Prof Span can reduce up to 75% and 80% of execution time compared with other two algorithms. Particularly, up to 50% of frequent patterns mined by Prof Span in hardware sample data are crossing basic block boundaries and can not be found by existing methods for source code or disassembly code. We also analyze three example patterns identified by Prof Span: consecutive loads, JIT entry sequence, and conditional code dependency sequence, to illustrate how Prof Span can benefit performance analysis. Finally, we apply patterns to module classification and obtain promising results.

Original languageEnglish (US)
Title of host publicationProceedings - 10th IEEE International Conference on Data Mining, ICDM 2010
Pages1205-1210
Number of pages6
DOIs
StatePublished - Dec 1 2010
Externally publishedYes
Event10th IEEE International Conference on Data Mining, ICDM 2010 - Sydney, NSW, Australia
Duration: Dec 14 2010Dec 17 2010

Publication series

NameProceedings - IEEE International Conference on Data Mining, ICDM
ISSN (Print)1550-4786

Other

Other10th IEEE International Conference on Data Mining, ICDM 2010
CountryAustralia
CitySydney, NSW
Period12/14/1012/17/10

Fingerprint

Hardware
Data mining
Computer systems
Systems analysis
Monitoring

Keywords

  • Hardware sample data
  • Performance analysis
  • Sequential patter mining

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Zou, J., Xiao, J., Hou, R., & Wang, Y. (2010). Frequent instruction sequential pattern mining in hardware sample data. In Proceedings - 10th IEEE International Conference on Data Mining, ICDM 2010 (pp. 1205-1210). [5694109] (Proceedings - IEEE International Conference on Data Mining, ICDM). https://doi.org/10.1109/ICDM.2010.123

Frequent instruction sequential pattern mining in hardware sample data. / Zou, Jia; Xiao, Jing; Hou, Rui; Wang, Yanqi.

Proceedings - 10th IEEE International Conference on Data Mining, ICDM 2010. 2010. p. 1205-1210 5694109 (Proceedings - IEEE International Conference on Data Mining, ICDM).

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

Zou, J, Xiao, J, Hou, R & Wang, Y 2010, Frequent instruction sequential pattern mining in hardware sample data. in Proceedings - 10th IEEE International Conference on Data Mining, ICDM 2010., 5694109, Proceedings - IEEE International Conference on Data Mining, ICDM, pp. 1205-1210, 10th IEEE International Conference on Data Mining, ICDM 2010, Sydney, NSW, Australia, 12/14/10. https://doi.org/10.1109/ICDM.2010.123
Zou J, Xiao J, Hou R, Wang Y. Frequent instruction sequential pattern mining in hardware sample data. In Proceedings - 10th IEEE International Conference on Data Mining, ICDM 2010. 2010. p. 1205-1210. 5694109. (Proceedings - IEEE International Conference on Data Mining, ICDM). https://doi.org/10.1109/ICDM.2010.123
Zou, Jia ; Xiao, Jing ; Hou, Rui ; Wang, Yanqi. / Frequent instruction sequential pattern mining in hardware sample data. Proceedings - 10th IEEE International Conference on Data Mining, ICDM 2010. 2010. pp. 1205-1210 (Proceedings - IEEE International Conference on Data Mining, ICDM).
@inproceedings{a634dd08d688487db28c05d81ea3fb33,
title = "Frequent instruction sequential pattern mining in hardware sample data",
abstract = "When parallelism and heterogeneity has become the trend for computer system design, both the size and the complexity of the hardware sample data generated by Performance Monitoring Unit (PMU) increase fast, thus automatic analysis methods, i.e. data mining methods, are urgently needed to accelerate hardware sample data analysis. We are the first to study instruction sequential pattern mining for hardware sample data. It is a challenging task due to the implicit sequential relationship contained in the data and due to the importance of low frequency patterns. As a solution, we i) provide a novel algorithm Prof Span; ii) adapt two existing algorithms, which are based on candidate generation and projected database generation, to hardware sample data. Our evaluation results show Prof Span can reduce up to 75{\%} and 80{\%} of execution time compared with other two algorithms. Particularly, up to 50{\%} of frequent patterns mined by Prof Span in hardware sample data are crossing basic block boundaries and can not be found by existing methods for source code or disassembly code. We also analyze three example patterns identified by Prof Span: consecutive loads, JIT entry sequence, and conditional code dependency sequence, to illustrate how Prof Span can benefit performance analysis. Finally, we apply patterns to module classification and obtain promising results.",
keywords = "Hardware sample data, Performance analysis, Sequential patter mining",
author = "Jia Zou and Jing Xiao and Rui Hou and Yanqi Wang",
year = "2010",
month = "12",
day = "1",
doi = "10.1109/ICDM.2010.123",
language = "English (US)",
isbn = "9780769542560",
series = "Proceedings - IEEE International Conference on Data Mining, ICDM",
pages = "1205--1210",
booktitle = "Proceedings - 10th IEEE International Conference on Data Mining, ICDM 2010",

}

TY - GEN

T1 - Frequent instruction sequential pattern mining in hardware sample data

AU - Zou, Jia

AU - Xiao, Jing

AU - Hou, Rui

AU - Wang, Yanqi

PY - 2010/12/1

Y1 - 2010/12/1

N2 - When parallelism and heterogeneity has become the trend for computer system design, both the size and the complexity of the hardware sample data generated by Performance Monitoring Unit (PMU) increase fast, thus automatic analysis methods, i.e. data mining methods, are urgently needed to accelerate hardware sample data analysis. We are the first to study instruction sequential pattern mining for hardware sample data. It is a challenging task due to the implicit sequential relationship contained in the data and due to the importance of low frequency patterns. As a solution, we i) provide a novel algorithm Prof Span; ii) adapt two existing algorithms, which are based on candidate generation and projected database generation, to hardware sample data. Our evaluation results show Prof Span can reduce up to 75% and 80% of execution time compared with other two algorithms. Particularly, up to 50% of frequent patterns mined by Prof Span in hardware sample data are crossing basic block boundaries and can not be found by existing methods for source code or disassembly code. We also analyze three example patterns identified by Prof Span: consecutive loads, JIT entry sequence, and conditional code dependency sequence, to illustrate how Prof Span can benefit performance analysis. Finally, we apply patterns to module classification and obtain promising results.

AB - When parallelism and heterogeneity has become the trend for computer system design, both the size and the complexity of the hardware sample data generated by Performance Monitoring Unit (PMU) increase fast, thus automatic analysis methods, i.e. data mining methods, are urgently needed to accelerate hardware sample data analysis. We are the first to study instruction sequential pattern mining for hardware sample data. It is a challenging task due to the implicit sequential relationship contained in the data and due to the importance of low frequency patterns. As a solution, we i) provide a novel algorithm Prof Span; ii) adapt two existing algorithms, which are based on candidate generation and projected database generation, to hardware sample data. Our evaluation results show Prof Span can reduce up to 75% and 80% of execution time compared with other two algorithms. Particularly, up to 50% of frequent patterns mined by Prof Span in hardware sample data are crossing basic block boundaries and can not be found by existing methods for source code or disassembly code. We also analyze three example patterns identified by Prof Span: consecutive loads, JIT entry sequence, and conditional code dependency sequence, to illustrate how Prof Span can benefit performance analysis. Finally, we apply patterns to module classification and obtain promising results.

KW - Hardware sample data

KW - Performance analysis

KW - Sequential patter mining

UR - http://www.scopus.com/inward/record.url?scp=79951738836&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79951738836&partnerID=8YFLogxK

U2 - 10.1109/ICDM.2010.123

DO - 10.1109/ICDM.2010.123

M3 - Conference contribution

AN - SCOPUS:79951738836

SN - 9780769542560

T3 - Proceedings - IEEE International Conference on Data Mining, ICDM

SP - 1205

EP - 1210

BT - Proceedings - 10th IEEE International Conference on Data Mining, ICDM 2010

ER -