Workload prediction using run-length encoding for runtime processor power management

S. W. Kim, T. M. Kim, Hyuck Yoo

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

In processor power management research, workload prediction is a requisite for adjusting frequency without performance loss, and previous studies have proposed various prediction algorithms. Among them, methods based on the workload history table are lightweight and have high prediction accuracy for a variable workload. However, such prediction algorithms lose their prediction accuracy in the case of repeated workload. An improved workload prediction method using run-length encoding is proposed, which handles workload repetition. Evaluation results show that the proposed algorithm improves the prediction of repeated workload by up to 14% and also improves 4% of energy saving.

Original languageEnglish
Pages (from-to)1759-1761
Number of pages3
JournalElectronics Letters
Volume51
Issue number22
DOIs
Publication statusPublished - 2015 Oct 22

Fingerprint

Power management
Energy conservation
History

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Workload prediction using run-length encoding for runtime processor power management. / Kim, S. W.; Kim, T. M.; Yoo, Hyuck.

In: Electronics Letters, Vol. 51, No. 22, 22.10.2015, p. 1759-1761.

Research output: Contribution to journalArticle

@article{176211ef903e415ba3716c1208b751ef,
title = "Workload prediction using run-length encoding for runtime processor power management",
abstract = "In processor power management research, workload prediction is a requisite for adjusting frequency without performance loss, and previous studies have proposed various prediction algorithms. Among them, methods based on the workload history table are lightweight and have high prediction accuracy for a variable workload. However, such prediction algorithms lose their prediction accuracy in the case of repeated workload. An improved workload prediction method using run-length encoding is proposed, which handles workload repetition. Evaluation results show that the proposed algorithm improves the prediction of repeated workload by up to 14{\%} and also improves 4{\%} of energy saving.",
author = "Kim, {S. W.} and Kim, {T. M.} and Hyuck Yoo",
year = "2015",
month = "10",
day = "22",
doi = "10.1049/el.2014.4529",
language = "English",
volume = "51",
pages = "1759--1761",
journal = "Electronics Letters",
issn = "0013-5194",
publisher = "Institution of Engineering and Technology",
number = "22",

}

TY - JOUR

T1 - Workload prediction using run-length encoding for runtime processor power management

AU - Kim, S. W.

AU - Kim, T. M.

AU - Yoo, Hyuck

PY - 2015/10/22

Y1 - 2015/10/22

N2 - In processor power management research, workload prediction is a requisite for adjusting frequency without performance loss, and previous studies have proposed various prediction algorithms. Among them, methods based on the workload history table are lightweight and have high prediction accuracy for a variable workload. However, such prediction algorithms lose their prediction accuracy in the case of repeated workload. An improved workload prediction method using run-length encoding is proposed, which handles workload repetition. Evaluation results show that the proposed algorithm improves the prediction of repeated workload by up to 14% and also improves 4% of energy saving.

AB - In processor power management research, workload prediction is a requisite for adjusting frequency without performance loss, and previous studies have proposed various prediction algorithms. Among them, methods based on the workload history table are lightweight and have high prediction accuracy for a variable workload. However, such prediction algorithms lose their prediction accuracy in the case of repeated workload. An improved workload prediction method using run-length encoding is proposed, which handles workload repetition. Evaluation results show that the proposed algorithm improves the prediction of repeated workload by up to 14% and also improves 4% of energy saving.

UR - http://www.scopus.com/inward/record.url?scp=84946013113&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84946013113&partnerID=8YFLogxK

U2 - 10.1049/el.2014.4529

DO - 10.1049/el.2014.4529

M3 - Article

AN - SCOPUS:84946013113

VL - 51

SP - 1759

EP - 1761

JO - Electronics Letters

JF - Electronics Letters

SN - 0013-5194

IS - 22

ER -