File Download
  Links for fulltext
     (May Require Subscription)
Supplementary

Article: EClass: An execution classification approach to improving the energy-efficiency of software via machine learning

TitleEClass: An execution classification approach to improving the energy-efficiency of software via machine learning
Authors
KeywordsDVFS
Energy optimization
Energy saving
Machine learning
Workload prediction
Issue Date2012
PublisherElsevier Inc. The Journal's web site is located at http://www.elsevier.com/locate/jss
Citation
Journal of Systems and Software, 2012, v. 85 n. 4, p. 960-973 How to Cite?
AbstractEnergy efficiency at the software level has gained much attention in the past decade. This paper presents a performance-aware frequency assignment algorithm for reducing processor energy consumption using Dynamic Voltage and Frequency Scaling (DVFS). Existing energy-saving techniques often rely on simplified predictions or domain knowledge to extract energy savings for specialized software (such as multimedia or mobile applications) or hardware (such as NPU or sensor nodes). We present an innovative framework, known as EClass, for general-purpose DVFS processors by recognizing short and repetitive utilization patterns efficiently using machine learning. Our algorithm is lightweight and can save up to 52.9% of the energy consumption compared with the classical PAST algorithm. It achieves an average savings of 9.1% when compared with an existing online learning algorithm that also utilizes the statistics from the current execution only. We have simulated the algorithms on a cycle-accurate power simulator. Experimental results show that EClass can effectively save energy for real life applications that exhibit mixed CPU utilization patterns during executions. Our research challenges an assumption among previous work in the research community that a simple and efficient heuristic should be used to adjust the processor frequency online. Our empirical result shows that the use of an advanced algorithm such as machine learning can not only compensate for the energy needed to run such an algorithm, but also outperforms prior techniques based on the above assumption. © 2011 Elsevier Inc. All rights reserved.
Persistent Identifierhttp://hdl.handle.net/10722/152496
ISSN
2023 Impact Factor: 3.7
2023 SCImago Journal Rankings: 1.160
ISI Accession Number ID
Funding AgencyGrant Number
Council of Hong Kong111410
717811
City University of Hong Kong7002673
Funding Information:

This work is supported in part by the General Research Fund of the Research Grants Council of Hong Kong (project numbers 111410 and 717811) and a Strategic Research Grant of City University of Hong Kong (project number 7002673).

References

 

DC FieldValueLanguage
dc.contributor.authorKan, EYYen_HK
dc.contributor.authorChan, WKen_HK
dc.contributor.authorTse, THen_HK
dc.date.accessioned2012-06-26T06:39:41Z-
dc.date.available2012-06-26T06:39:41Z-
dc.date.issued2012en_HK
dc.identifier.citationJournal of Systems and Software, 2012, v. 85 n. 4, p. 960-973en_HK
dc.identifier.issn0164-1212en_HK
dc.identifier.urihttp://hdl.handle.net/10722/152496-
dc.description.abstractEnergy efficiency at the software level has gained much attention in the past decade. This paper presents a performance-aware frequency assignment algorithm for reducing processor energy consumption using Dynamic Voltage and Frequency Scaling (DVFS). Existing energy-saving techniques often rely on simplified predictions or domain knowledge to extract energy savings for specialized software (such as multimedia or mobile applications) or hardware (such as NPU or sensor nodes). We present an innovative framework, known as EClass, for general-purpose DVFS processors by recognizing short and repetitive utilization patterns efficiently using machine learning. Our algorithm is lightweight and can save up to 52.9% of the energy consumption compared with the classical PAST algorithm. It achieves an average savings of 9.1% when compared with an existing online learning algorithm that also utilizes the statistics from the current execution only. We have simulated the algorithms on a cycle-accurate power simulator. Experimental results show that EClass can effectively save energy for real life applications that exhibit mixed CPU utilization patterns during executions. Our research challenges an assumption among previous work in the research community that a simple and efficient heuristic should be used to adjust the processor frequency online. Our empirical result shows that the use of an advanced algorithm such as machine learning can not only compensate for the energy needed to run such an algorithm, but also outperforms prior techniques based on the above assumption. © 2011 Elsevier Inc. All rights reserved.en_HK
dc.languageengen_US
dc.publisherElsevier Inc. The Journal's web site is located at http://www.elsevier.com/locate/jssen_HK
dc.relation.ispartofJournal of Systems and Softwareen_HK
dc.rightsNOTICE: this is the author’s version of a work that was accepted for publication in Journal of Systems and Software. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Journal of Systems and Software, 2012, v. 85 n. 4, p. 960-973. DOI: 10.1016/j.jss.2011.11.1010-
dc.rightsThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.-
dc.subjectDVFSen_HK
dc.subjectEnergy optimizationen_HK
dc.subjectEnergy savingen_HK
dc.subjectMachine learningen_HK
dc.subjectWorkload predictionen_HK
dc.titleEClass: An execution classification approach to improving the energy-efficiency of software via machine learningen_HK
dc.typeArticleen_HK
dc.identifier.emailTse, TH: thtse@cs.hku.hken_HK
dc.identifier.authorityTse, TH=rp00546en_HK
dc.description.naturepostprinten_US
dc.identifier.doi10.1016/j.jss.2011.11.1010en_HK
dc.identifier.scopuseid_2-s2.0-84857364286en_HK
dc.identifier.hkuros203896-
dc.relation.referenceshttp://www.scopus.com/mlt/select.url?eid=2-s2.0-84857364286&selection=ref&src=s&origin=recordpageen_HK
dc.identifier.volume85en_HK
dc.identifier.issue4en_HK
dc.identifier.spage960en_HK
dc.identifier.epage973en_HK
dc.identifier.isiWOS:000301630700015-
dc.publisher.placeUnited Statesen_HK
dc.identifier.scopusauthoridKan, EYY=36573652000en_HK
dc.identifier.scopusauthoridChan, WK=23967779900en_HK
dc.identifier.scopusauthoridTse, TH=7005496974en_HK
dc.identifier.citeulike10103145-
dc.identifier.issnl0164-1212-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats