Reducing Energy Usage with Memory and Computation-Aware Dynamic Frequency Scaling
Identifieur interne : 000761 ( Main/Exploration ); précédent : 000760; suivant : 000762Reducing Energy Usage with Memory and Computation-Aware Dynamic Frequency Scaling
Auteurs : A. Laurenzano [États-Unis] ; Mitesh Meswani [États-Unis] ; Laura Carrington [États-Unis] ; Allan Snavely [États-Unis] ; M. Tikir [États-Unis] ; Stephen Poole [États-Unis]Source :
- Lecture Notes in Computer Science [ 0302-9743 ] ; 2011.
Abstract
Abstract: Over the life of a modern supercomputer, the energy cost of running the system can exceed the cost of the original hardware purchase. This has driven the community to attempt to understand and minimize energy costs wherever possible. Towards these ends, we present an automated, fine-grained approach to selecting per-loop processor clock frequencies. The clock frequency selection criteria is established through a combination of lightweight static analysis and runtime tracing that automatically acquires application signatures - characterizations of the patterns of execution of each loop in an application. This application characterization is matched with one of a series of benchmark loops, which have been run on the target system and probe it in various ways. These benchmarks form a covering set, a machine characterization of the expected power consumption and performance traits of the machine over the space of execution patterns and clock frequencies. The frequency that confers the optimal behavior in terms of power-delay product for the benchmark that most closely resembles each application loop is the one chosen for that loop. The set of tools that implement this scheme is fully automated, built on top of freely available open source software, and uses an inexpensive power measurement apparatus. We use these tools to show a measured, system-wide energy savings of up to 7.6% on an 8-core Intel Xeon E5530 and 10.6% on a 32-core AMD Opteron 8380 (a Sun X4600 Node) across a range of workloads.
Url:
DOI: 10.1007/978-3-642-23400-2_9
Affiliations:
Links toward previous steps (curation, corpus...)
- to stream Istex, to step Corpus: 000383
- to stream Istex, to step Curation: 000383
- to stream Istex, to step Checkpoint: 000181
- to stream Main, to step Merge: 000763
- to stream Main, to step Curation: 000761
Le document en format XML
<record><TEI wicri:istexFullTextTei="biblStruct"><teiHeader><fileDesc><titleStmt><title xml:lang="en">Reducing Energy Usage with Memory and Computation-Aware Dynamic Frequency Scaling</title>
<author><name sortKey="Laurenzano, A" sort="Laurenzano, A" uniqKey="Laurenzano A" first="A." last="Laurenzano">A. Laurenzano</name>
</author>
<author><name sortKey="Meswani, Mitesh" sort="Meswani, Mitesh" uniqKey="Meswani M" first="Mitesh" last="Meswani">Mitesh Meswani</name>
</author>
<author><name sortKey="Carrington, Laura" sort="Carrington, Laura" uniqKey="Carrington L" first="Laura" last="Carrington">Laura Carrington</name>
</author>
<author><name sortKey="Snavely, Allan" sort="Snavely, Allan" uniqKey="Snavely A" first="Allan" last="Snavely">Allan Snavely</name>
</author>
<author><name sortKey="Tikir, M" sort="Tikir, M" uniqKey="Tikir M" first="M." last="Tikir">M. Tikir</name>
</author>
<author><name sortKey="Poole, Stephen" sort="Poole, Stephen" uniqKey="Poole S" first="Stephen" last="Poole">Stephen Poole</name>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">ISTEX</idno>
<idno type="RBID">ISTEX:85C771693C049D6774CFC213948CA9939456C56E</idno>
<date when="2011" year="2011">2011</date>
<idno type="doi">10.1007/978-3-642-23400-2_9</idno>
<idno type="url">https://api.istex.fr/document/85C771693C049D6774CFC213948CA9939456C56E/fulltext/pdf</idno>
<idno type="wicri:Area/Istex/Corpus">000383</idno>
<idno type="wicri:Area/Istex/Curation">000383</idno>
<idno type="wicri:Area/Istex/Checkpoint">000181</idno>
<idno type="wicri:doubleKey">0302-9743:2011:Laurenzano A:reducing:energy:usage</idno>
<idno type="wicri:Area/Main/Merge">000763</idno>
<idno type="wicri:Area/Main/Curation">000761</idno>
<idno type="wicri:Area/Main/Exploration">000761</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title level="a" type="main" xml:lang="en">Reducing Energy Usage with Memory and Computation-Aware Dynamic Frequency Scaling</title>
<author><name sortKey="Laurenzano, A" sort="Laurenzano, A" uniqKey="Laurenzano A" first="A." last="Laurenzano">A. Laurenzano</name>
<affiliation wicri:level="2"><country xml:lang="fr">États-Unis</country>
<wicri:regionArea>San Diego Supercomputer Center, La Jolla, CA</wicri:regionArea>
<placeName><region type="state">Californie</region>
</placeName>
</affiliation>
<affiliation wicri:level="1"><country wicri:rule="url">États-Unis</country>
</affiliation>
</author>
<author><name sortKey="Meswani, Mitesh" sort="Meswani, Mitesh" uniqKey="Meswani M" first="Mitesh" last="Meswani">Mitesh Meswani</name>
<affiliation wicri:level="2"><country xml:lang="fr">États-Unis</country>
<wicri:regionArea>San Diego Supercomputer Center, La Jolla, CA</wicri:regionArea>
<placeName><region type="state">Californie</region>
</placeName>
</affiliation>
<affiliation wicri:level="1"><country wicri:rule="url">États-Unis</country>
</affiliation>
</author>
<author><name sortKey="Carrington, Laura" sort="Carrington, Laura" uniqKey="Carrington L" first="Laura" last="Carrington">Laura Carrington</name>
<affiliation wicri:level="2"><country xml:lang="fr">États-Unis</country>
<wicri:regionArea>San Diego Supercomputer Center, La Jolla, CA</wicri:regionArea>
<placeName><region type="state">Californie</region>
</placeName>
</affiliation>
<affiliation wicri:level="1"><country wicri:rule="url">États-Unis</country>
</affiliation>
</author>
<author><name sortKey="Snavely, Allan" sort="Snavely, Allan" uniqKey="Snavely A" first="Allan" last="Snavely">Allan Snavely</name>
<affiliation wicri:level="2"><country xml:lang="fr">États-Unis</country>
<wicri:regionArea>San Diego Supercomputer Center, La Jolla, CA</wicri:regionArea>
<placeName><region type="state">Californie</region>
</placeName>
</affiliation>
<affiliation wicri:level="1"><country wicri:rule="url">États-Unis</country>
</affiliation>
</author>
<author><name sortKey="Tikir, M" sort="Tikir, M" uniqKey="Tikir M" first="M." last="Tikir">M. Tikir</name>
<affiliation wicri:level="2"><country xml:lang="fr">États-Unis</country>
<wicri:regionArea>Google, Inc, Mountain View, CA</wicri:regionArea>
<placeName><region type="state">Californie</region>
</placeName>
</affiliation>
<affiliation><wicri:noCountry code="no comma">E-mail: mustafa.m.tikir@gmail.com</wicri:noCountry>
</affiliation>
</author>
<author><name sortKey="Poole, Stephen" sort="Poole, Stephen" uniqKey="Poole S" first="Stephen" last="Poole">Stephen Poole</name>
<affiliation wicri:level="2"><country xml:lang="fr">États-Unis</country>
<wicri:regionArea>Oak Ridge National Laboratory, Oak Ridge, TN</wicri:regionArea>
<placeName><region type="state">Tennessee</region>
</placeName>
</affiliation>
<affiliation><wicri:noCountry code="no comma">E-mail: spoole@ornl.gov</wicri:noCountry>
</affiliation>
</author>
</analytic>
<monogr></monogr>
<series><title level="s">Lecture Notes in Computer Science</title>
<imprint><date>2011</date>
</imprint>
<idno type="ISSN">0302-9743</idno>
<idno type="eISSN">1611-3349</idno>
<idno type="ISSN">0302-9743</idno>
</series>
<idno type="istex">85C771693C049D6774CFC213948CA9939456C56E</idno>
<idno type="DOI">10.1007/978-3-642-23400-2_9</idno>
<idno type="ChapterID">9</idno>
<idno type="ChapterID">Chap9</idno>
</biblStruct>
</sourceDesc>
<seriesStmt><idno type="ISSN">0302-9743</idno>
</seriesStmt>
</fileDesc>
<profileDesc><textClass></textClass>
<langUsage><language ident="en">en</language>
</langUsage>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en">Abstract: Over the life of a modern supercomputer, the energy cost of running the system can exceed the cost of the original hardware purchase. This has driven the community to attempt to understand and minimize energy costs wherever possible. Towards these ends, we present an automated, fine-grained approach to selecting per-loop processor clock frequencies. The clock frequency selection criteria is established through a combination of lightweight static analysis and runtime tracing that automatically acquires application signatures - characterizations of the patterns of execution of each loop in an application. This application characterization is matched with one of a series of benchmark loops, which have been run on the target system and probe it in various ways. These benchmarks form a covering set, a machine characterization of the expected power consumption and performance traits of the machine over the space of execution patterns and clock frequencies. The frequency that confers the optimal behavior in terms of power-delay product for the benchmark that most closely resembles each application loop is the one chosen for that loop. The set of tools that implement this scheme is fully automated, built on top of freely available open source software, and uses an inexpensive power measurement apparatus. We use these tools to show a measured, system-wide energy savings of up to 7.6% on an 8-core Intel Xeon E5530 and 10.6% on a 32-core AMD Opteron 8380 (a Sun X4600 Node) across a range of workloads.</div>
</front>
</TEI>
<affiliations><list><country><li>États-Unis</li>
</country>
<region><li>Californie</li>
<li>Tennessee</li>
</region>
</list>
<tree><country name="États-Unis"><region name="Californie"><name sortKey="Laurenzano, A" sort="Laurenzano, A" uniqKey="Laurenzano A" first="A." last="Laurenzano">A. Laurenzano</name>
</region>
<name sortKey="Carrington, Laura" sort="Carrington, Laura" uniqKey="Carrington L" first="Laura" last="Carrington">Laura Carrington</name>
<name sortKey="Carrington, Laura" sort="Carrington, Laura" uniqKey="Carrington L" first="Laura" last="Carrington">Laura Carrington</name>
<name sortKey="Laurenzano, A" sort="Laurenzano, A" uniqKey="Laurenzano A" first="A." last="Laurenzano">A. Laurenzano</name>
<name sortKey="Meswani, Mitesh" sort="Meswani, Mitesh" uniqKey="Meswani M" first="Mitesh" last="Meswani">Mitesh Meswani</name>
<name sortKey="Meswani, Mitesh" sort="Meswani, Mitesh" uniqKey="Meswani M" first="Mitesh" last="Meswani">Mitesh Meswani</name>
<name sortKey="Poole, Stephen" sort="Poole, Stephen" uniqKey="Poole S" first="Stephen" last="Poole">Stephen Poole</name>
<name sortKey="Snavely, Allan" sort="Snavely, Allan" uniqKey="Snavely A" first="Allan" last="Snavely">Allan Snavely</name>
<name sortKey="Snavely, Allan" sort="Snavely, Allan" uniqKey="Snavely A" first="Allan" last="Snavely">Allan Snavely</name>
<name sortKey="Tikir, M" sort="Tikir, M" uniqKey="Tikir M" first="M." last="Tikir">M. Tikir</name>
</country>
</tree>
</affiliations>
</record>
Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Ticri/CIDE/explor/CyberinfraV1/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000761 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 000761 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien |wiki= Ticri/CIDE |area= CyberinfraV1 |flux= Main |étape= Exploration |type= RBID |clé= ISTEX:85C771693C049D6774CFC213948CA9939456C56E |texte= Reducing Energy Usage with Memory and Computation-Aware Dynamic Frequency Scaling }}
This area was generated with Dilib version V0.6.25. |