Academic Journal

A Comparative Evaluation of Software Techniques to Hide Memory Latency

التفاصيل البيبلوغرافية
العنوان: A Comparative Evaluation of Software Techniques to Hide Memory Latency
المؤلفون: Lizy Kurian, Vinod Reddy, Paul T. Hulina, Lee D. Coraor
المساهمون: The Pennsylvania State University CiteSeerX Archives
المصدر: http://www.ece.utexas.edu/projects/ece/lca/ps/HICCS.ps.
سنة النشر: 1995
المجموعة: CiteSeerX
مصطلحات موضوعية: Compiler Optimization, Data Prefetching, Loop Unrolling, Memory Latency, Software Cache Prefetching, Software Pipelining, Static Scheduling
الوصف: Software oriented techniques to hide memory latency in superscalar and superpipelined machines include loop unrolling, software pipelining, and software cache prefetching. Issuing the data fetch request prior to actual need for data allows overlap of accessing with useful computations. Loop unrolling and software pipelining do not necessitate microarchitecture or instruction set architecture changes, whereas software controlled prefetching does. While studies on the benefits of the individual techniques have been done, no study evaluates all of these techniques within a consistent framework. This paper attempts to remedy this by providing a comparative evaluation of the features and benefits of the techniques. Loop unrolling and static scheduling of loads is seen to produce significant improvement in performance at lower latencies. Software pipelining is observed to be better than software controlled prefetching at lower latencies, but at higher latencies, software prefetching outperfo.
نوع الوثيقة: text
وصف الملف: application/postscript
اللغة: English
Relation: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.48.3359; http://www.ece.utexas.edu/projects/ece/lca/ps/HICCS.ps
الاتاحة: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.48.3359
http://www.ece.utexas.edu/projects/ece/lca/ps/HICCS.ps
Rights: Metadata may be used without restrictions as long as the oai identifier remains attached to it.
رقم الانضمام: edsbas.22CB28D8
قاعدة البيانات: BASE
ResultId 1
Header edsbas
BASE
edsbas.22CB28D8
820
3
Academic Journal
academicJournal
819.799560546875
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&scope=site&db=edsbas&AN=edsbas.22CB28D8&custid=s6537998&authtype=sso
FullText Array ( [Availability] => 0 )
Array ( [0] => Array ( [Url] => http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.48.3359# [Name] => EDS - BASE [Category] => fullText [Text] => View record in BASE [MouseOverText] => View record in BASE ) )
Items Array ( [Name] => Title [Label] => Title [Group] => Ti [Data] => A Comparative Evaluation of Software Techniques to Hide Memory Latency )
Array ( [Name] => Author [Label] => Authors [Group] => Au [Data] => <searchLink fieldCode="AR" term="%22Lizy+Kurian%22">Lizy Kurian</searchLink><br /><searchLink fieldCode="AR" term="%22Vinod+Reddy%22">Vinod Reddy</searchLink><br /><searchLink fieldCode="AR" term="%22Paul+T%2E+Hulina%22">Paul T. Hulina</searchLink><br /><searchLink fieldCode="AR" term="%22Lee+D%2E+Coraor%22">Lee D. Coraor</searchLink> )
Array ( [Name] => Author [Label] => Contributors [Group] => Au [Data] => The Pennsylvania State University CiteSeerX Archives )
Array ( [Name] => TitleSource [Label] => Source [Group] => Src [Data] => <i>http://www.ece.utexas.edu/projects/ece/lca/ps/HICCS.ps</i>. )
Array ( [Name] => DatePubCY [Label] => Publication Year [Group] => Date [Data] => 1995 )
Array ( [Name] => Subset [Label] => Collection [Group] => HoldingsInfo [Data] => CiteSeerX )
Array ( [Name] => Subject [Label] => Subject Terms [Group] => Su [Data] => <searchLink fieldCode="DE" term="%22Compiler+Optimization%22">Compiler Optimization</searchLink><br /><searchLink fieldCode="DE" term="%22Data+Prefetching%22">Data Prefetching</searchLink><br /><searchLink fieldCode="DE" term="%22Loop+Unrolling%22">Loop Unrolling</searchLink><br /><searchLink fieldCode="DE" term="%22Memory+Latency%22">Memory Latency</searchLink><br /><searchLink fieldCode="DE" term="%22Software+Cache+Prefetching%22">Software Cache Prefetching</searchLink><br /><searchLink fieldCode="DE" term="%22Software+Pipelining%22">Software Pipelining</searchLink><br /><searchLink fieldCode="DE" term="%22Static+Scheduling%22">Static Scheduling</searchLink> )
Array ( [Name] => Abstract [Label] => Description [Group] => Ab [Data] => Software oriented techniques to hide memory latency in superscalar and superpipelined machines include loop unrolling, software pipelining, and software cache prefetching. Issuing the data fetch request prior to actual need for data allows overlap of accessing with useful computations. Loop unrolling and software pipelining do not necessitate microarchitecture or instruction set architecture changes, whereas software controlled prefetching does. While studies on the benefits of the individual techniques have been done, no study evaluates all of these techniques within a consistent framework. This paper attempts to remedy this by providing a comparative evaluation of the features and benefits of the techniques. Loop unrolling and static scheduling of loads is seen to produce significant improvement in performance at lower latencies. Software pipelining is observed to be better than software controlled prefetching at lower latencies, but at higher latencies, software prefetching outperfo. )
Array ( [Name] => TypeDocument [Label] => Document Type [Group] => TypDoc [Data] => text )
Array ( [Name] => Format [Label] => File Description [Group] => SrcInfo [Data] => application/postscript )
Array ( [Name] => Language [Label] => Language [Group] => Lang [Data] => English )
Array ( [Name] => NoteTitleSource [Label] => Relation [Group] => SrcInfo [Data] => http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.48.3359; http://www.ece.utexas.edu/projects/ece/lca/ps/HICCS.ps )
Array ( [Name] => URL [Label] => Availability [Group] => URL [Data] => http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.48.3359<br />http://www.ece.utexas.edu/projects/ece/lca/ps/HICCS.ps )
Array ( [Name] => Copyright [Label] => Rights [Group] => Cpyrght [Data] => Metadata may be used without restrictions as long as the oai identifier remains attached to it. )
Array ( [Name] => AN [Label] => Accession Number [Group] => ID [Data] => edsbas.22CB28D8 )
RecordInfo Array ( [BibEntity] => Array ( [Languages] => Array ( [0] => Array ( [Text] => English ) ) [Subjects] => Array ( [0] => Array ( [SubjectFull] => Compiler Optimization [Type] => general ) [1] => Array ( [SubjectFull] => Data Prefetching [Type] => general ) [2] => Array ( [SubjectFull] => Loop Unrolling [Type] => general ) [3] => Array ( [SubjectFull] => Memory Latency [Type] => general ) [4] => Array ( [SubjectFull] => Software Cache Prefetching [Type] => general ) [5] => Array ( [SubjectFull] => Software Pipelining [Type] => general ) [6] => Array ( [SubjectFull] => Static Scheduling [Type] => general ) ) [Titles] => Array ( [0] => Array ( [TitleFull] => A Comparative Evaluation of Software Techniques to Hide Memory Latency [Type] => main ) ) ) [BibRelationships] => Array ( [HasContributorRelationships] => Array ( [0] => Array ( [PersonEntity] => Array ( [Name] => Array ( [NameFull] => Lizy Kurian ) ) ) [1] => Array ( [PersonEntity] => Array ( [Name] => Array ( [NameFull] => Vinod Reddy ) ) ) [2] => Array ( [PersonEntity] => Array ( [Name] => Array ( [NameFull] => Paul T. Hulina ) ) ) [3] => Array ( [PersonEntity] => Array ( [Name] => Array ( [NameFull] => Lee D. Coraor ) ) ) [4] => Array ( [PersonEntity] => Array ( [Name] => Array ( [NameFull] => The Pennsylvania State University CiteSeerX Archives ) ) ) ) [IsPartOfRelationships] => Array ( [0] => Array ( [BibEntity] => Array ( [Dates] => Array ( [0] => Array ( [D] => 01 [M] => 01 [Type] => published [Y] => 1995 ) ) [Identifiers] => Array ( [0] => Array ( [Type] => issn-locals [Value] => edsbas ) [1] => Array ( [Type] => issn-locals [Value] => edsbas.oa ) ) [Titles] => Array ( [0] => Array ( [TitleFull] => http://www.ece.utexas.edu/projects/ece/lca/ps/HICCS.ps [Type] => main ) ) ) ) ) ) )
IllustrationInfo