The Effect of Contention on the Scalability of Page-Based Software Shared Memory Systems
In this paper, we examine the causes and effects of contention for shared data access in parallel programs running on a software dis- tributed shared memory (DSM) system. Specifically, we experiment on two widely-used, page-based protocols, Princeton’s home-based lazy re- lease consistency (HLRC) an...
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Buchkapitel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 169 |
|---|---|
| container_issue | |
| container_start_page | 155 |
| container_title | |
| container_volume | 1915 |
| creator | de Lara, Eyal Lu, Honghui Charlie, Y. Cox, Alan L. Zwaenepoel, Willy |
| description | In this paper, we examine the causes and effects of contention for shared data access in parallel programs running on a software dis- tributed shared memory (DSM) system. Specifically, we experiment on two widely-used, page-based protocols, Princeton’s home-based lazy re- lease consistency (HLRC) and TreadMarks. For most of our programs, these protocols were equally affected by latency increases caused by con- tention and achieved similar performance. Where they differ significantly, HLRC’s ability to manually eliminate load imbalance was the largest fac- tor accounting for the difference. To quantify the effects of contention we either modified the application to eliminate the cause of the contention or modified the underlying protocol to efficiently handle it. Overall, we find that contention has profound effects on performance: eliminating contention reduced execution time by 64% in the most extreme case, even at the relatively modest scale of 32 nodes that we consider in this paper. |
| doi_str_mv | 10.1007/3-540-40889-4_12 |
| format | Book Chapter |
| fullrecord | <record><control><sourceid>proquest_pasca</sourceid><recordid>TN_cdi_pascalfrancis_primary_779803</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>EBC3071623_18_163</sourcerecordid><originalsourceid>FETCH-LOGICAL-p308t-3cb1afcd919e2ead812f91669a4fd2273b9d052d74c02eeeb0e43f9dcf8b66233</originalsourceid><addsrcrecordid>eNotkE1vFDEMhsOnupS9cxyJc4oTZzLJEVaFIhWBtEXiFmUyTndgd2ZJUqH992S2tSxb8vvash7G3gm4EgDdB-StAq7AGMuVE_IZW9vOYB2eZ-1zthJaCI6o7Av25iwIYVr5kq0AQXLbKXzNVhZbaUG3-oKtc_4NNVAqBL1iv-521FzHSKE0c2w281RoKuM8NTVL1bbB730_7sdyWgw__D3xTz7T0GznWP75VC27WofmGx3mdGq2p1zokN-yV9HvM62f-iX7-fn6bnPDb79_-br5eMuPCKZwDL3wMQxWWJLkByNktEJr61UcpOywtwO0cuhUAElEPZDCaIcQTa-1RLxk7x_vHn2ur8bkpzBmd0zjwaeT6zprYHFdPbpyFaZ7Sq6f5z_ZCXALaYeusnNnqm4hXRfw6Wya_z5QLo6WjVDhJL8PO38slLJD6ET9wgnjhEb8D4vEe6s</addsrcrecordid><sourcetype>Index Database</sourcetype><iscdi>true</iscdi><recordtype>book_chapter</recordtype><pqid>EBC3071623_18_163</pqid></control><display><type>book_chapter</type><title>The Effect of Contention on the Scalability of Page-Based Software Shared Memory Systems</title><source>Springer Books</source><creator>de Lara, Eyal ; Lu, Honghui ; Charlie, Y. ; Cox, Alan L. ; Zwaenepoel, Willy</creator><contributor>Dwarkadas, Sandhya</contributor><creatorcontrib>de Lara, Eyal ; Lu, Honghui ; Charlie, Y. ; Cox, Alan L. ; Zwaenepoel, Willy ; Dwarkadas, Sandhya</creatorcontrib><description>In this paper, we examine the causes and effects of contention for shared data access in parallel programs running on a software dis- tributed shared memory (DSM) system. Specifically, we experiment on two widely-used, page-based protocols, Princeton’s home-based lazy re- lease consistency (HLRC) and TreadMarks. For most of our programs, these protocols were equally affected by latency increases caused by con- tention and achieved similar performance. Where they differ significantly, HLRC’s ability to manually eliminate load imbalance was the largest fac- tor accounting for the difference. To quantify the effects of contention we either modified the application to eliminate the cause of the contention or modified the underlying protocol to efficiently handle it. Overall, we find that contention has profound effects on performance: eliminating contention reduced execution time by 64% in the most extreme case, even at the relatively modest scale of 32 nodes that we consider in this paper.</description><identifier>ISSN: 0302-9743</identifier><identifier>ISBN: 3540411852</identifier><identifier>ISBN: 9783540411857</identifier><identifier>EISSN: 1611-3349</identifier><identifier>EISBN: 9783540408895</identifier><identifier>EISBN: 3540408894</identifier><identifier>DOI: 10.1007/3-540-40889-4_12</identifier><identifier>OCLC: 935290656</identifier><identifier>LCCallNum: QA76.6-76.66</identifier><language>eng</language><publisher>Germany: Springer Berlin / Heidelberg</publisher><subject>Applied sciences ; Average Response Time ; Computer science; control theory; systems ; Exact sciences and technology ; Execution Time ; Home Node ; Language processing and microprogramming ; Load Imbalance ; Shared Memory ; Software</subject><ispartof>Languages, Compilers, and Run-Time Systems for Scalable Computers, 2000, Vol.1915, p.155-169</ispartof><rights>Springer-Verlag Berlin Heidelberg 2000</rights><rights>2001 INIST-CNRS</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><relation>Lecture Notes in Computer Science</relation></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Uhttps://ebookcentral.proquest.com/covers/3071623-l.jpg</thumbnail><linktopdf>$$Uhttps://link.springer.com/content/pdf/10.1007/3-540-40889-4_12$$EPDF$$P50$$Gspringer$$H</linktopdf><linktohtml>$$Uhttps://link.springer.com/10.1007/3-540-40889-4_12$$EHTML$$P50$$Gspringer$$H</linktohtml><link.rule.ids>309,310,775,776,780,785,786,789,4036,4037,27902,38232,41418,42487</link.rule.ids><backlink>$$Uhttp://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=779803$$DView record in Pascal Francis$$Hfree_for_read</backlink></links><search><contributor>Dwarkadas, Sandhya</contributor><creatorcontrib>de Lara, Eyal</creatorcontrib><creatorcontrib>Lu, Honghui</creatorcontrib><creatorcontrib>Charlie, Y.</creatorcontrib><creatorcontrib>Cox, Alan L.</creatorcontrib><creatorcontrib>Zwaenepoel, Willy</creatorcontrib><title>The Effect of Contention on the Scalability of Page-Based Software Shared Memory Systems</title><title>Languages, Compilers, and Run-Time Systems for Scalable Computers</title><description>In this paper, we examine the causes and effects of contention for shared data access in parallel programs running on a software dis- tributed shared memory (DSM) system. Specifically, we experiment on two widely-used, page-based protocols, Princeton’s home-based lazy re- lease consistency (HLRC) and TreadMarks. For most of our programs, these protocols were equally affected by latency increases caused by con- tention and achieved similar performance. Where they differ significantly, HLRC’s ability to manually eliminate load imbalance was the largest fac- tor accounting for the difference. To quantify the effects of contention we either modified the application to eliminate the cause of the contention or modified the underlying protocol to efficiently handle it. Overall, we find that contention has profound effects on performance: eliminating contention reduced execution time by 64% in the most extreme case, even at the relatively modest scale of 32 nodes that we consider in this paper.</description><subject>Applied sciences</subject><subject>Average Response Time</subject><subject>Computer science; control theory; systems</subject><subject>Exact sciences and technology</subject><subject>Execution Time</subject><subject>Home Node</subject><subject>Language processing and microprogramming</subject><subject>Load Imbalance</subject><subject>Shared Memory</subject><subject>Software</subject><issn>0302-9743</issn><issn>1611-3349</issn><isbn>3540411852</isbn><isbn>9783540411857</isbn><isbn>9783540408895</isbn><isbn>3540408894</isbn><fulltext>true</fulltext><rsrctype>book_chapter</rsrctype><creationdate>2000</creationdate><recordtype>book_chapter</recordtype><recordid>eNotkE1vFDEMhsOnupS9cxyJc4oTZzLJEVaFIhWBtEXiFmUyTndgd2ZJUqH992S2tSxb8vvash7G3gm4EgDdB-StAq7AGMuVE_IZW9vOYB2eZ-1zthJaCI6o7Av25iwIYVr5kq0AQXLbKXzNVhZbaUG3-oKtc_4NNVAqBL1iv-521FzHSKE0c2w281RoKuM8NTVL1bbB730_7sdyWgw__D3xTz7T0GznWP75VC27WofmGx3mdGq2p1zokN-yV9HvM62f-iX7-fn6bnPDb79_-br5eMuPCKZwDL3wMQxWWJLkByNktEJr61UcpOywtwO0cuhUAElEPZDCaIcQTa-1RLxk7x_vHn2ur8bkpzBmd0zjwaeT6zprYHFdPbpyFaZ7Sq6f5z_ZCXALaYeusnNnqm4hXRfw6Wya_z5QLo6WjVDhJL8PO38slLJD6ET9wgnjhEb8D4vEe6s</recordid><startdate>2000</startdate><enddate>2000</enddate><creator>de Lara, Eyal</creator><creator>Lu, Honghui</creator><creator>Charlie, Y.</creator><creator>Cox, Alan L.</creator><creator>Zwaenepoel, Willy</creator><general>Springer Berlin / Heidelberg</general><general>Springer Berlin Heidelberg</general><general>Springer</general><scope>FFUUA</scope><scope>IQODW</scope></search><sort><creationdate>2000</creationdate><title>The Effect of Contention on the Scalability of Page-Based Software Shared Memory Systems</title><author>de Lara, Eyal ; Lu, Honghui ; Charlie, Y. ; Cox, Alan L. ; Zwaenepoel, Willy</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-p308t-3cb1afcd919e2ead812f91669a4fd2273b9d052d74c02eeeb0e43f9dcf8b66233</frbrgroupid><rsrctype>book_chapters</rsrctype><prefilter>book_chapters</prefilter><language>eng</language><creationdate>2000</creationdate><topic>Applied sciences</topic><topic>Average Response Time</topic><topic>Computer science; control theory; systems</topic><topic>Exact sciences and technology</topic><topic>Execution Time</topic><topic>Home Node</topic><topic>Language processing and microprogramming</topic><topic>Load Imbalance</topic><topic>Shared Memory</topic><topic>Software</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>de Lara, Eyal</creatorcontrib><creatorcontrib>Lu, Honghui</creatorcontrib><creatorcontrib>Charlie, Y.</creatorcontrib><creatorcontrib>Cox, Alan L.</creatorcontrib><creatorcontrib>Zwaenepoel, Willy</creatorcontrib><collection>ProQuest Ebook Central - Book Chapters - Demo use only</collection><collection>Pascal-Francis</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext</fulltext></delivery><addata><au>de Lara, Eyal</au><au>Lu, Honghui</au><au>Charlie, Y.</au><au>Cox, Alan L.</au><au>Zwaenepoel, Willy</au><au>Dwarkadas, Sandhya</au><format>book</format><genre>bookitem</genre><ristype>CHAP</ristype><atitle>The Effect of Contention on the Scalability of Page-Based Software Shared Memory Systems</atitle><btitle>Languages, Compilers, and Run-Time Systems for Scalable Computers</btitle><seriestitle>Lecture Notes in Computer Science</seriestitle><date>2000</date><risdate>2000</risdate><volume>1915</volume><spage>155</spage><epage>169</epage><pages>155-169</pages><issn>0302-9743</issn><eissn>1611-3349</eissn><isbn>3540411852</isbn><isbn>9783540411857</isbn><eisbn>9783540408895</eisbn><eisbn>3540408894</eisbn><abstract>In this paper, we examine the causes and effects of contention for shared data access in parallel programs running on a software dis- tributed shared memory (DSM) system. Specifically, we experiment on two widely-used, page-based protocols, Princeton’s home-based lazy re- lease consistency (HLRC) and TreadMarks. For most of our programs, these protocols were equally affected by latency increases caused by con- tention and achieved similar performance. Where they differ significantly, HLRC’s ability to manually eliminate load imbalance was the largest fac- tor accounting for the difference. To quantify the effects of contention we either modified the application to eliminate the cause of the contention or modified the underlying protocol to efficiently handle it. Overall, we find that contention has profound effects on performance: eliminating contention reduced execution time by 64% in the most extreme case, even at the relatively modest scale of 32 nodes that we consider in this paper.</abstract><cop>Germany</cop><pub>Springer Berlin / Heidelberg</pub><doi>10.1007/3-540-40889-4_12</doi><oclcid>935290656</oclcid><tpages>15</tpages><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext |
| identifier | ISSN: 0302-9743 |
| ispartof | Languages, Compilers, and Run-Time Systems for Scalable Computers, 2000, Vol.1915, p.155-169 |
| issn | 0302-9743 1611-3349 |
| language | eng |
| recordid | cdi_pascalfrancis_primary_779803 |
| source | Springer Books |
| subjects | Applied sciences Average Response Time Computer science control theory systems Exact sciences and technology Execution Time Home Node Language processing and microprogramming Load Imbalance Shared Memory Software |
| title | The Effect of Contention on the Scalability of Page-Based Software Shared Memory Systems |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-11T02%3A32%3A13IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_pasca&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=bookitem&rft.atitle=The%20Effect%20of%20Contention%20on%20the%20Scalability%20of%20Page-Based%20Software%20Shared%20Memory%20Systems&rft.btitle=Languages,%20Compilers,%20and%20Run-Time%20Systems%20for%20Scalable%20Computers&rft.au=de%20Lara,%20Eyal&rft.date=2000&rft.volume=1915&rft.spage=155&rft.epage=169&rft.pages=155-169&rft.issn=0302-9743&rft.eissn=1611-3349&rft.isbn=3540411852&rft.isbn_list=9783540411857&rft_id=info:doi/10.1007/3-540-40889-4_12&rft_dat=%3Cproquest_pasca%3EEBC3071623_18_163%3C/proquest_pasca%3E%3Curl%3E%3C/url%3E&rft.eisbn=9783540408895&rft.eisbn_list=3540408894&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=EBC3071623_18_163&rft_id=info:pmid/&rfr_iscdi=true |