Unexpected Performance of Intel® Optane™ DC Persistent Memory
We evaluated Intel® Optane TM DC Persistent Memory and found that Intel's persistent memory is highly sensitive to data locality, size, and access patterns, which becomes clearer by optimizing both virtual memory page size and data layout for locality. Using the Polybench high-performance compu...
Gespeichert in:
| Veröffentlicht in: | IEEE computer architecture letters 2020-01, Vol.19 (1), p.55-58 |
|---|---|
| Hauptverfasser: | , , , |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 58 |
|---|---|
| container_issue | 1 |
| container_start_page | 55 |
| container_title | IEEE computer architecture letters |
| container_volume | 19 |
| creator | Mason, Tony Doudali, Thaleia Dimitra Seltzer, Margo Gavrilovska, Ada |
| description | We evaluated Intel® Optane TM DC Persistent Memory and found that Intel's persistent memory is highly sensitive to data locality, size, and access patterns, which becomes clearer by optimizing both virtual memory page size and data layout for locality. Using the Polybench high-performance computing benchmark suite and controlling for mapped page size, we evaluate persistent meemory (PMEM) performance relative to DRAM. In particular, the Linux PMEM support maps preferentially maps persistent memory in large pages while always mapping DRAM to small pages. We observed using large pages for PMEM and small pages for DRAM can create a 5x difference in performance, dwarfing other effects discussed in the literature. We found PMEM performance comparable to DRAM performance for the majority of tests when controlled for page size and optimized for data locality. |
| doi_str_mv | 10.1109/LCA.2020.2987303 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_proquest_journals_2408658089</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9072482</ieee_id><sourcerecordid>2408658089</sourcerecordid><originalsourceid>FETCH-LOGICAL-c333t-2e60ff2890a8f46716f9af716d12ab6711ec48da15c35de640eb6156b42c129d3</originalsourceid><addsrcrecordid>eNo9kEtKA0EQhhtRMEb3gpsB1xOrn-neGcZXIBIXZt10eqohIZkZuydg9p7EA3gIj-JJnJCQ1V8F318FHyHXFAaUgrmbFKMBAwYDZvSQAz8hPSqlyhUocXqcpTonFyktAYTiWvTI_azCzwZ9i2X2hjHUce0qj1kdsnHV4ur3J5s2ravw7-s7eyh2TFqkFqs2e8V1HbeX5Cy4VcKrQ_bJ7OnxvXjJJ9PncTGa5J5z3uYMFYTAtAGng1BDqoJxoYuSMjfvdope6NJR6bksUQnAuaJSzQXzlJmS98nt_m4T648NptYu602supeWCdBKatCmo2BP-VinFDHYJi7WLm4tBbvzZDtPdufJHjx1lZt9ZYGIR9zAkAnN-D--d2PZ</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>2408658089</pqid></control><display><type>article</type><title>Unexpected Performance of Intel® Optane™ DC Persistent Memory</title><source>IEEE Electronic Library (IEL)</source><creator>Mason, Tony ; Doudali, Thaleia Dimitra ; Seltzer, Margo ; Gavrilovska, Ada</creator><creatorcontrib>Mason, Tony ; Doudali, Thaleia Dimitra ; Seltzer, Margo ; Gavrilovska, Ada</creatorcontrib><description>We evaluated Intel® Optane TM DC Persistent Memory and found that Intel's persistent memory is highly sensitive to data locality, size, and access patterns, which becomes clearer by optimizing both virtual memory page size and data layout for locality. Using the Polybench high-performance computing benchmark suite and controlling for mapped page size, we evaluate persistent meemory (PMEM) performance relative to DRAM. In particular, the Linux PMEM support maps preferentially maps persistent memory in large pages while always mapping DRAM to small pages. We observed using large pages for PMEM and small pages for DRAM can create a 5x difference in performance, dwarfing other effects discussed in the literature. We found PMEM performance comparable to DRAM performance for the majority of tests when controlled for page size and optimized for data locality.</description><identifier>ISSN: 1556-6056</identifier><identifier>EISSN: 1556-6064</identifier><identifier>DOI: 10.1109/LCA.2020.2987303</identifier><identifier>CODEN: ICALC3</identifier><language>eng</language><publisher>New York: IEEE</publisher><subject>Dynamic random access memory ; File systems ; Linux ; Mapping ; Memory management ; Optimization ; Page layout ; Persistent memory ; Random access memory ; Resource management ; Virtual memory systems</subject><ispartof>IEEE computer architecture letters, 2020-01, Vol.19 (1), p.55-58</ispartof><rights>Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2020</rights><lds50>peer_reviewed</lds50><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c333t-2e60ff2890a8f46716f9af716d12ab6711ec48da15c35de640eb6156b42c129d3</citedby><cites>FETCH-LOGICAL-c333t-2e60ff2890a8f46716f9af716d12ab6711ec48da15c35de640eb6156b42c129d3</cites><orcidid>0000-0002-3197-839X ; 0000-0003-4199-2512 ; 0000-0002-0651-5019</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9072482$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,776,780,792,4010,27900,27901,27902,54733</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/9072482$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Mason, Tony</creatorcontrib><creatorcontrib>Doudali, Thaleia Dimitra</creatorcontrib><creatorcontrib>Seltzer, Margo</creatorcontrib><creatorcontrib>Gavrilovska, Ada</creatorcontrib><title>Unexpected Performance of Intel® Optane™ DC Persistent Memory</title><title>IEEE computer architecture letters</title><addtitle>LCA</addtitle><description>We evaluated Intel® Optane TM DC Persistent Memory and found that Intel's persistent memory is highly sensitive to data locality, size, and access patterns, which becomes clearer by optimizing both virtual memory page size and data layout for locality. Using the Polybench high-performance computing benchmark suite and controlling for mapped page size, we evaluate persistent meemory (PMEM) performance relative to DRAM. In particular, the Linux PMEM support maps preferentially maps persistent memory in large pages while always mapping DRAM to small pages. We observed using large pages for PMEM and small pages for DRAM can create a 5x difference in performance, dwarfing other effects discussed in the literature. We found PMEM performance comparable to DRAM performance for the majority of tests when controlled for page size and optimized for data locality.</description><subject>Dynamic random access memory</subject><subject>File systems</subject><subject>Linux</subject><subject>Mapping</subject><subject>Memory management</subject><subject>Optimization</subject><subject>Page layout</subject><subject>Persistent memory</subject><subject>Random access memory</subject><subject>Resource management</subject><subject>Virtual memory systems</subject><issn>1556-6056</issn><issn>1556-6064</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2020</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNo9kEtKA0EQhhtRMEb3gpsB1xOrn-neGcZXIBIXZt10eqohIZkZuydg9p7EA3gIj-JJnJCQ1V8F318FHyHXFAaUgrmbFKMBAwYDZvSQAz8hPSqlyhUocXqcpTonFyktAYTiWvTI_azCzwZ9i2X2hjHUce0qj1kdsnHV4ur3J5s2ravw7-s7eyh2TFqkFqs2e8V1HbeX5Cy4VcKrQ_bJ7OnxvXjJJ9PncTGa5J5z3uYMFYTAtAGng1BDqoJxoYuSMjfvdope6NJR6bksUQnAuaJSzQXzlJmS98nt_m4T648NptYu602supeWCdBKatCmo2BP-VinFDHYJi7WLm4tBbvzZDtPdufJHjx1lZt9ZYGIR9zAkAnN-D--d2PZ</recordid><startdate>202001</startdate><enddate>202001</enddate><creator>Mason, Tony</creator><creator>Doudali, Thaleia Dimitra</creator><creator>Seltzer, Margo</creator><creator>Gavrilovska, Ada</creator><general>IEEE</general><general>The Institute of Electrical and Electronics Engineers, Inc. (IEEE)</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>7SC</scope><scope>7SP</scope><scope>8FD</scope><scope>JQ2</scope><scope>L7M</scope><scope>L~C</scope><scope>L~D</scope><orcidid>https://orcid.org/0000-0002-3197-839X</orcidid><orcidid>https://orcid.org/0000-0003-4199-2512</orcidid><orcidid>https://orcid.org/0000-0002-0651-5019</orcidid></search><sort><creationdate>202001</creationdate><title>Unexpected Performance of Intel® Optane™ DC Persistent Memory</title><author>Mason, Tony ; Doudali, Thaleia Dimitra ; Seltzer, Margo ; Gavrilovska, Ada</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c333t-2e60ff2890a8f46716f9af716d12ab6711ec48da15c35de640eb6156b42c129d3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2020</creationdate><topic>Dynamic random access memory</topic><topic>File systems</topic><topic>Linux</topic><topic>Mapping</topic><topic>Memory management</topic><topic>Optimization</topic><topic>Page layout</topic><topic>Persistent memory</topic><topic>Random access memory</topic><topic>Resource management</topic><topic>Virtual memory systems</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Mason, Tony</creatorcontrib><creatorcontrib>Doudali, Thaleia Dimitra</creatorcontrib><creatorcontrib>Seltzer, Margo</creatorcontrib><creatorcontrib>Gavrilovska, Ada</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><collection>CrossRef</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics & Communications Abstracts</collection><collection>Technology Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><jtitle>IEEE computer architecture letters</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Mason, Tony</au><au>Doudali, Thaleia Dimitra</au><au>Seltzer, Margo</au><au>Gavrilovska, Ada</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Unexpected Performance of Intel® Optane™ DC Persistent Memory</atitle><jtitle>IEEE computer architecture letters</jtitle><stitle>LCA</stitle><date>2020-01</date><risdate>2020</risdate><volume>19</volume><issue>1</issue><spage>55</spage><epage>58</epage><pages>55-58</pages><issn>1556-6056</issn><eissn>1556-6064</eissn><coden>ICALC3</coden><abstract>We evaluated Intel® Optane TM DC Persistent Memory and found that Intel's persistent memory is highly sensitive to data locality, size, and access patterns, which becomes clearer by optimizing both virtual memory page size and data layout for locality. Using the Polybench high-performance computing benchmark suite and controlling for mapped page size, we evaluate persistent meemory (PMEM) performance relative to DRAM. In particular, the Linux PMEM support maps preferentially maps persistent memory in large pages while always mapping DRAM to small pages. We observed using large pages for PMEM and small pages for DRAM can create a 5x difference in performance, dwarfing other effects discussed in the literature. We found PMEM performance comparable to DRAM performance for the majority of tests when controlled for page size and optimized for data locality.</abstract><cop>New York</cop><pub>IEEE</pub><doi>10.1109/LCA.2020.2987303</doi><tpages>4</tpages><orcidid>https://orcid.org/0000-0002-3197-839X</orcidid><orcidid>https://orcid.org/0000-0003-4199-2512</orcidid><orcidid>https://orcid.org/0000-0002-0651-5019</orcidid><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 1556-6056 |
| ispartof | IEEE computer architecture letters, 2020-01, Vol.19 (1), p.55-58 |
| issn | 1556-6056 1556-6064 |
| language | eng |
| recordid | cdi_proquest_journals_2408658089 |
| source | IEEE Electronic Library (IEL) |
| subjects | Dynamic random access memory File systems Linux Mapping Memory management Optimization Page layout Persistent memory Random access memory Resource management Virtual memory systems |
| title | Unexpected Performance of Intel® Optane™ DC Persistent Memory |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-11T07%3A01%3A52IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Unexpected%20Performance%20of%20Intel%C2%AE%20Optane%E2%84%A2%20DC%20Persistent%20Memory&rft.jtitle=IEEE%20computer%20architecture%20letters&rft.au=Mason,%20Tony&rft.date=2020-01&rft.volume=19&rft.issue=1&rft.spage=55&rft.epage=58&rft.pages=55-58&rft.issn=1556-6056&rft.eissn=1556-6064&rft.coden=ICALC3&rft_id=info:doi/10.1109/LCA.2020.2987303&rft_dat=%3Cproquest_RIE%3E2408658089%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=2408658089&rft_id=info:pmid/&rft_ieee_id=9072482&rfr_iscdi=true |