{"id":"https://openalex.org/W2949486320","doi":"https://doi.org/10.1145/3307650.3322224","title":"Interplay between hardware prefetcher and page eviction policy in CPU-GPU unified virtual memory","display_name":"Interplay between hardware prefetcher and page eviction policy in CPU-GPU unified virtual memory","publication_year":2019,"publication_date":"2019-06-14","ids":{"openalex":"https://openalex.org/W2949486320","doi":"https://doi.org/10.1145/3307650.3322224","mag":"2949486320"},"language":"en","primary_location":{"id":"doi:10.1145/3307650.3322224","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3307650.3322224","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 46th International Symposium on Computer Architecture","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5061926291","display_name":"Debashis Ganguly","orcid":null},"institutions":[{"id":"https://openalex.org/I170201317","display_name":"University of Pittsburgh","ror":"https://ror.org/01an3r305","country_code":"US","type":"education","lineage":["https://openalex.org/I170201317"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Debashis Ganguly","raw_affiliation_strings":["University of Pittsburgh"],"affiliations":[{"raw_affiliation_string":"University of Pittsburgh","institution_ids":["https://openalex.org/I170201317"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100706053","display_name":"Ziyu Zhang","orcid":"https://orcid.org/0000-0003-2495-4880"},"institutions":[{"id":"https://openalex.org/I170201317","display_name":"University of Pittsburgh","ror":"https://ror.org/01an3r305","country_code":"US","type":"education","lineage":["https://openalex.org/I170201317"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ziyu Zhang","raw_affiliation_strings":["University of Pittsburgh"],"affiliations":[{"raw_affiliation_string":"University of Pittsburgh","institution_ids":["https://openalex.org/I170201317"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101605460","display_name":"Jun Yang","orcid":"https://orcid.org/0000-0001-8372-6541"},"institutions":[{"id":"https://openalex.org/I170201317","display_name":"University of Pittsburgh","ror":"https://ror.org/01an3r305","country_code":"US","type":"education","lineage":["https://openalex.org/I170201317"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jun Yang","raw_affiliation_strings":["University of Pittsburgh"],"affiliations":[{"raw_affiliation_string":"University of Pittsburgh","institution_ids":["https://openalex.org/I170201317"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5038997010","display_name":"Rami Melhem","orcid":"https://orcid.org/0000-0001-6403-5446"},"institutions":[{"id":"https://openalex.org/I170201317","display_name":"University of Pittsburgh","ror":"https://ror.org/01an3r305","country_code":"US","type":"education","lineage":["https://openalex.org/I170201317"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Rami Melhem","raw_affiliation_strings":["University of Pittsburgh"],"affiliations":[{"raw_affiliation_string":"University of Pittsburgh","institution_ids":["https://openalex.org/I170201317"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5061926291"],"corresponding_institution_ids":["https://openalex.org/I170201317"],"apc_list":null,"apc_paid":null,"fwci":10.3685,"has_fulltext":false,"cited_by_count":81,"citation_normalized_percentile":{"value":0.9891153,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"224","last_page":"235"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11181","display_name":"Advanced Data Storage Technologies","score":0.9998999834060669,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10101","display_name":"Cloud Computing and Resource Management","score":0.9994000196456909,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8650370836257935},{"id":"https://openalex.org/keywords/virtual-memory","display_name":"Virtual memory","score":0.5783432722091675},{"id":"https://openalex.org/keywords/demand-paging","display_name":"Demand paging","score":0.5263441801071167},{"id":"https://openalex.org/keywords/instruction-prefetch","display_name":"Instruction prefetch","score":0.5200102925300598},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.49576887488365173},{"id":"https://openalex.org/keywords/cache","display_name":"Cache","score":0.47131258249282837},{"id":"https://openalex.org/keywords/page-fault","display_name":"Page fault","score":0.4653419256210327},{"id":"https://openalex.org/keywords/memory-bandwidth","display_name":"Memory bandwidth","score":0.4514361023902893},{"id":"https://openalex.org/keywords/interleaved-memory","display_name":"Interleaved memory","score":0.42155206203460693},{"id":"https://openalex.org/keywords/cas-latency","display_name":"CAS latency","score":0.41900426149368286},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.396106481552124},{"id":"https://openalex.org/keywords/memory-controller","display_name":"Memory controller","score":0.3859838545322418},{"id":"https://openalex.org/keywords/memory-management","display_name":"Memory management","score":0.3791006803512573},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.3763887584209442},{"id":"https://openalex.org/keywords/overlay","display_name":"Overlay","score":0.16439548134803772},{"id":"https://openalex.org/keywords/semiconductor-memory","display_name":"Semiconductor memory","score":0.16001784801483154}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8650370836257935},{"id":"https://openalex.org/C76399640","wikidata":"https://www.wikidata.org/wiki/Q189401","display_name":"Virtual memory","level":4,"score":0.5783432722091675},{"id":"https://openalex.org/C188873839","wikidata":"https://www.wikidata.org/wiki/Q5255045","display_name":"Demand paging","level":5,"score":0.5263441801071167},{"id":"https://openalex.org/C133588205","wikidata":"https://www.wikidata.org/wiki/Q28455645","display_name":"Instruction prefetch","level":3,"score":0.5200102925300598},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.49576887488365173},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.47131258249282837},{"id":"https://openalex.org/C193343404","wikidata":"https://www.wikidata.org/wiki/Q1928607","display_name":"Page fault","level":5,"score":0.4653419256210327},{"id":"https://openalex.org/C188045654","wikidata":"https://www.wikidata.org/wiki/Q17148339","display_name":"Memory bandwidth","level":2,"score":0.4514361023902893},{"id":"https://openalex.org/C63511323","wikidata":"https://www.wikidata.org/wiki/Q908936","display_name":"Interleaved memory","level":4,"score":0.42155206203460693},{"id":"https://openalex.org/C189930140","wikidata":"https://www.wikidata.org/wiki/Q1112878","display_name":"CAS latency","level":4,"score":0.41900426149368286},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.396106481552124},{"id":"https://openalex.org/C100800780","wikidata":"https://www.wikidata.org/wiki/Q1175867","display_name":"Memory controller","level":3,"score":0.3859838545322418},{"id":"https://openalex.org/C176649486","wikidata":"https://www.wikidata.org/wiki/Q2308807","display_name":"Memory management","level":3,"score":0.3791006803512573},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.3763887584209442},{"id":"https://openalex.org/C136085584","wikidata":"https://www.wikidata.org/wiki/Q910289","display_name":"Overlay","level":2,"score":0.16439548134803772},{"id":"https://openalex.org/C98986596","wikidata":"https://www.wikidata.org/wiki/Q1143031","display_name":"Semiconductor memory","level":2,"score":0.16001784801483154}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3307650.3322224","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3307650.3322224","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 46th International Symposium on Computer Architecture","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.7099999785423279,"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":20,"referenced_works":["https://openalex.org/W1979527452","https://openalex.org/W1983235612","https://openalex.org/W2056816375","https://openalex.org/W2061812855","https://openalex.org/W2080592089","https://openalex.org/W2087866311","https://openalex.org/W2100537799","https://openalex.org/W2100926301","https://openalex.org/W2102843684","https://openalex.org/W2107349848","https://openalex.org/W2120138088","https://openalex.org/W2128120785","https://openalex.org/W2142444503","https://openalex.org/W2330672121","https://openalex.org/W2613639279","https://openalex.org/W2764065518","https://openalex.org/W2934021599","https://openalex.org/W3022971582","https://openalex.org/W4249082578","https://openalex.org/W4251259849"],"related_works":["https://openalex.org/W2407941772","https://openalex.org/W2165601128","https://openalex.org/W1437641643","https://openalex.org/W2170496111","https://openalex.org/W1480691199","https://openalex.org/W4360831822","https://openalex.org/W1981606011","https://openalex.org/W104390747","https://openalex.org/W3194276159","https://openalex.org/W4254597226"],"abstract_inverted_index":{"Memory":[0,54],"capacity":[1],"in":[2,82,103,163],"GPGPUs":[3],"is":[4,171],"a":[5,19,27],"major":[6],"challenge":[7],"for":[8],"data-intensive":[9],"applications":[10],"with":[11,147,240],"their":[12],"ever":[13],"increasing":[14],"memory":[15,25,74,86,170,268],"requirement.":[16],"To":[17,45,136],"fit":[18],"workload":[20,34],"into":[21],"the":[22,33,37,47,66,73,119,123,168,192,202,216,235,241,267,278],"limited":[23],"GPU":[24,85,116,164,169,272],"space,":[26],"programmer":[28,48],"needs":[29],"to":[30,58,65,115,131,151,160,173,181,185,254,276],"manually":[31],"divide":[32],"by":[35,77],"tiling":[36],"working":[38],"set":[39],"and":[40,62,222,228,249,258],"perform":[41],"user-level":[42],"data":[43,112],"migration.":[44],"relieve":[46],"from":[49,113],"this":[50,137],"burden,":[51],"Unified":[52],"Virtual":[53],"(UVM)":[55],"was":[56],"developed":[57],"support":[59],"on-demand":[60],"paging":[61],"migration,":[63],"transparent":[64],"user.":[67],"It":[68],"further":[69,265],"takes":[70],"care":[71],"of":[72,94,100,111,194,201,218,247,271],"over-subscription":[75],"issue":[76],"automatically":[78],"performing":[79],"page":[80,95,120,260],"replacement":[81,261],"an":[83,156,245],"oversubscribed":[84],"situation.":[87],"However,":[88,166],"we":[89,106,139],"found":[90],"that":[91,108,153,198,233],"na\u00efve":[92],"handling":[93],"faults":[96],"can":[97,117,129],"cause":[98],"orders":[99],"magnitude":[101],"slowdown":[102],"performance.":[104],"Moreover,":[105],"observed":[107],"although":[109],"prefetching":[110,127,177],"CPU":[114],"hide":[118],"fault":[121],"latency,":[122],"difference":[124],"among":[125],"various":[126],"mechanisms":[128],"lead":[130],"drastically":[132],"different":[133],"performance":[134,229,251,280],"results.":[135],"end,":[138],"performed":[140],"extensive":[141],"experiments":[142],"on":[143],"GeForceGTX":[144],"1080ti":[145],"GPUs":[146],"PCI-e":[148],"3.0":[149],"16x":[150],"discover":[152],"there":[154],"exists":[155],"effective":[157],"prefetch":[158],"mechanism":[159,178],"enhance":[161],"locality":[162,186],"memory.":[165],"as":[167],"filled":[172],"its":[174],"capacity,":[175],"such":[176],"quickly":[179],"proves":[180],"be":[182],"counterproductive":[183],"due":[184],"unaware":[187],"eviction":[188,196],"policy.":[189],"This":[190],"necessitates":[191],"design":[193],"new":[195,209],"policies":[197,213],"are":[199],"aware":[200],"hardware":[203,220,242],"prefetcher":[204,221,243],"semantics.":[205],"We":[206,231,264],"propose":[207],"two":[208],"programmer-agnostic,":[210],"locality-aware":[211],"pre-eviction":[212,238],"which":[214],"leverage":[215],"mechanics":[217],"existing":[219],"thus":[223],"incur":[224],"no":[225],"additional":[226],"implementation":[227],"overhead.":[230],"demonstrate":[232],"combining":[234],"proposed":[236],"tree-based":[237],"policy":[239],"provides":[244],"average":[246],"93%":[248],"18.5%":[250],"speed-up":[252],"compared":[253],"LRU":[255],"based":[256],"4KB":[257],"2MB":[259],"strategies,":[262],"respectively.":[263],"examine":[266],"access":[269],"pattern":[270],"workloads":[273],"under":[274],"consideration":[275],"analyze":[277],"achieved":[279],"speed-up.":[281]},"counts_by_year":[{"year":2026,"cited_by_count":4},{"year":2025,"cited_by_count":15},{"year":2024,"cited_by_count":9},{"year":2023,"cited_by_count":11},{"year":2022,"cited_by_count":9},{"year":2021,"cited_by_count":14},{"year":2020,"cited_by_count":14},{"year":2019,"cited_by_count":5}],"updated_date":"2026-04-11T08:14:18.477133","created_date":"2025-10-10T00:00:00"}
