{"id":"https://openalex.org/W4411600395","doi":"https://doi.org/10.1109/lca.2025.3553143","title":"SAFE: Sharing-Aware Prefetching for Efficient GPU Memory Management With Unified Virtual Memory","display_name":"SAFE: Sharing-Aware Prefetching for Efficient GPU Memory Management With Unified Virtual Memory","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4411600395","doi":"https://doi.org/10.1109/lca.2025.3553143"},"language":"en","primary_location":{"id":"doi:10.1109/lca.2025.3553143","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lca.2025.3553143","pdf_url":null,"source":{"id":"https://openalex.org/S17643076","display_name":"IEEE Computer Architecture Letters","issn_l":"1556-6056","issn":["1556-6056","1556-6064","2473-2575"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Computer Architecture Letters","raw_type":"journal-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/A5100889264","display_name":"Hyunkyun Shin","orcid":null},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyunkyun Shin","raw_affiliation_strings":["Yonsei University, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0000-3105-7286","affiliations":[{"raw_affiliation_string":"Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112058052","display_name":"S.H. Bang","orcid":"https://orcid.org/0000-0002-2928-1403"},"institutions":[{"id":"https://openalex.org/I193352282","display_name":"Daegu Gyeongbuk Institute of Science and Technology","ror":"https://ror.org/03frjya69","country_code":"KR","type":"education","lineage":["https://openalex.org/I193352282"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Seongtae Bang","raw_affiliation_strings":["Daegu-Gyeongbuk Institute of Science and Technology, Daegu, South Korea"],"raw_orcid":"https://orcid.org/0000-0002-2928-1403","affiliations":[{"raw_affiliation_string":"Daegu-Gyeongbuk Institute of Science and Technology, Daegu, South Korea","institution_ids":["https://openalex.org/I193352282"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5039453868","display_name":"Hyoungwon Park","orcid":"https://orcid.org/0000-0002-7614-6415"},"institutions":[{"id":"https://openalex.org/I193352282","display_name":"Daegu Gyeongbuk Institute of Science and Technology","ror":"https://ror.org/03frjya69","country_code":"KR","type":"education","lineage":["https://openalex.org/I193352282"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyungwon Park","raw_affiliation_strings":["Daegu-Gyeongbuk Institute of Science and Technology, Daegu, South Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Daegu-Gyeongbuk Institute of Science and Technology, Daegu, South Korea","institution_ids":["https://openalex.org/I193352282"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101885459","display_name":"Daehoon Kim","orcid":"https://orcid.org/0000-0003-0837-0877"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Daehoon Kim","raw_affiliation_strings":["Yonsei University, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0000-0003-0837-0877","affiliations":[{"raw_affiliation_string":"Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.377,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.78484349,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":"24","issue":"1","first_page":"117","last_page":"120"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9990000128746033,"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":0.9990000128746033,"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.9983999729156494,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.9977999925613403,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8551714420318604},{"id":"https://openalex.org/keywords/memory-management","display_name":"Memory management","score":0.6583822965621948},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.6194170713424683},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.5553933382034302},{"id":"https://openalex.org/keywords/storage-management","display_name":"Storage management","score":0.5419365763664246},{"id":"https://openalex.org/keywords/virtual-memory","display_name":"Virtual memory","score":0.42102959752082825},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.29395031929016113},{"id":"https://openalex.org/keywords/overlay","display_name":"Overlay","score":0.2839643955230713}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8551714420318604},{"id":"https://openalex.org/C176649486","wikidata":"https://www.wikidata.org/wiki/Q2308807","display_name":"Memory management","level":3,"score":0.6583822965621948},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.6194170713424683},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.5553933382034302},{"id":"https://openalex.org/C2984984529","wikidata":"https://www.wikidata.org/wiki/Q7619925","display_name":"Storage management","level":2,"score":0.5419365763664246},{"id":"https://openalex.org/C76399640","wikidata":"https://www.wikidata.org/wiki/Q189401","display_name":"Virtual memory","level":4,"score":0.42102959752082825},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.29395031929016113},{"id":"https://openalex.org/C136085584","wikidata":"https://www.wikidata.org/wiki/Q910289","display_name":"Overlay","level":2,"score":0.2839643955230713}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/lca.2025.3553143","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lca.2025.3553143","pdf_url":null,"source":{"id":"https://openalex.org/S17643076","display_name":"IEEE Computer Architecture Letters","issn_l":"1556-6056","issn":["1556-6056","1556-6064","2473-2575"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Computer Architecture Letters","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","score":0.7599999904632568,"display_name":"Affordable and clean energy"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":10,"referenced_works":["https://openalex.org/W2080592089","https://openalex.org/W2128120785","https://openalex.org/W2934021599","https://openalex.org/W3011293047","https://openalex.org/W3043528907","https://openalex.org/W4221032334","https://openalex.org/W4318541551","https://openalex.org/W4381894554","https://openalex.org/W6763585923","https://openalex.org/W6781267947"],"related_works":["https://openalex.org/W1976766385","https://openalex.org/W2078640694","https://openalex.org/W261562921","https://openalex.org/W1421493983","https://openalex.org/W4243333834","https://openalex.org/W1698699620","https://openalex.org/W71081774","https://openalex.org/W2138847","https://openalex.org/W2373327185","https://openalex.org/W2051837050"],"abstract_inverted_index":{"As":[0],"the":[1,54,90,98,106,116,124,145,149,158,162],"demand":[2],"for":[3,175,207],"GPU":[4,17,32],"memory":[5,18,33,44,103,151,163,212],"from":[6],"applications":[7],"such":[8],"as":[9],"machine":[10],"learning":[11],"continues":[12],"to":[13,41,198],"grow":[14],"exponentially,":[15],"maximizing":[16],"capacity":[19],"has":[20],"become":[21],"increasingly":[22],"important.":[23],"Unified":[24],"Virtual":[25],"Memory":[26],"(UVM),":[27],"which":[28,112,138],"combines":[29],"host":[30,61],"and":[31,80,170,184],"into":[34],"a":[35,199],"unified":[36,167],"address":[37],"space,":[38],"allows":[39],"GPUs":[40],"utilize":[42],"more":[43],"than":[45],"their":[46],"physical":[47],"capacity.":[48],"However,":[49],"this":[50,85],"advantage":[51],"comes":[52],"at":[53],"cost":[55],"of":[56,92,101,110,118,148,161,219],"significant":[57],"overheads":[58],"when":[59,75],"accessing":[60],"memory.":[62],"Although":[63],"existing":[64],"prefetching":[65,141,173],"techniques":[66],"help":[67],"alleviate":[68],"these":[69],"overheads,":[70],"they":[71],"still":[72],"encounter":[73],"challenges":[74],"dealing":[76],"with":[77,97,209,215],"irregular":[78,211],"workloads":[79,93,208],"dynamic":[81],"mixed":[82],"workloads.":[83],"In":[84,120],"paper,":[86],"we":[87,122],"demonstrate":[88],"that":[89,192],"regularity":[91],"is":[94],"strongly":[95],"correlated":[96],"sharing":[99,146,159],"status":[100,147,160],"UVM":[102,204],"blocks":[104,164],"among":[105],"Streaming":[107],"Multiprocessors":[108],"(SMs)":[109],"GPUs,":[111],"in":[113],"turn":[114],"impacts":[115],"effectiveness":[117],"prefetching.":[119],"addition,":[121],"propose":[123],"<bold":[125,128],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[126,129,132,136,154,194],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">S</b>haring":[127],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">A</b>ware":[130],"pre<bold":[131],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">FE</b>tching":[133],"technique,":[134],"<monospace":[135,153,193],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">SAFE</monospace>,":[137],"dynamically":[139],"adjusts":[140],"strategies":[142],"based":[143],"on":[144],"accessed":[150],"blocks.":[152],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">SAFE</monospace>":[155,195],"efficiently":[156],"tracks":[157],"by":[165],"leveraging":[166],"TLBs":[168],"(uTLBs)":[169],"enforces":[171],"tailored":[172],"configurations":[174],"each":[176],"block.":[177],"This":[178],"approach":[179],"requires":[180],"no":[181],"hardware":[182],"modifications":[183],"incurs":[185],"negligible":[186],"performance":[187,201],"overhead.":[188],"Our":[189],"evaluation":[190],"shows":[191],"achieves":[196],"up":[197],"6.5\u00d7":[200],"improvement":[202,218],"over":[203],"default":[205],"prefetcher":[206],"predominantly":[210],"access":[213],"patterns,":[214],"an":[216],"average":[217],"3.6\u00d7.":[220]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}
