{"id":"https://openalex.org/W4414198422","doi":"https://doi.org/10.1109/dac63849.2025.11132724","title":"NVR: Vector Runahead on NPUs for Sparse Memory Access","display_name":"NVR: Vector Runahead on NPUs for Sparse Memory Access","publication_year":2025,"publication_date":"2025-06-22","ids":{"openalex":"https://openalex.org/W4414198422","doi":"https://doi.org/10.1109/dac63849.2025.11132724"},"language":"en","primary_location":{"id":"doi:10.1109/dac63849.2025.11132724","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac63849.2025.11132724","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 62nd ACM/IEEE Design Automation Conference (DAC)","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/A5004478535","display_name":"Hui Wang","orcid":"https://orcid.org/0000-0002-3948-4415"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Hui Wang","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010501555","display_name":"Zhengpeng Zhao","orcid":"https://orcid.org/0000-0003-1347-1680"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhengpeng Zhao","raw_affiliation_strings":["Huazhong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100452376","display_name":"Jing Wang","orcid":"https://orcid.org/0009-0005-5089-9987"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Jing Wang","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101992422","display_name":"Yingying Du","orcid":"https://orcid.org/0000-0002-5149-4632"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Yushu Du","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100582162","display_name":"Yuan Cheng","orcid":"https://orcid.org/0000-0003-1830-7951"},"institutions":[{"id":"https://openalex.org/I4210091958","display_name":"Hanoi Open University","ror":"https://ror.org/00gjkwr85","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210091958"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Yuan Cheng","raw_affiliation_strings":["Houmo AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI","institution_ids":["https://openalex.org/I4210091958"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101795051","display_name":"Bing Guo","orcid":"https://orcid.org/0000-0002-3998-5981"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Bing Guo","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111004576","display_name":"Xiao He","orcid":"https://orcid.org/0000-0002-1065-4927"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"He Xiao","raw_affiliation_strings":["Harbin Institute of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Harbin Institute of Technology","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101827445","display_name":"Congcong Ma","orcid":"https://orcid.org/0009-0008-2701-9875"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Chenhao Ma","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030201424","display_name":"Xiaomeng Han","orcid":null},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Xiaomeng Han","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5116349631","display_name":"Dean You","orcid":null},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Dean You","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104243916","display_name":"Jiapeng Guan","orcid":null},"institutions":[{"id":"https://openalex.org/I27357992","display_name":"Dalian University of Technology","ror":"https://ror.org/023hj5876","country_code":"CN","type":"education","lineage":["https://openalex.org/I27357992"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiapeng Guan","raw_affiliation_strings":["Dalian University of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dalian University of Technology","institution_ids":["https://openalex.org/I27357992"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077319702","display_name":"Ran Wei","orcid":"https://orcid.org/0000-0003-2178-5014"},"institutions":[{"id":"https://openalex.org/I27357992","display_name":"Dalian University of Technology","ror":"https://ror.org/023hj5876","country_code":"CN","type":"education","lineage":["https://openalex.org/I27357992"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ran Wei","raw_affiliation_strings":["Dalian University of Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Dalian University of Technology","institution_ids":["https://openalex.org/I27357992"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100917814","display_name":"Dawei Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210091958","display_name":"Hanoi Open University","ror":"https://ror.org/00gjkwr85","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210091958"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Dawei Yang","raw_affiliation_strings":["Houmo AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI","institution_ids":["https://openalex.org/I4210091958"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5110656023","display_name":"Zhe Jiang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Zhe Jiang","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":14,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.23131673,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9952999949455261,"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":0.9952999949455261,"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/T12326","display_name":"Network Packet Processing and Optimization","score":0.9433000087738037,"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.9348000288009644,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/cache","display_name":"Cache","score":0.8138999938964844},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.6226999759674072},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.6078000068664551},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5586000084877014},{"id":"https://openalex.org/keywords/cache-only-memory-architecture","display_name":"Cache-only memory architecture","score":0.5116999745368958},{"id":"https://openalex.org/keywords/cpu-cache","display_name":"CPU cache","score":0.49380001425743103},{"id":"https://openalex.org/keywords/cache-algorithms","display_name":"Cache algorithms","score":0.46459999680519104},{"id":"https://openalex.org/keywords/cache-pollution","display_name":"Cache pollution","score":0.4293999969959259},{"id":"https://openalex.org/keywords/compiler","display_name":"Compiler","score":0.413100004196167}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8400999903678894},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.8138999938964844},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.7426000237464905},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.6226999759674072},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.6078000068664551},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5586000084877014},{"id":"https://openalex.org/C3720319","wikidata":"https://www.wikidata.org/wiki/Q5015937","display_name":"Cache-only memory architecture","level":5,"score":0.5116999745368958},{"id":"https://openalex.org/C189783530","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"CPU cache","level":3,"score":0.49380001425743103},{"id":"https://openalex.org/C38556500","wikidata":"https://www.wikidata.org/wiki/Q13404475","display_name":"Cache algorithms","level":4,"score":0.46459999680519104},{"id":"https://openalex.org/C113166858","wikidata":"https://www.wikidata.org/wiki/Q5015981","display_name":"Cache pollution","level":5,"score":0.4293999969959259},{"id":"https://openalex.org/C169590947","wikidata":"https://www.wikidata.org/wiki/Q47506","display_name":"Compiler","level":2,"score":0.413100004196167},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.41190001368522644},{"id":"https://openalex.org/C201148951","wikidata":"https://www.wikidata.org/wiki/Q5015976","display_name":"Cache coloring","level":4,"score":0.3546000123023987},{"id":"https://openalex.org/C167713795","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"Smart Cache","level":5,"score":0.3384999930858612},{"id":"https://openalex.org/C190902152","wikidata":"https://www.wikidata.org/wiki/Q1325106","display_name":"Optimizing compiler","level":3,"score":0.31779998540878296},{"id":"https://openalex.org/C176649486","wikidata":"https://www.wikidata.org/wiki/Q2308807","display_name":"Memory management","level":3,"score":0.3165000081062317},{"id":"https://openalex.org/C59687516","wikidata":"https://www.wikidata.org/wiki/Q5015938","display_name":"Cache-oblivious algorithm","level":5,"score":0.31279999017715454},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.29660001397132874},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.29030001163482666},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.2712000012397766},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.26409998536109924},{"id":"https://openalex.org/C2779602883","wikidata":"https://www.wikidata.org/wiki/Q15544750","display_name":"Memory architecture","level":2,"score":0.2630000114440918},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.25360000133514404},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.2522999942302704},{"id":"https://openalex.org/C133371097","wikidata":"https://www.wikidata.org/wiki/Q868014","display_name":"Non-uniform memory access","level":5,"score":0.2513999938964844}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/dac63849.2025.11132724","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac63849.2025.11132724","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 62nd ACM/IEEE Design Automation Conference (DAC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320324856","display_name":"Southeast University","ror":"https://ror.org/04ct4d772"},{"id":"https://openalex.org/F4320337504","display_name":"Research and Development","ror":"https://ror.org/027s68j25"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W201215413","https://openalex.org/W1964471912","https://openalex.org/W2030812654","https://openalex.org/W2101738713","https://openalex.org/W2123608497","https://openalex.org/W2166620913","https://openalex.org/W2234355962","https://openalex.org/W2234679013","https://openalex.org/W2289252105","https://openalex.org/W2329976284","https://openalex.org/W2516141709","https://openalex.org/W2748573297","https://openalex.org/W2850910281","https://openalex.org/W2889895098","https://openalex.org/W2899244816","https://openalex.org/W2980113464","https://openalex.org/W3011338798","https://openalex.org/W3016777042","https://openalex.org/W3131922516","https://openalex.org/W3146366485","https://openalex.org/W3147011727","https://openalex.org/W3194960030","https://openalex.org/W3213528054","https://openalex.org/W4237150160","https://openalex.org/W4360606782","https://openalex.org/W4388739256","https://openalex.org/W4391136507","https://openalex.org/W4394945474","https://openalex.org/W4401211642"],"related_works":[],"abstract_inverted_index":{"Deep":[0],"Neural":[1],"Networks":[2],"are":[3],"increasingly":[4],"leveraging":[5],"sparsity":[6,21],"to":[7,27,33,52,76,125],"reduce":[8],"the":[9,77,107,146,156,176,182],"scaling":[10],"up":[11],"of":[12,80,148],"model":[13],"parameter":[14],"size.":[15],"However,":[16],"reducing":[17],"wall-clock":[18],"time":[19],"through":[20],"and":[22,69],"pruning":[23],"remains":[24],"challenging":[25],"due":[26],"irregular":[28],"memory":[29,64],"access":[30],"patterns,":[31],"leading":[32],"frequent":[34],"cache":[35,54,122,152,168,179],"misses.":[36],"In":[37],"this":[38,166],"paper,":[39],"we":[40,144],"present":[41],"NPU":[42,157],"Vector":[43],"Runahead":[44],"(NVR),":[45],"a":[46,84,100,150],"prefetching":[47,127],"mechanism":[48],"tailored":[49],"for":[50,88],"NPUs":[51,140],"address":[53],"miss":[55],"problems":[56],"in":[57,121,128],"sparse":[58,89,137],"DNN":[59,90],"workloads.":[60],"Rather":[61],"than":[62,174],"optimising":[63],"patterns":[65],"with":[66,109,159],"high":[67],"overhead":[68,112],"poor":[70],"portability,":[71],"NVR":[72,82,115],"adapts":[73],"runahead":[74],"execution":[75],"unique":[78],"architecture":[79],"NPUs.":[81],"provides":[83],"general":[85],"micro-architectural":[86],"solution":[87],"workloads":[91,138],"without":[92,141],"requiring":[93],"compiler":[94],"or":[95],"algorithmic":[96],"support,":[97],"operating":[98],"as":[99],"decoupled,":[101],"speculative,":[102],"lightweight":[103],"hardware":[104,111],"sub-thread":[105],"alongside":[106],"NPU,":[108],"minimal":[110],"(under":[113],"5%).":[114],"achieves":[116],"an":[117],"average":[118,134],"90%":[119],"reduction":[120],"misses":[123],"compared":[124],"SOTA":[126],"general-purpose":[129],"processors,":[130],"delivering":[131],"4":[132],"x":[133],"speedup":[135],"on":[136],"versus":[139],"prefetching.":[142],"Moreover,":[143],"investigate":[145],"advantages":[147],"incorporating":[149],"small":[151],"(16":[153],"KB)":[154],"into":[155],"combined":[158],"NVR.":[160],"Our":[161],"evaluation":[162],"shows":[163],"that":[164],"expanding":[165],"modest":[167],"delivers":[169],"5x":[170],"higher":[171],"performance":[172],"benefits":[173],"increasing":[175],"$\\mathbf{L":[177],"2}$":[178],"size":[180],"by":[181],"same":[183],"amount.":[184]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}
