{"id":"https://openalex.org/W4408925909","doi":"https://doi.org/10.1145/3676641.3716009","title":"PAPI: Exploiting Dynamic Parallelism in Large Language Model Decoding with a Processing-In-Memory-Enabled Computing System","display_name":"PAPI: Exploiting Dynamic Parallelism in Large Language Model Decoding with a Processing-In-Memory-Enabled Computing System","publication_year":2025,"publication_date":"2025-03-27","ids":{"openalex":"https://openalex.org/W4408925909","doi":"https://doi.org/10.1145/3676641.3716009"},"language":"en","primary_location":{"id":"doi:10.1145/3676641.3716009","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3676641.3716009","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3676641.3716009","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5090389605","display_name":"Yintao He","orcid":"https://orcid.org/0000-0003-3054-0617"},"institutions":[{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]},{"id":"https://openalex.org/I4210165038","display_name":"University of Chinese Academy of Sciences","ror":"https://ror.org/05qbk4x57","country_code":"CN","type":"education","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yintao He","raw_affiliation_strings":["SKLP, Institute of Computing Technology, CAS, Beijing, China and University of Chinese Academy of Sciences, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-3054-0617","affiliations":[{"raw_affiliation_string":"SKLP, Institute of Computing Technology, CAS, Beijing, China and University of Chinese Academy of Sciences, Beijing, China","institution_ids":["https://openalex.org/I4210090176","https://openalex.org/I4210165038"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008354867","display_name":"Haiyu Mao","orcid":"https://orcid.org/0000-0002-7393-4504"},"institutions":[{"id":"https://openalex.org/I183935753","display_name":"King's College London","ror":"https://ror.org/0220mzb33","country_code":"GB","type":"education","lineage":["https://openalex.org/I124357947","https://openalex.org/I183935753"]},{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH","GB"],"is_corresponding":false,"raw_author_name":"Haiyu Mao","raw_affiliation_strings":["King's College London, London, United Kingdom and ETH Z\u00fcrich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-7393-4504","affiliations":[{"raw_affiliation_string":"King's College London, London, United Kingdom and ETH Z\u00fcrich, Zurich, Switzerland","institution_ids":["https://openalex.org/I183935753","https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044162748","display_name":"Christina Giannoula","orcid":"https://orcid.org/0000-0003-0162-4547"},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]},{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]},{"id":"https://openalex.org/I4210127509","display_name":"Vector Institute","ror":"https://ror.org/03kqdja62","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210127509"]}],"countries":["CA","CH"],"is_corresponding":false,"raw_author_name":"Christina Giannoula","raw_affiliation_strings":["University of Toronto, Toronto, Canada, Vector Institute, Toronto, Canada, and ETH Z\u00fcrich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0003-0162-4547","affiliations":[{"raw_affiliation_string":"University of Toronto, Toronto, Canada, Vector Institute, Toronto, Canada, and ETH Z\u00fcrich, Zurich, Switzerland","institution_ids":["https://openalex.org/I185261750","https://openalex.org/I35440088","https://openalex.org/I4210127509"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008240365","display_name":"Mohammad Sadrosadati","orcid":"https://orcid.org/0000-0002-4029-0175"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Mohammad Sadrosadati","raw_affiliation_strings":["ETH Z\u00fcrich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-4029-0175","affiliations":[{"raw_affiliation_string":"ETH Z\u00fcrich, Zurich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044416322","display_name":"Juan G\u00f3mez-Luna","orcid":"https://orcid.org/0000-0002-6514-1571"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Juan G\u00f3mez-Luna","raw_affiliation_strings":["NVIDIA, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-6514-1571","affiliations":[{"raw_affiliation_string":"NVIDIA, Zurich, Switzerland","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100768288","display_name":"Huawei Li","orcid":"https://orcid.org/0000-0001-8082-4218"},"institutions":[{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]},{"id":"https://openalex.org/I4210165038","display_name":"University of Chinese Academy of Sciences","ror":"https://ror.org/05qbk4x57","country_code":"CN","type":"education","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huawei Li","raw_affiliation_strings":["SKLP, Institute of Computing Technology, CAS, Beijing, China and University of Chinese Academy of Sciences, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0001-8082-4218","affiliations":[{"raw_affiliation_string":"SKLP, Institute of Computing Technology, CAS, Beijing, China and University of Chinese Academy of Sciences, Beijing, China","institution_ids":["https://openalex.org/I4210090176","https://openalex.org/I4210165038"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023380073","display_name":"Xiaowei Li","orcid":"https://orcid.org/0000-0002-0874-814X"},"institutions":[{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]},{"id":"https://openalex.org/I4210165038","display_name":"University of Chinese Academy of Sciences","ror":"https://ror.org/05qbk4x57","country_code":"CN","type":"education","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaowei Li","raw_affiliation_strings":["SKLP, Institute of Computing Technology, CAS, Beijing, China and University of Chinese Academy of Sciences, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-0874-814X","affiliations":[{"raw_affiliation_string":"SKLP, Institute of Computing Technology, CAS, Beijing, China and University of Chinese Academy of Sciences, Beijing, China","institution_ids":["https://openalex.org/I4210090176","https://openalex.org/I4210165038"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100346965","display_name":"Ying Wang","orcid":"https://orcid.org/0000-0001-5172-4736"},"institutions":[{"id":"https://openalex.org/I4210090176","display_name":"Institute of Computing Technology","ror":"https://ror.org/0090r4d87","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210090176"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ying Wang","raw_affiliation_strings":["SKLP, Institute of Computing Technology, CAS, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0001-5172-4736","affiliations":[{"raw_affiliation_string":"SKLP, Institute of Computing Technology, CAS, Beijing, China","institution_ids":["https://openalex.org/I4210090176"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5050695684","display_name":"Onur Mutlu","orcid":"https://orcid.org/0000-0002-0075-2312"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Onur Mutlu","raw_affiliation_strings":["ETH Z\u00fcrich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-0075-2312","affiliations":[{"raw_affiliation_string":"ETH Z\u00fcrich, Zurich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]}],"institutions":[],"countries_distinct_count":4,"institutions_distinct_count":6,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":24,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"766","last_page":"782"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9983000159263611,"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":0.9983000159263611,"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":0.9921000003814697,"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/T10320","display_name":"Neural Networks and Applications","score":0.9914000034332275,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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":0.8748887181282043},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.7068749666213989},{"id":"https://openalex.org/keywords/parallelism","display_name":"Parallelism (grammar)","score":0.6655859351158142},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.6473942399024963},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.4523315727710724},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.3278074264526367},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.11791369318962097}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8748887181282043},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.7068749666213989},{"id":"https://openalex.org/C2781172179","wikidata":"https://www.wikidata.org/wiki/Q853109","display_name":"Parallelism (grammar)","level":2,"score":0.6655859351158142},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.6473942399024963},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.4523315727710724},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.3278074264526367},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.11791369318962097}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3676641.3716009","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3676641.3716009","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3676641.3716009","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3676641.3716009","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G5319519996","display_name":null,"funder_award_id":"62090024, 62222411","funder_id":"https://openalex.org/F4320323817","funder_display_name":"Universitas Brawijaya"}],"funders":[{"id":"https://openalex.org/F4320323817","display_name":"Universitas Brawijaya","ror":"https://ror.org/01wk3d929"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":32,"referenced_works":["https://openalex.org/W755906292","https://openalex.org/W2003838150","https://openalex.org/W2034861439","https://openalex.org/W2155263300","https://openalex.org/W2765234579","https://openalex.org/W2766489088","https://openalex.org/W3025585506","https://openalex.org/W3093933627","https://openalex.org/W3111721177","https://openalex.org/W3159727696","https://openalex.org/W4214734582","https://openalex.org/W4221001402","https://openalex.org/W4280496502","https://openalex.org/W4297097348","https://openalex.org/W4301607055","https://openalex.org/W4308083827","https://openalex.org/W4312933868","https://openalex.org/W4360831786","https://openalex.org/W4366341968","https://openalex.org/W4381894570","https://openalex.org/W4389576338","https://openalex.org/W4393407018","https://openalex.org/W4393407040","https://openalex.org/W4393407316","https://openalex.org/W4399402229","https://openalex.org/W4399677840","https://openalex.org/W4402672007","https://openalex.org/W4403334538","https://openalex.org/W4407692569","https://openalex.org/W6600171454","https://openalex.org/W6600291067","https://openalex.org/W6604985974"],"related_works":["https://openalex.org/W2045183646","https://openalex.org/W2162409446","https://openalex.org/W2109463584","https://openalex.org/W2161474341","https://openalex.org/W2504075107","https://openalex.org/W4302615923","https://openalex.org/W3203142394","https://openalex.org/W2351061015","https://openalex.org/W1595672120","https://openalex.org/W4230999561"],"abstract_inverted_index":{"Large":[0],"language":[1,9],"models":[2],"(LLMs)":[3],"are":[4],"widely":[5],"used":[6],"for":[7],"natural":[8],"understanding":[10],"and":[11,46,55,63,78,120,124],"text":[12],"generation.":[13],"An":[14],"LLM":[15,23,38,50,88],"model":[16],"relies":[17],"on":[18,33],"a":[19,69,95,135],"time-consuming":[20],"step":[21],"called":[22],"decoding":[24,39,51,89],"to":[25,68,103,106,117,134],"generate":[26],"output":[27],"tokens.":[28],"Several":[29],"prior":[30,59],"works":[31,60],"focus":[32],"improving":[34],"the":[35],"performance":[36],"of":[37,73,87,128,138],"using":[40],"parallelism":[41],"techniques,":[42],"such":[43],"as":[44],"batching":[45],"speculative":[47],"decoding.":[48],"State-of-the-art":[49],"has":[52],"both":[53,74],"compute-bound":[54],"memory-bound":[56,142],"kernels.":[57,143],"Some":[58],"statically":[61],"identify":[62],"map":[64],"these":[65],"different":[66],"kernels":[67,90],"heterogeneous":[70],"architecture":[71],"consisting":[72],"processing-in-memory":[75],"(PIM)":[76],"units":[77,119,131],"computation-centric":[79,121],"accelerators":[80,122],"(e.g.,":[81,91],"GPUs).":[82],"We":[83],"observe":[84],"that":[85],"characteristics":[86],"whether":[92],"or":[93],"not":[94],"kernel":[96,115],"is":[97],"memory-bound)":[98],"can":[99],"change":[100],"dynamically":[101],"due":[102,133],"parameter":[104],"changes":[105],"meet":[107],"user":[108],"and/or":[109],"system":[110],"demands,":[111],"making":[112],"(1)":[113],"static":[114],"mapping":[116],"PIM":[118,130],"suboptimal,":[123],"(2)":[125],"one-size-fits-all":[126],"approach":[127],"designing":[129],"inefficient":[132],"large":[136],"degree":[137],"heterogeneity":[139],"even":[140],"in":[141]},"counts_by_year":[{"year":2026,"cited_by_count":15},{"year":2025,"cited_by_count":9}],"updated_date":"2026-07-22T07:51:19.307946","created_date":"2025-10-10T00:00:00"}
