{"id":"https://openalex.org/W4415934811","doi":"https://doi.org/10.1109/tc.2025.3628193","title":"Bit-Serial Acceleration of LLM Inference With Mixture-of-Datatype Quantization","display_name":"Bit-Serial Acceleration of LLM Inference With Mixture-of-Datatype Quantization","publication_year":2025,"publication_date":"2025-11-05","ids":{"openalex":"https://openalex.org/W4415934811","doi":"https://doi.org/10.1109/tc.2025.3628193"},"language":null,"primary_location":{"id":"doi:10.1109/tc.2025.3628193","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2025.3628193","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","0016-9340","1557-9956","2326-3814"],"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 Transactions on Computers","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/A5101930621","display_name":"Yuzong Chen","orcid":"https://orcid.org/0000-0001-6387-327X"},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yuzong Chen","raw_affiliation_strings":["Cornell University, New York, NY, USA"],"raw_orcid":"https://orcid.org/0000-0001-6387-327X","affiliations":[{"raw_affiliation_string":"Cornell University, New York, NY, USA","institution_ids":["https://openalex.org/I205783295"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005153980","display_name":"Chi-Chih Chang","orcid":"https://orcid.org/0000-0002-3154-1745"},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chi-Chih Chang","raw_affiliation_strings":["Cornell University, New York, NY, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cornell University, New York, NY, USA","institution_ids":["https://openalex.org/I205783295"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5094114735","display_name":"Xilai Dai","orcid":"https://orcid.org/0009-0009-9802-9402"},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xilai Dai","raw_affiliation_strings":["Cornell University, New York, NY, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Cornell University, New York, NY, USA","institution_ids":["https://openalex.org/I205783295"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069242854","display_name":"Ahmed F. AbouElhamayed","orcid":"https://orcid.org/0000-0001-6381-2936"},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ahmed F. AbouElhamayed","raw_affiliation_strings":["Cornell University, New York, NY, USA"],"raw_orcid":"https://orcid.org/0000-0001-6381-2936","affiliations":[{"raw_affiliation_string":"Cornell University, New York, NY, USA","institution_ids":["https://openalex.org/I205783295"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5092829415","display_name":"Marta Andronic","orcid":null},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Marta Andronic","raw_affiliation_strings":["Imperial College London, London, U.K","Imperial College London, London, UK"],"raw_orcid":"https://orcid.org/0009-0004-6153-0491","affiliations":[{"raw_affiliation_string":"Imperial College London, London, U.K","institution_ids":["https://openalex.org/I47508984"]},{"raw_affiliation_string":"Imperial College London, London, UK","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029829952","display_name":"George A. Constantinides","orcid":"https://orcid.org/0000-0002-0201-310X"},"institutions":[{"id":"https://openalex.org/I47508984","display_name":"Imperial College London","ror":"https://ror.org/041kmwe10","country_code":"GB","type":"education","lineage":["https://openalex.org/I47508984"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"George A. Constantinides","raw_affiliation_strings":["Imperial College London, London, U.K","Imperial College London, London, UK"],"raw_orcid":"https://orcid.org/0000-0002-0201-310X","affiliations":[{"raw_affiliation_string":"Imperial College London, London, U.K","institution_ids":["https://openalex.org/I47508984"]},{"raw_affiliation_string":"Imperial College London, London, UK","institution_ids":["https://openalex.org/I47508984"]}]},{"author_position":"last","author":{"id":null,"display_name":"Mohamed S. Abdelfattah","orcid":"https://orcid.org/0000-0002-4568-8932"},"institutions":[{"id":"https://openalex.org/I205783295","display_name":"Cornell University","ror":"https://ror.org/05bnh6r87","country_code":"US","type":"education","lineage":["https://openalex.org/I205783295"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mohamed S. Abdelfattah","raw_affiliation_strings":["Cornell University, New York, NY, USA"],"raw_orcid":"https://orcid.org/0000-0002-4568-8932","affiliations":[{"raw_affiliation_string":"Cornell University, New York, NY, USA","institution_ids":["https://openalex.org/I205783295"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":{"value":2645,"currency":"USD","value_usd":2645},"apc_paid":null,"fwci":1.1309,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.8554607,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":"75","issue":"2","first_page":"567","last_page":"581"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.46050000190734863,"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"}},"topics":[{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.46050000190734863,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.15369999408721924,"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"}},{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.05719999969005585,"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/quantization","display_name":"Quantization (signal processing)","score":0.7407000064849854},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6284999847412109},{"id":"https://openalex.org/keywords/discriminative-model","display_name":"Discriminative model","score":0.5978999733924866},{"id":"https://openalex.org/keywords/generative-model","display_name":"Generative model","score":0.5317999720573425},{"id":"https://openalex.org/keywords/hardware-acceleration","display_name":"Hardware acceleration","score":0.4894999861717224},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.4381999969482422},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4147999882698059},{"id":"https://openalex.org/keywords/memory-footprint","display_name":"Memory footprint","score":0.3946000039577484}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7663000226020813},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.7407000064849854},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6284999847412109},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.5978999733924866},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.5317999720573425},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.4894999861717224},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.44190001487731934},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.4381999969482422},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4147999882698059},{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.3946000039577484},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.3856000006198883},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.3824999928474426},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3732999861240387},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3693999946117401},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.3465999960899353},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.34209999442100525},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.3249000012874603},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3133000135421753},{"id":"https://openalex.org/C58489278","wikidata":"https://www.wikidata.org/wiki/Q1172284","display_name":"Data set","level":2,"score":0.31150001287460327},{"id":"https://openalex.org/C138958017","wikidata":"https://www.wikidata.org/wiki/Q190087","display_name":"Data type","level":2,"score":0.30979999899864197},{"id":"https://openalex.org/C199833920","wikidata":"https://www.wikidata.org/wiki/Q612536","display_name":"Vector quantization","level":2,"score":0.3009999990463257},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2757999897003174},{"id":"https://openalex.org/C2777299769","wikidata":"https://www.wikidata.org/wiki/Q3707858","display_name":"Type (biology)","level":2,"score":0.271699994802475},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.26510000228881836},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.2531999945640564}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tc.2025.3628193","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2025.3628193","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","0016-9340","1557-9956","2326-3814"],"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 Transactions on Computers","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3496335909","display_name":"Centre for Spatial Computational Learning","funder_award_id":"EP/S030069/1","funder_id":"https://openalex.org/F4320334627","funder_display_name":"Engineering and Physical Sciences Research Council"}],"funders":[{"id":"https://openalex.org/F4320334627","display_name":"Engineering and Physical Sciences Research Council","ror":"https://ror.org/0439y7842"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":19,"referenced_works":["https://openalex.org/W2725159389","https://openalex.org/W2946609015","https://openalex.org/W2996908057","https://openalex.org/W2998617917","https://openalex.org/W3006586535","https://openalex.org/W3205209748","https://openalex.org/W3213241618","https://openalex.org/W4247198796","https://openalex.org/W4308083739","https://openalex.org/W4366341968","https://openalex.org/W4380874652","https://openalex.org/W4393145114","https://openalex.org/W4393406875","https://openalex.org/W4393407021","https://openalex.org/W4404954664","https://openalex.org/W4409248600","https://openalex.org/W4409248709","https://openalex.org/W4415796210","https://openalex.org/W4415799217"],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"have":[4],"achieved":[5],"significant":[6],"breakthroughs":[7],"on":[8,138,153],"machine":[9],"learning":[10],"tasks.":[11,158],"Yet":[12],"the":[13,39,66,90,95,160,171],"substantial":[14],"memory":[15],"footprint":[16],"of":[17,60,69,142,177],"LLMs":[18,143],"significantly":[19,147],"hinders":[20],"their":[21],"wide":[22],"deployment.":[23,37],"In":[24],"this":[25],"paper,":[26],"we":[27,120],"propose":[28],"BitMoD,":[29],"an":[30,165],"algorithm-hardware":[31],"co-design":[32],"solution":[33],"for":[34],"efficient":[35,166],"LLM":[36,78,150,173],"On":[38,89],"algorithm":[40],"side,":[41,92],"BitMoD":[42,73,93,146,169],"introduces":[43],"\u201cfine-grained":[44],"data":[45,52,71,106],"type":[46,53],"adaptation\u201d,":[47],"which":[48],"uses":[49],"a":[50,56,110,139],"different":[51],"to":[54,76,82,99,125],"quantize":[55,77],"group":[57],"(e.g.,":[58],"128)":[59],"weights":[61,79],"and":[62,80,105,116,131,156,181],"key-value-cache":[63],"(KV-cache).":[64],"Through":[65],"careful":[67],"design":[68,121],"these":[70],"types,":[72,107],"is":[74],"able":[75],"KV-cache":[81,129],"sub-4-bit":[83],"precision":[84],"while":[85],"maintaining":[86],"high":[87],"accuracy.":[88],"hardware":[91,117,123,179],"employs":[94],"bit-serial":[96],"computing":[97],"paradigm":[98],"easily":[100],"support":[101],"multiple":[102],"numerical":[103],"precisions":[104],"thus":[108],"providing":[109],"flexible":[111],"trade-off":[112],"between":[113],"model":[114,162],"accuracy":[115],"efficiency.":[118,183],"Furthermore,":[119],"low-cost":[122],"components":[124],"effectively":[126],"handle":[127],"online":[128],"quantization":[130,151],"per-group":[132],"partial":[133],"sum":[134],"dequantization.":[135],"Our":[136],"evaluation":[137],"diverse":[140],"set":[141],"demonstrates":[144],"that":[145],"outperforms":[148],"state-of-the-art":[149,172],"methods":[152],"both":[154,178],"discriminative":[155],"generative":[157],"Combining":[159],"superior":[161],"performance":[163,180],"with":[164],"accelerator":[167,174],"design,":[168],"surpasses":[170],"in":[175],"terms":[176],"energy":[182]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-08-21T09:56:20.448147","created_date":"2025-11-05T00:00:00"}