{"id":"https://openalex.org/W4408354081","doi":"https://doi.org/10.1109/icassp49660.2025.10890716","title":"QoKV: Comprehending and Surpassing the Hurdles of KV Cache Quantization","display_name":"QoKV: Comprehending and Surpassing the Hurdles of KV Cache Quantization","publication_year":2025,"publication_date":"2025-03-12","ids":{"openalex":"https://openalex.org/W4408354081","doi":"https://doi.org/10.1109/icassp49660.2025.10890716"},"language":"en","primary_location":{"id":"doi:10.1109/icassp49660.2025.10890716","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10890716","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5033106421","display_name":"Jinguang Wang","orcid":"https://orcid.org/0000-0003-2430-6701"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]},{"id":"https://openalex.org/I4392021250","display_name":"State Key Laboratory of Networking and Switching Technology","ror":"https://ror.org/00qtv5q45","country_code":null,"type":"facility","lineage":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jinguang Wang","raw_affiliation_strings":["Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China","institution_ids":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064859462","display_name":"Yuexi Yin","orcid":null},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]},{"id":"https://openalex.org/I4392021250","display_name":"State Key Laboratory of Networking and Switching Technology","ror":"https://ror.org/00qtv5q45","country_code":null,"type":"facility","lineage":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuexi Yin","raw_affiliation_strings":["Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China","institution_ids":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008194128","display_name":"Haifeng Sun","orcid":"https://orcid.org/0000-0003-3072-7422"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]},{"id":"https://openalex.org/I4392021250","display_name":"State Key Laboratory of Networking and Switching Technology","ror":"https://ror.org/00qtv5q45","country_code":null,"type":"facility","lineage":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haifeng Sun","raw_affiliation_strings":["Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China","institution_ids":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083657836","display_name":"Tingting Yang","orcid":"https://orcid.org/0000-0002-7406-2170"},"institutions":[{"id":"https://openalex.org/I4210136793","display_name":"Peng Cheng Laboratory","ror":"https://ror.org/03qdqbt06","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210136793"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Tingting Yang","raw_affiliation_strings":["PengCheng Laboratory,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"PengCheng Laboratory,Shenzhen,China","institution_ids":["https://openalex.org/I4210136793"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5104141685","display_name":"Qi Qi","orcid":"https://orcid.org/0009-0002-4302-9973"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]},{"id":"https://openalex.org/I4392021250","display_name":"State Key Laboratory of Networking and Switching Technology","ror":"https://ror.org/00qtv5q45","country_code":null,"type":"facility","lineage":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qi Qi","raw_affiliation_strings":["Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China","institution_ids":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076969590","display_name":"Zirui Zhuang","orcid":"https://orcid.org/0000-0003-3345-1732"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]},{"id":"https://openalex.org/I4392021250","display_name":"State Key Laboratory of Networking and Switching Technology","ror":"https://ror.org/00qtv5q45","country_code":null,"type":"facility","lineage":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zirui Zhuang","raw_affiliation_strings":["Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China","institution_ids":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100432460","display_name":"Jingyu Wang","orcid":"https://orcid.org/0000-0002-2182-2228"},"institutions":[{"id":"https://openalex.org/I139759216","display_name":"Beijing University of Posts and Telecommunications","ror":"https://ror.org/04w9fbh59","country_code":"CN","type":"education","lineage":["https://openalex.org/I139759216"]},{"id":"https://openalex.org/I4392021250","display_name":"State Key Laboratory of Networking and Switching Technology","ror":"https://ror.org/00qtv5q45","country_code":null,"type":"facility","lineage":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jingyu Wang","raw_affiliation_strings":["Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Beijing University of Posts and Telecommunications,State Key Laboratory of Networking and Switching Technology,Beijing,China","institution_ids":["https://openalex.org/I139759216","https://openalex.org/I4392021250"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11269","display_name":"Algorithms and Data Compression","score":0.9855999946594238,"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/T11269","display_name":"Algorithms and Data Compression","score":0.9855999946594238,"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/T10901","display_name":"Advanced Data Compression Techniques","score":0.9746999740600586,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.9728000164031982,"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/cache","display_name":"Cache","score":0.7964690923690796},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6920738220214844},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.6285330057144165},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.24877724051475525},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.07634106278419495}],"concepts":[{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.7964690923690796},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6920738220214844},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.6285330057144165},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.24877724051475525},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.07634106278419495}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp49660.2025.10890716","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp49660.2025.10890716","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320311687","display_name":"Ministry of Education","ror":"https://ror.org/03m01yf64"},{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320335787","display_name":"Fundamental Research Funds for the Central Universities","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":20,"referenced_works":["https://openalex.org/W2093647425","https://openalex.org/W2890894339","https://openalex.org/W2923014074","https://openalex.org/W2963015836","https://openalex.org/W2998617917","https://openalex.org/W3194676777","https://openalex.org/W4389524393","https://openalex.org/W4406650295","https://openalex.org/W6727099177","https://openalex.org/W6764401283","https://openalex.org/W6769627184","https://openalex.org/W6771626834","https://openalex.org/W6788175385","https://openalex.org/W6790328413","https://openalex.org/W6811340617","https://openalex.org/W6838589991","https://openalex.org/W6846659131","https://openalex.org/W6847478871","https://openalex.org/W6851828392","https://openalex.org/W6852408377"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"have":[4],"demonstrated":[5],"outstanding":[6],"performance":[7,135],"in":[8,79],"various":[9,123],"tasks.":[10],"However,":[11],"the":[12,16,38,51,65,102,140],"memory":[13],"footprint":[14],"of":[15,37,45,76,139],"key-value":[17],"(KV)":[18],"cache":[19,40,81,141],"generated":[20],"during":[21],"model":[22,29],"inference":[23],"poses":[24],"significant":[25],"challenges":[26],"for":[27,101],"efficient":[28,95],"deployment.":[30],"This":[31],"paper":[32],"presents":[33],"a":[34,55,73,83,111],"detailed":[35],"analysis":[36],"KV":[39,80,103],"and":[41,67,70,116],"identifies":[42],"potential":[43],"sources":[44],"quantization":[46,97,109],"errors.":[47],"We":[48],"find":[49],"that":[50,71,126],"attention":[52,63],"scores":[53],"exhibit":[54],"strong":[56],"power-law":[57],"distribution":[58],"while":[59,136],"showing":[60],"an":[61,94],"additional":[62],"to":[64],"initial":[66],"recent":[68],"tokens,":[69],"only":[72],"small":[74],"subset":[75],"outlier":[77],"channels":[78],"span":[82],"wide":[84],"numerical":[85,117],"range.":[86],"Based":[87],"on":[88],"these":[89],"findings,":[90],"we":[91],"propose":[92],"QoKV,":[93],"post-training":[96],"(PTQ)":[98],"method":[99,129],"designed":[100],"cache,":[104],"which":[105],"facilitates":[106],"accurate":[107],"low-bit":[108],"through":[110],"novel":[112],"group":[113],"evaluation":[114],"strategy":[115],"scaling.":[118],"Extensive":[119],"experimental":[120],"results":[121],"across":[122],"tasks":[124],"demonstrate":[125],"our":[127],"proposed":[128],"outperforms":[130],"existing":[131],"methods,":[132],"achieving":[133],"near-floating-point":[134],"saving":[137],"73.75%":[138],"footprint.":[142]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}