{"id":"https://openalex.org/W4410153353","doi":"https://doi.org/10.1109/lca.2025.3567844","title":"OASIS: Outlier-Aware KV Cache Clustering for Scaling LLM Inference in CXL Memory Systems","display_name":"OASIS: Outlier-Aware KV Cache Clustering for Scaling LLM Inference in CXL Memory Systems","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4410153353","doi":"https://doi.org/10.1109/lca.2025.3567844"},"language":"en","primary_location":{"id":"doi:10.1109/lca.2025.3567844","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lca.2025.3567844","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/A5102025104","display_name":"Minseok Seo","orcid":"https://orcid.org/0000-0003-3648-5575"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Minseok Seo","raw_affiliation_strings":["Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0003-3648-5575","affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I139264467"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Jungi Hyun","orcid":"https://orcid.org/0009-0001-7716-8886"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jungi Hyun","raw_affiliation_strings":["Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea"],"raw_orcid":"https://orcid.org/0009-0001-7716-8886","affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I139264467"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Seongho Jeong","orcid":"https://orcid.org/0009-0002-4454-3800"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Seongho Jeong","raw_affiliation_strings":["Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea"],"raw_orcid":"https://orcid.org/0009-0002-4454-3800","affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I139264467"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014900695","display_name":"Xuan Truong Nguyen","orcid":"https://orcid.org/0000-0002-7527-6971"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Xuan Truong Nguyen","raw_affiliation_strings":["Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0002-7527-6971","affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I139264467"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102861073","display_name":"Hyuk-Jae Lee","orcid":"https://orcid.org/0000-0001-8895-9117"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]},{"id":"https://openalex.org/I57664883","display_name":"Ajou University","ror":"https://ror.org/03tzb2h73","country_code":"KR","type":"education","lineage":["https://openalex.org/I57664883"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyuk-Jae Lee","raw_affiliation_strings":["Department of ECE, Ajou University, Suwon-si, South Korea","Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0001-8895-9117","affiliations":[{"raw_affiliation_string":"Department of ECE, Ajou University, Suwon-si, South Korea","institution_ids":["https://openalex.org/I57664883"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE) and Inter-university Semiconductor Research Center, Seoul National University, Seoul, Republic of Korea","institution_ids":["https://openalex.org/I139264467"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5079409207","display_name":"Hyokeun Lee","orcid":"https://orcid.org/0000-0002-0824-6238"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]},{"id":"https://openalex.org/I57664883","display_name":"Ajou University","ror":"https://ror.org/03tzb2h73","country_code":"KR","type":"education","lineage":["https://openalex.org/I57664883"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyokeun Lee","raw_affiliation_strings":["Department of ECE, Ajou University, Suwon-si, South Korea","Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","Department of ECE, Ajou University, Gyeonggido, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0002-0824-6238","affiliations":[{"raw_affiliation_string":"Department of ECE, Ajou University, Suwon-si, South Korea","institution_ids":["https://openalex.org/I57664883"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering (ECE), Inter-university Semiconductor Research Center, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]},{"raw_affiliation_string":"Department of ECE, Ajou University, Gyeonggido, Republic of Korea","institution_ids":["https://openalex.org/I57664883"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5102025104"],"corresponding_institution_ids":["https://openalex.org/I139264467"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.04179155,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"24","issue":"1","first_page":"165","last_page":"168"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.9186000227928162,"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/T10320","display_name":"Neural Networks and Applications","score":0.9186000227928162,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.9021999835968018,"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":0.8492051362991333},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.6695891618728638},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6159982681274414},{"id":"https://openalex.org/keywords/cache","display_name":"Cache","score":0.5826995372772217},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.4668505787849426},{"id":"https://openalex.org/keywords/outlier","display_name":"Outlier","score":0.43220147490501404},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.40928930044174194},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.23737871646881104}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8492051362991333},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.6695891618728638},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6159982681274414},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.5826995372772217},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.4668505787849426},{"id":"https://openalex.org/C79337645","wikidata":"https://www.wikidata.org/wiki/Q779824","display_name":"Outlier","level":2,"score":0.43220147490501404},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.40928930044174194},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.23737871646881104},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/lca.2025.3567844","is_oa":false,"landing_page_url":"https://doi.org/10.1109/lca.2025.3567844","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":[],"awards":[{"id":"https://openalex.org/G1037224261","display_name":null,"funder_award_id":"RS-2025-02305531","funder_id":"https://openalex.org/F4320322741","funder_display_name":"Ministerstvo Pr\u016fmyslu a Obchodu"}],"funders":[{"id":"https://openalex.org/F4320322741","display_name":"Ministerstvo Pr\u016fmyslu a Obchodu","ror":"https://ror.org/03j4eb467"},{"id":"https://openalex.org/F4320328359","display_name":"Ministry of Science and ICT, South Korea","ror":"https://ror.org/01wpjm123"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":14,"referenced_works":["https://openalex.org/W2807925339","https://openalex.org/W4390280328","https://openalex.org/W4393407255","https://openalex.org/W4395073431","https://openalex.org/W4401211590","https://openalex.org/W4402670433","https://openalex.org/W4402671766","https://openalex.org/W6850927664","https://openalex.org/W6857690716","https://openalex.org/W6861788325","https://openalex.org/W6862025885","https://openalex.org/W6864878861","https://openalex.org/W6873917596","https://openalex.org/W6875634212"],"related_works":["https://openalex.org/W3006513224","https://openalex.org/W2046456988","https://openalex.org/W2357409937","https://openalex.org/W2510582230","https://openalex.org/W2978674666","https://openalex.org/W2074430941","https://openalex.org/W141820298","https://openalex.org/W2113096305","https://openalex.org/W1977636359","https://openalex.org/W2772305933"],"abstract_inverted_index":{"The":[0],"key-value":[1],"(KV)":[2],"cache":[3,68],"in":[4,42,74,89],"large":[5],"language":[6],"models":[7],"(LLM)":[8],"now":[9],"necessitates":[10],"a":[11,32,43],"substantial":[12],"amount":[13],"of":[14,58,107,128],"memory":[15,30,37,41,76],"capacity":[16],"as":[17],"its":[18],"size":[19],"proportionally":[20],"grows":[21],"with":[22,125],"the":[23,55,83,118],"context's":[24],"size.":[25],"Recently,":[26],"Compute-Express":[27],"Link":[28],"(CXL)":[29],"becomes":[31],"promising":[33],"method":[34,79],"to":[35,54,97,117],"secure":[36],"capacity.":[38],"However,":[39],"CXL":[40,59,75],"GPU-based":[44],"LLM":[45,72],"inference":[46,73],"platform":[47],"entails":[48],"performance":[49,93],"and":[50,94],"scalability":[51],"challenges":[52],"due":[53],"limited":[56],"bandwidth":[57],"memory.":[60],"This":[61],"paper":[62],"proposes":[63],"OASIS,":[64],"an":[65],"outlier-aware":[66],"KV":[67,130],"clustering":[69,86,104,121],"for":[70],"scaling":[71],"systems.":[77],"Our":[78,109],"is":[80,87,105],"based":[81],"on":[82],"observation":[84],"that":[85],"effective":[88],"trading":[90],"off":[91],"between":[92],"accuracy":[95,124],"compared":[96,116],"previous":[98],"quantization-":[99],"or":[100],"selection-based":[101],"approaches":[102],"if":[103],"aware":[106],"outliers.":[108],"evaluation":[110],"shows":[111],"OASIS":[112],"yields":[113],"3.6\u00d7":[114],"speedup":[115],"case":[119],"without":[120],"while":[122],"preserving":[123],"just":[126],"5%":[127],"full":[129],"cache.":[131]},"counts_by_year":[],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}