{"id":"https://openalex.org/W4414371540","doi":"https://doi.org/10.1145/3768628","title":"RAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generation","display_name":"RAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generation","publication_year":2025,"publication_date":"2025-09-20","ids":{"openalex":"https://openalex.org/W4414371540","doi":"https://doi.org/10.1145/3768628"},"language":"en","primary_location":{"id":"doi:10.1145/3768628","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3768628","pdf_url":null,"source":{"id":"https://openalex.org/S193109227","display_name":"ACM Transactions on Computer Systems","issn_l":"0734-2071","issn":["0734-2071","1557-7333"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Computer Systems","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/A5102021404","display_name":"Chao Jin","orcid":"https://orcid.org/0009-0006-1355-4995"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chao Jin","raw_affiliation_strings":["School of Computer Science, Peking University","School of Computer Science, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0006-1355-4995","affiliations":[{"raw_affiliation_string":"School of Computer Science, Peking University","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Computer Science, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100603108","display_name":"Zili Zhang","orcid":"https://orcid.org/0000-0003-4209-9451"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zili Zhang","raw_affiliation_strings":["School of Computer Science, Peking University","School of Computer Science, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-4209-9451","affiliations":[{"raw_affiliation_string":"School of Computer Science, Peking University","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Computer Science, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102652594","display_name":"Xuanlin Jiang","orcid":null},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuanlin Jiang","raw_affiliation_strings":["School of Electronics Engineering and Computer Science, Peking University","School of Electronics Engineering and Computer Science, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0006-2265-1877","affiliations":[{"raw_affiliation_string":"School of Electronics Engineering and Computer Science, Peking University","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Electronics Engineering and Computer Science, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111178462","display_name":"Fangyue Liu","orcid":"https://orcid.org/0009-0001-9791-046X"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fangyue Liu","raw_affiliation_strings":["School of Computer Science, Peking University","School of Computer Science, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0001-9791-046X","affiliations":[{"raw_affiliation_string":"School of Computer Science, Peking University","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Computer Science, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Shufan Liu","orcid":"https://orcid.org/0009-0007-1978-4983"},"institutions":[{"id":"https://openalex.org/I231086526","display_name":"University of Siedlce","ror":"https://ror.org/01wkb9987","country_code":"PL","type":"education","lineage":["https://openalex.org/I231086526"]}],"countries":["PL"],"is_corresponding":false,"raw_author_name":"Shufan Liu","raw_affiliation_strings":["ByteDance Seed","ByteDance Seed, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0007-1978-4983","affiliations":[{"raw_affiliation_string":"ByteDance Seed","institution_ids":["https://openalex.org/I231086526"]},{"raw_affiliation_string":"ByteDance Seed, Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052249316","display_name":"Xuanzhe Liu","orcid":"https://orcid.org/0000-0002-7908-8484"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuanzhe Liu","raw_affiliation_strings":["School of Computer Science, Peking University","School of Computer Science, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-7908-8484","affiliations":[{"raw_affiliation_string":"School of Computer Science, Peking University","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Computer Science, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101882004","display_name":"Xin Jin","orcid":"https://orcid.org/0000-0001-8741-5847"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xin Jin","raw_affiliation_strings":["School of Computer Science, Peking University","School of Computer Science, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0001-8741-5847","affiliations":[{"raw_affiliation_string":"School of Computer Science, Peking University","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Computer Science, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":30.9824,"has_fulltext":false,"cited_by_count":33,"citation_normalized_percentile":{"value":0.99755491,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":99,"max":100},"biblio":{"volume":"44","issue":"1","first_page":"1","last_page":"27"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11478","display_name":"Caching and Content Delivery","score":0.9932000041007996,"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"}},"topics":[{"id":"https://openalex.org/T11478","display_name":"Caching and Content Delivery","score":0.9932000041007996,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9921000003814697,"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/T10203","display_name":"Recommender Systems and Techniques","score":0.9811000227928162,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/exploit","display_name":"Exploit","score":0.7379999756813049},{"id":"https://openalex.org/keywords/cache","display_name":"Cache","score":0.640999972820282},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.5807999968528748},{"id":"https://openalex.org/keywords/host","display_name":"Host (biology)","score":0.4788999855518341},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.4510999917984009},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.3467999994754791},{"id":"https://openalex.org/keywords/data-retrieval","display_name":"Data retrieval","score":0.3215999901294708}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.9235000014305115},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.7379999756813049},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.640999972820282},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.5807999968528748},{"id":"https://openalex.org/C126831891","wikidata":"https://www.wikidata.org/wiki/Q221673","display_name":"Host (biology)","level":2,"score":0.4788999855518341},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.4510999917984009},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.3467999994754791},{"id":"https://openalex.org/C551230270","wikidata":"https://www.wikidata.org/wiki/Q4368942","display_name":"Data retrieval","level":2,"score":0.3215999901294708},{"id":"https://openalex.org/C189783530","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"CPU cache","level":3,"score":0.3131999969482422},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.3046000003814697},{"id":"https://openalex.org/C2776187449","wikidata":"https://www.wikidata.org/wiki/Q1513879","display_name":"Natural language generation","level":3,"score":0.29899999499320984},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.2809999883174896},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2809999883174896},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.27300000190734863},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.27160000801086426},{"id":"https://openalex.org/C141917322","wikidata":"https://www.wikidata.org/wiki/Q1025017","display_name":"Cache coherence","level":5,"score":0.2590000033378601},{"id":"https://openalex.org/C161156560","wikidata":"https://www.wikidata.org/wiki/Q1638872","display_name":"Document retrieval","level":2,"score":0.25859999656677246},{"id":"https://openalex.org/C2778348673","wikidata":"https://www.wikidata.org/wiki/Q739302","display_name":"Production (economics)","level":2,"score":0.2551000118255615},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3768628","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3768628","pdf_url":null,"source":{"id":"https://openalex.org/S193109227","display_name":"ACM Transactions on Computer Systems","issn_l":"0734-2071","issn":["0734-2071","1557-7333"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Computer Systems","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":26,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W2340222647","https://openalex.org/W2912924812","https://openalex.org/W2954698171","https://openalex.org/W2963339397","https://openalex.org/W2963469388","https://openalex.org/W3029693508","https://openalex.org/W4225108562","https://openalex.org/W4317898419","https://openalex.org/W4378718572","https://openalex.org/W4382319938","https://openalex.org/W4385571271","https://openalex.org/W4387321091","https://openalex.org/W4388584589","https://openalex.org/W4388778348","https://openalex.org/W4391876619","https://openalex.org/W4393147129","https://openalex.org/W4401175629","https://openalex.org/W4401176373","https://openalex.org/W4401399383","https://openalex.org/W4402671542","https://openalex.org/W4405418476","https://openalex.org/W4408848702","https://openalex.org/W4411471721","https://openalex.org/W4411485933","https://openalex.org/W4411486516"],"related_works":[],"abstract_inverted_index":{"Retrieval-Augmented":[0],"Generation":[1],"(RAG)":[2],"has":[3],"demonstrated":[4],"substantial":[5],"advancements":[6],"in":[7,39,97],"various":[8],"natural":[9],"language":[10,19],"processing":[11],"tasks":[12],"by":[13,161,169],"integrating":[14],"the":[15,27,53,84,91,102,119,155,167],"strengths":[16],"of":[17,94],"large":[18],"models":[20],"(LLMs)":[21],"and":[22,34,49,68,89,104,138,140,146,165],"external":[23],"knowledge":[24,96,99],"databases.":[25],"However,":[26],"retrieval":[28,54,66,85,124],"step":[29,55],"introduces":[30],"long":[31,40],"sequence":[32],"generation":[33,110,127],"extra":[35],"data":[36],"dependency,":[37],"resulting":[38],"end-to-end":[41,130],"latency.":[42,131],"Our":[43],"analysis":[44],"benchmarks":[45],"current":[46],"RAG":[47],"systems":[48],"reveals":[50],"that,":[51],"while":[52],"poses":[56],"performance":[57],"challenges,":[58],"it":[59,142],"also":[60],"offers":[61],"optimization":[62],"opportunities":[63],"through":[64],"its":[65],"pattern":[67,86],"streaming":[69,120],"search":[70,121],"behavior.":[71],"We":[72,132],"propose":[73],"RAGCache,":[74],"a":[75,98],"latency-optimized":[76],"serving":[77],"system":[78],"tailored":[79],"for":[80],"RAG.":[81],"RAGCache":[82,112,134,153],"leverages":[83],"to":[87,117,128,157,163,171,174],"organize":[88],"cache":[90],"intermediate":[92],"states":[93],"retrieved":[95],"tree":[100],"across":[101],"GPU":[103],"host":[105],"memory":[106],"hierarchy,":[107],"reducing":[108],"LLM":[109,126],"time.":[111],"employs":[113],"dynamic":[114],"speculative":[115],"pipelining":[116],"exploit":[118],"behavior,":[122],"overlapping":[123],"with":[125,177],"minimize":[129],"implement":[133],"based":[135],"on":[136,143],"vLLM":[137,175],"Faiss,":[139],"evaluate":[141],"both":[144],"open-source":[145],"production":[147],"datasets.":[148],"Experimental":[149],"results":[150],"demonstrate":[151],"that":[152],"reduces":[154],"time":[156],"first":[158],"token":[159],"(TTFT)":[160],"up":[162,170],"4\u00d7":[164],"improves":[166],"throughput":[168],"2.1\u00d7":[172],"compared":[173],"integrated":[176],"Faiss.":[178]},"counts_by_year":[{"year":2026,"cited_by_count":17},{"year":2025,"cited_by_count":16}],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2025-10-10T00:00:00"}
