{"id":"https://openalex.org/W7214733153","doi":"https://doi.org/10.48550/arxiv.2609.32999","title":"Tessera: Demand-Driven KV Cache Management for Retrieval-Augmented LLM Serving","display_name":"Tessera: Demand-Driven KV Cache Management for Retrieval-Augmented LLM Serving","publication_year":2026,"publication_date":"2026-09-26","ids":{"openalex":"https://openalex.org/W7214733153","doi":"https://doi.org/10.48550/arxiv.2609.32999"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2609.32999","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.32999","pdf_url":"https://arxiv.org/pdf/2609.32999","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2609.32999","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5153518455","display_name":"Fei Fang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Fang, Fei","raw_affiliation_strings":["University of California , Santa Cruz"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California , Santa Cruz","institution_ids":["https://openalex.org/I185103710"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153478776","display_name":"Chung-Hsiang Lo","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Lo, Chung-Hsiang","raw_affiliation_strings":["University of California , Santa Cruz"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California , Santa Cruz","institution_ids":["https://openalex.org/I185103710"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5153414716","display_name":"Yi Liu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Liu, Yi","raw_affiliation_strings":["University of California , Santa Cruz"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California , Santa Cruz","institution_ids":["https://openalex.org/I185103710"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5022189369","display_name":"Yifan Hua","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hua, Yifan","raw_affiliation_strings":["University of California , Santa Cruz"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California , Santa Cruz","institution_ids":["https://openalex.org/I185103710"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100737734","display_name":"Chen Qian","orcid":"https://orcid.org/0000-0002-6882-9590","observed_orcids":["https://orcid.org/0000-0002-6882-9590","https://orcid.org/0000-0001-6498-4262"]},"institutions":[{"id":"https://openalex.org/I185103710","display_name":"University of California, Santa Cruz","ror":"https://ror.org/03s65by71","country_code":"US","type":"education","lineage":["https://openalex.org/I185103710"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qian, Chen","raw_affiliation_strings":["University of California , Santa Cruz"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California , Santa Cruz","institution_ids":["https://openalex.org/I185103710"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I185103710"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.23319999873638153,"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/T10028","display_name":"Topic Modeling","score":0.23319999873638153,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.13040000200271606,"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/T10317","display_name":"Advanced Database Systems and Queries","score":0.12489999830722809,"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/retrieval-augmented-generation","display_name":"retrieval-augmented generation","score":0.9559999704360962},{"id":"https://openalex.org/keywords/large-language-model-inference","display_name":"large language model inference","score":0.75},{"id":"https://openalex.org/keywords/agent-memory","display_name":"agent memory","score":0.703000009059906},{"id":"https://openalex.org/keywords/request-routing","display_name":"request routing","score":0.871999979019165}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8062000274658203},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.632099986076355},{"id":"https://openalex.org/C141603448","wikidata":"https://www.wikidata.org/wiki/Q134830","display_name":"Prefix","level":2,"score":0.6219000220298767},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.6190000176429749},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.5860000252723694},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.49639999866485596},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.4083000123500824},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.3440000116825104},{"id":"https://openalex.org/C82687282","wikidata":"https://www.wikidata.org/wiki/Q66221","display_name":"Auxiliary memory","level":2,"score":0.3368000090122223},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.3292999863624573},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.3257000148296356},{"id":"https://openalex.org/C115519274","wikidata":"https://www.wikidata.org/wiki/Q267903","display_name":"EPIC","level":2,"score":0.3253999948501587},{"id":"https://openalex.org/C53833338","wikidata":"https://www.wikidata.org/wiki/Q1061424","display_name":"Context switch","level":2,"score":0.2996000051498413},{"id":"https://openalex.org/C190290938","wikidata":"https://www.wikidata.org/wiki/Q387015","display_name":"Trie","level":3,"score":0.2888000011444092},{"id":"https://openalex.org/C189783530","wikidata":"https://www.wikidata.org/wiki/Q352090","display_name":"CPU cache","level":3,"score":0.288100004196167},{"id":"https://openalex.org/C2779530757","wikidata":"https://www.wikidata.org/wiki/Q1207505","display_name":"Quality (philosophy)","level":2,"score":0.28049999475479126},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.26840001344680786},{"id":"https://openalex.org/C5119721","wikidata":"https://www.wikidata.org/wiki/Q220501","display_name":"Quality of service","level":2,"score":0.26499998569488525},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.2644999921321869},{"id":"https://openalex.org/C2778000800","wikidata":"https://www.wikidata.org/wiki/Q830043","display_name":"Handshake","level":3,"score":0.2590999901294708},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.25839999318122864},{"id":"https://openalex.org/C2778618852","wikidata":"https://www.wikidata.org/wiki/Q1128613","display_name":"Content-addressable storage","level":4,"score":0.2556999921798706},{"id":"https://openalex.org/C162319229","wikidata":"https://www.wikidata.org/wiki/Q175263","display_name":"Data structure","level":2,"score":0.25060001015663147}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2609.32999","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.32999","pdf_url":"https://arxiv.org/pdf/2609.32999","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2609.32999","doi":"https://doi.org/10.48550/arxiv.2609.32999","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2609.32999","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2609.32999","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2609.32999","pdf_url":"https://arxiv.org/pdf/2609.32999","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7214733153.pdf","grobid_xml":"https://content.openalex.org/works/W7214733153.grobid-xml"},"referenced_works_count":9,"referenced_works":["https://openalex.org/W4387636003","https://openalex.org/W4389984066","https://openalex.org/W4395686609","https://openalex.org/W4400104234","https://openalex.org/W4407831980","https://openalex.org/W4416034797","https://openalex.org/W4417136556","https://openalex.org/W4417150458","https://openalex.org/W6929286924"],"related_works":["https://openalex.org/W4411403197","https://openalex.org/W7203662383","https://openalex.org/W4413360541","https://openalex.org/W7214734835","https://openalex.org/W7220726744","https://openalex.org/W7140323352","https://openalex.org/W7090856496","https://openalex.org/W7160700801","https://openalex.org/W4416541269","https://openalex.org/W7214786013"],"abstract_inverted_index":{"RAG":[0,166],"and":[1,15,134,141,152,167,180,188],"retrieval-based":[2],"agent":[3],"memory":[4,17,58],"both":[5],"inject":[6],"retrieved":[7],"content":[8,22],"into":[9],"LLM":[10],"prompts,":[11],"as":[12],"document":[13],"chunks":[14],"recalled":[16],"records,":[18],"respectively.":[19],"The":[20],"same":[21],"can":[23],"recur":[24],"across":[25],"requests":[26],"at":[27,184,192],"different":[28,33],"prompt":[29],"positions":[30],"or":[31,91],"after":[32],"preceding":[34],"contexts,":[35],"preventing":[36],"reuse":[37,67],"through":[38],"conventional":[39],"prefix":[40,51],"caching.":[41],"Our":[42],"characterization":[43],"finds":[44],"that":[45,104],"records":[46],"recurring":[47,79],"outside":[48],"the":[49,107,115,161,195,200,204],"matching":[50,199],"account":[52],"for":[53,110],"over":[54,178],"70%":[55],"of":[56,203],"injected":[57],"tokens":[59],"in":[60,68],"agent-memory":[61,168],"workloads.":[62],"Composable":[63],"KV-reuse":[64],"methods":[65],"enable":[66],"such":[69],"cases,":[70],"but":[71],"online":[72],"serving":[73,102],"introduces":[74],"a":[75,78,100],"management":[76,140],"problem:":[77],"unit's":[80],"KV":[81,111,132],"states":[82,159],"may":[83,87,92],"not":[84],"yet":[85],"exist,":[86],"have":[88],"been":[89],"evicted,":[90],"reside":[93],"on":[94],"another":[95],"node.":[96],"We":[97],"present":[98],"Tessera,":[99],"disaggregated":[101],"system":[103],"makes":[105],"retrieval":[106,122,130],"control":[108],"plane":[109],"reuse.":[112],"By":[113],"exposing":[114],"context":[116],"units":[117],"needed":[118],"before":[119],"model":[120],"execution,":[121],"allows":[123],"Tessera":[124,170],"to":[125,137,176],"combine":[126],"current":[127],"demand":[128],"with":[129,182],"history,":[131],"residency,":[133],"generation":[135],"load":[136],"coordinate":[138],"cache":[139],"request":[142,186],"routing.":[143],"Generation":[144],"nodes":[145],"concurrently":[146],"prepare":[147],"locally":[148],"cached,":[149,151],"remotely":[150],"missing":[153],"states,":[154],"while":[155,198],"retaining":[156],"newly":[157],"computed":[158],"off":[160],"request's":[162],"critical":[163],"path.":[164],"Across":[165],"workloads,":[169],"lowers":[171],"mean":[172],"TTFT":[173,191],"by":[174],"up":[175],"3.6x":[177],"SGLang":[179],"LMCache":[181],"EPIC":[183],"matched":[185],"rates,":[187],"sustains":[189],"low":[190],"rates":[193],"where":[194],"baselines":[196],"saturate,":[197],"answer":[201],"quality":[202],"underlying":[205],"composition":[206],"policy.":[207]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-09-30T00:00:00"}
