{"id":"https://openalex.org/W7225422778","doi":"https://doi.org/10.5281/zenodo.24004705","title":"Predictive Latency Optimization for Generative AI: Jointly Optimizing Precomputation Time and User Waiting Latency","display_name":"Predictive Latency Optimization for Generative AI: Jointly Optimizing Precomputation Time and User Waiting Latency","publication_year":2026,"publication_date":"2026-11-06","ids":{"openalex":"https://openalex.org/W7225422778","doi":"https://doi.org/10.5281/zenodo.24004705"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.24004705","doi":"https://doi.org/10.5281/zenodo.24004705","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24004705","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.24004705","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100638045","display_name":"Jincheng Zhang","orcid":"https://orcid.org/0000-0001-7332-6704","observed_orcids":["https://orcid.org/0000-0001-7332-6704","https://orcid.org/0009-0005-1860-0009"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Zhang, Jincheng","raw_affiliation_strings":[],"raw_orcid":"https://orcid.org/0009-0005-1860-0009","affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5100638045"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.44510000944137573,"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"}},"topics":[{"id":"https://openalex.org/T10101","display_name":"Cloud Computing and Resource Management","score":0.44510000944137573,"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"}},{"id":"https://openalex.org/T14098","display_name":"Energy Efficiency in Computing","score":0.07540000230073929,"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/T12127","display_name":"Software System Performance and Reliability","score":0.07339999824762344,"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/precomputation","display_name":"precomputation","score":0.8029999732971191},{"id":"https://openalex.org/keywords/inference-latency","display_name":"inference latency","score":0.8610000014305115},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.843999981880188},{"id":"https://openalex.org/keywords/deadline-aware-scheduling","display_name":"deadline-aware scheduling","score":0.8029999732971191},{"id":"https://openalex.org/keywords/multi-objective-optimization","display_name":"multi-objective optimization","score":0.824999988079071},{"id":"https://openalex.org/keywords/service-level-objectives","display_name":"service level objectives","score":0.6029999852180481}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8561000227928162},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.6765000224113464},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6718999743461609},{"id":"https://openalex.org/C127964446","wikidata":"https://www.wikidata.org/wiki/Q1092142","display_name":"Computational resource","level":3,"score":0.5619999766349792},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.5031999945640564},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.476500004529953},{"id":"https://openalex.org/C159379195","wikidata":"https://www.wikidata.org/wiki/Q7239568","display_name":"Precomputation","level":3,"score":0.4498000144958496},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4381999969482422},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4278999865055084},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3853999972343445},{"id":"https://openalex.org/C66024118","wikidata":"https://www.wikidata.org/wiki/Q1122506","display_name":"Computational model","level":2,"score":0.36800000071525574},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.3495999872684479},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.3488999903202057},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.31360000371932983},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.31130000948905945},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.3043000102043152},{"id":"https://openalex.org/C45804977","wikidata":"https://www.wikidata.org/wiki/Q7239673","display_name":"Predictive modelling","level":2,"score":0.27549999952316284},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.2700999975204468},{"id":"https://openalex.org/C158600405","wikidata":"https://www.wikidata.org/wiki/Q5054566","display_name":"Causal inference","level":2,"score":0.25540000200271606},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.25279998779296875},{"id":"https://openalex.org/C917703","wikidata":"https://www.wikidata.org/wiki/Q7239668","display_name":"Predictive inference","level":5,"score":0.25270000100135803}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.24004705","doi":"https://doi.org/10.5281/zenodo.24004705","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24004705","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.5281/zenodo.24004706","doi":"https://doi.org/10.5281/zenodo.24004706","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24004706","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.5281/zenodo.24004705","doi":"https://doi.org/10.5281/zenodo.24004705","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24004705","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7225267314","https://openalex.org/W7225376051","https://openalex.org/W7225396636","https://openalex.org/W7225408802","https://openalex.org/W7225423500","https://openalex.org/W7225375373","https://openalex.org/W7225425116","https://openalex.org/W7225378531","https://openalex.org/W7225392952","https://openalex.org/W7225400169"],"abstract_inverted_index":{"The":[0,182,234,294],"increasing":[1,111],"deployment":[2],"of":[3,115],"generative":[4,356],"artificial":[5],"intelligence":[6],"(AI)":[7],"systems":[8],"has":[9],"created":[10],"substantial":[11],"challenges":[12],"in":[13,355],"managing":[14],"inference":[15,44,100,139,143],"latency,":[16],"particularly":[17],"when":[18],"large":[19],"language":[20],"models":[21,163],"(LLMs)":[22],"operate":[23],"under":[24,84],"fluctuating":[25],"workloads,":[26],"heterogeneous":[27],"computational":[28,55,74,135,159,191,238,241,271],"requirements,":[29],"and":[30,50,76,107,137,152,178,197,204,221,231,247,269,290,319,336],"strict":[31],"user-facing":[32,138],"service-level":[33,232],"objectives.":[34],"Predictive":[35,121],"computing":[36,354],"offers":[37],"a":[38,71,128,347],"potential":[39],"mechanism":[40],"for":[41,102,124,351],"reducing":[42],"future":[43,158,172],"workload":[45],"by":[46],"anticipating":[47],"user":[48,218,337],"demand":[49,317],"preparing":[51],"responses":[52],"or":[53,83],"reusable":[54],"states":[56],"before":[57,316],"corresponding":[58],"requests":[59,101,150],"arrive.":[60],"However,":[61],"advance":[62,134],"computation":[63,148,324],"introduces":[64,120],"an":[65],"additional":[66],"temporal":[67,165,343],"optimization":[68,211,282],"problem.":[69],"Preparing":[70],"response":[72,180,199],"requires":[73],"time,":[75,261],"performing":[77],"preparation":[78,95,136,170,206,223,259,262,302,312],"too":[79,81],"early,":[80],"late,":[82],"inappropriate":[85],"resource":[86,194,265],"conditions":[87],"may":[88],"reduce":[89],"its":[90,307],"practical":[91],"value.":[92],"Moreover,":[93],"predictive":[94,153,185,222,280,292,301,353],"can":[96,313],"compete":[97],"with":[98,283],"active":[99],"GPU":[103,227],"resources,":[104],"memory":[105],"bandwidth,":[106],"execution":[108,202],"capacity,":[109,228],"potentially":[110],"the":[112,164,256,321],"waiting":[113,219,249,338],"time":[114,239,242,250,339],"existing":[116],"users.":[117],"This":[118],"paper":[119],"Latency":[122],"Optimization":[123],"Generative":[125],"AI":[126,357],"(PLO-GAI),":[127],"conceptual":[129],"framework":[130,183,235],"that":[131,300],"jointly":[132,341],"optimizes":[133],"latency.":[140],"Unlike":[141],"conventional":[142,286],"optimization,":[144],"which":[145,155],"primarily":[146,156],"reduces":[147,304],"after":[149],"arrive,":[151],"computing,":[154],"anticipates":[157],"demand,":[160],"PLO-GAI":[161,345],"explicitly":[162],"relationship":[166],"between":[167,237,258],"prediction":[168,267],"initiation,":[169],"completion,":[171],"request":[173],"arrival,":[174],"residual":[175],"online":[176,323],"computation,":[177],"final":[179],"delivery.":[181],"integrates":[184],"request-arrival":[186],"estimation,":[187],"preparation-duration":[188],"prediction,":[189],"deadline-aware":[190],"scheduling,":[192],"latency-aware":[193,352],"allocation,":[195],"partial":[196],"complete":[198],"preparation,":[200],"queue-aware":[201],"control,":[203],"adaptive":[205],"cancellation.":[207],"A":[208],"unified":[209],"multi-objective":[210],"formulation":[212],"is":[213,276,325],"developed":[214],"to":[215,244,278],"minimize":[216],"expected":[217],"latency":[220,281],"expenditure":[224],"while":[225],"satisfying":[226],"memory,":[229],"response-quality,":[230],"constraints.":[233],"distinguishes":[236],"saved,":[240],"transferred":[243],"earlier":[245],"periods,":[246],"user-visible":[248],"actually":[251],"avoided.":[252],"It":[253],"further":[254],"investigates":[255],"trade-offs":[257],"lead":[260],"completion":[263],"probability,":[264],"contention,":[266],"uncertainty,":[268],"request-specific":[270],"requirements.":[272],"An":[273],"experimental":[274],"methodology":[275],"proposed":[277,295],"compare":[279],"reactive":[284],"inference,":[285],"caching,":[287],"latency-unaware":[288],"precomputation,":[289],"resource-aware":[291],"preparation.":[293],"research":[296,349],"does":[297],"not":[298],"assume":[299],"necessarily":[303],"end-to-end":[305],"latency;":[306],"effectiveness":[308],"depends":[309],"on":[310],"whether":[311,320],"be":[314],"completed":[315],"materializes":[318],"remaining":[322],"sufficiently":[326],"reduced":[327],"without":[328],"degrading":[329],"concurrent":[330],"service.":[331],"By":[332],"treating":[333],"precomputation":[334],"duration":[335],"as":[340],"controllable":[342],"variables,":[344],"establishes":[346],"distinct":[348],"direction":[350],"infrastructure.":[358]},"counts_by_year":[],"updated_date":"2026-11-08T12:12:28.789561","created_date":"2026-11-07T00:00:00"}