{"id":"https://openalex.org/W7224788681","doi":"https://doi.org/10.5281/zenodo.23788173","title":"\u9762\u5bf9\u5927\u6a21\u578b\u9ad8\u6548\u80fd\u5b58\u7b97\u5206\u79bb\u67b6\u6784","display_name":"\u9762\u5bf9\u5927\u6a21\u578b\u9ad8\u6548\u80fd\u5b58\u7b97\u5206\u79bb\u67b6\u6784","publication_year":2026,"publication_date":"2026-10-30","ids":{"openalex":"https://openalex.org/W7224788681","doi":"https://doi.org/10.5281/zenodo.23788173"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.23788173","doi":"https://doi.org/10.5281/zenodo.23788173","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23788173","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.23788173","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5158001796","display_name":"\u7b11\u7136 \u6f58","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"\u6f58, \u7b11\u7136","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5158001796"],"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.43849998712539673,"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"}},"topics":[{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.43849998712539673,"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/T10028","display_name":"Topic Modeling","score":0.16769999265670776,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.09969999641180038,"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/large-language-models","display_name":"large language models","score":0.8090000152587891},{"id":"https://openalex.org/keywords/key-value-cache","display_name":"key-value cache","score":0.6269999742507935},{"id":"https://openalex.org/keywords/edge-computing","display_name":"edge computing","score":0.4480000138282776},{"id":"https://openalex.org/keywords/energy-efficiency","display_name":"energy efficiency","score":0.7170000076293945},{"id":"https://openalex.org/keywords/content-safety","display_name":"content safety","score":0.5680000185966492},{"id":"https://openalex.org/keywords/data-privacy","display_name":"data privacy","score":0.6230000257492065}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8075000047683716},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5543000102043152},{"id":"https://openalex.org/C193702766","wikidata":"https://www.wikidata.org/wiki/Q1414548","display_name":"Concurrency","level":2,"score":0.5090000033378601},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.4869999885559082},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.43849998712539673},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.40939998626708984},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.376800000667572},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.3709000051021576},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.3513000011444092},{"id":"https://openalex.org/C7366592","wikidata":"https://www.wikidata.org/wiki/Q1255620","display_name":"Dram","level":2,"score":0.3504999876022339},{"id":"https://openalex.org/C46743427","wikidata":"https://www.wikidata.org/wiki/Q1341685","display_name":"Inference engine","level":3,"score":0.3353999853134155},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.3327000141143799},{"id":"https://openalex.org/C128896722","wikidata":"https://www.wikidata.org/wiki/Q1501387","display_name":"Transitive closure","level":2,"score":0.30160000920295715},{"id":"https://openalex.org/C187691185","wikidata":"https://www.wikidata.org/wiki/Q2020720","display_name":"Grid","level":2,"score":0.2985999882221222},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.2978000044822693},{"id":"https://openalex.org/C150495011","wikidata":"https://www.wikidata.org/wiki/Q128392","display_name":"Concurrent computing","level":2,"score":0.2973000109195709},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2849999964237213},{"id":"https://openalex.org/C159023740","wikidata":"https://www.wikidata.org/wiki/Q623276","display_name":"Deadlock","level":2,"score":0.28220000863075256},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.2660999894142151},{"id":"https://openalex.org/C191399111","wikidata":"https://www.wikidata.org/wiki/Q64861","display_name":"Transitive relation","level":2,"score":0.2558000087738037},{"id":"https://openalex.org/C2779602883","wikidata":"https://www.wikidata.org/wiki/Q15544750","display_name":"Memory architecture","level":2,"score":0.25369998812675476}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.23788173","doi":"https://doi.org/10.5281/zenodo.23788173","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23788173","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.23788174","doi":"https://doi.org/10.5281/zenodo.23788174","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23788174","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.23788173","doi":"https://doi.org/10.5281/zenodo.23788173","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23788173","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/W7147482908","https://openalex.org/W4416159316","https://openalex.org/W7119643653","https://openalex.org/W7165457725","https://openalex.org/W7165480801","https://openalex.org/W7117408316","https://openalex.org/W7133538689","https://openalex.org/W7214935083","https://openalex.org/W4406178148","https://openalex.org/W7165467156"],"abstract_inverted_index":{"Abstract":[0],"The":[1,25,129,257,273],"evolution":[2],"of":[3,20,35,69,76,78,125,141,150,214,237,253],"contemporary":[4],"large":[5],"language":[6],"models":[7],"(LLMs)":[8],"remains":[9],"constrained":[10],"by":[11,16,31],"an":[12,166],"architectural":[13],"bottleneck":[14],"dominated":[15],"the":[17,32,43,66,134,207,211,220,246],"brute-force":[18],"scaling":[19],"compute":[21,118],"and":[22,39,102,127,137,160,181,185,202,223,234,270,281],"GPU":[23,233],"memory.":[24],"mainstream":[26],"Transformer":[27],"architecture":[28,131,209,258],"is":[29,276,286],"limited":[30],"O(N^2)":[33],"complexity":[34,203,213],"its":[36],"attention":[37],"mechanism":[38],"therefore":[40],"must":[41],"keep":[42],"full":[44],"key\u2013value":[45],"cache":[46],"(KV":[47],"cache)":[48],"resident":[49],"in":[50,98],"extremely":[51],"expensive":[52],"high-bandwidth":[53],"memory":[54,126,169],"(HBM).":[55],"Technology":[56],"firms":[57],"have":[58],"attempted":[59],"to":[60,94,103,145,217,240,288],"compensate":[61],"for":[62],"software-level":[63],"inefficiencies":[64],"through":[65],"physical":[67],"expansion":[68],"cloud-based":[70],"intelligent":[71],"computing":[72],"clusters":[73],"comprising":[74],"tens":[75],"thousands":[77],"GPUs.":[79],"This":[80,107],"approach":[81],"not":[82],"only":[83],"incurs":[84],"substantial":[85],"energy":[86,242],"consumption\u2014the":[87],"so-called":[88],"\u201cVRAM":[89],"wall\u201d":[90],"problem\u2014but":[91],"also":[92,259],"leads":[93],"severe":[95],"concurrency":[96],"failures":[97],"centralized":[99],"cloud":[100],"inference":[101],"increased":[104],"market":[105],"concentration.":[106],"paper":[108],"proposes":[109],"a":[110,147,182,194,225,231,279],"system-level":[111],"paradigm":[112],"shift:":[113],"rather":[114],"than":[115],"pursuing":[116],"further":[117],"scaling,":[119],"it":[120],"advocates":[121],"deep":[122,155],"topological":[123],"optimization":[124],"storage.":[128],"proposed":[130,208],"draws":[132],"on":[133],"\u201cchunk":[135],"loading\u201d":[136],"\u201cspatial":[138],"teleportation\u201d":[139],"logic":[140],"open-world":[142],"game":[143],"engines":[144],"achieve":[146,241],"complete":[148],"decoupling":[149],"logical":[151,156],"retrieval":[152,216],"scheduling":[153],"from":[154],"inference.":[157],"We":[158],"propose":[159],"formally":[161],"define":[162],"five":[163],"core":[164],"mechanisms:":[165],"independent":[167],"navigation":[168],"layer,":[170],"semantic":[171],"closure":[172],"anchors,":[173],"Temporal":[174],"Teleport":[175],"Mapping":[176],"(TP-Mapping),":[177],"spawn-point":[178],"persistent":[179],"snapshots,":[180],"native":[183,304],"compliance":[184],"safety":[186],"circuit-breaking":[187],"layer.":[188],"On":[189],"this":[190],"basis,":[191],"we":[192],"construct":[193],"fat-client/thin-server":[195],"device\u2013cloud":[196],"collaborative":[197],"concurrency-offloading":[198],"model.":[199],"Theoretical":[200],"derivation":[201],"analysis":[204],"indicate":[205],"that":[206,252],"reduces":[210],"time":[212],"long-text":[215],"O(1)":[218],"at":[219],"addressing":[221],"stage":[222],"enables":[224],"single":[226],"personal":[227],"computer":[228],"equipped":[229],"with":[230],"consumer-grade":[232],"32":[235],"GB":[236],"commodity":[238],"DRAM":[239],"efficiency":[243],"that,":[244],"under":[245],"stated":[247],"analytical":[248,282],"assumptions,":[249],"may":[250],"exceed":[251],"industrial-grade":[254],"multi-GPU":[255],"clusters.":[256],"eliminates":[260],"high-concurrency":[261],"downtime":[262],"risk":[263],"while":[264],"providing":[265],"strong":[266],"data":[267],"privacy":[268],"protection":[269],"content-safety":[271],"compliance.":[272],"present":[274],"work":[275],"intended":[277],"as":[278],"systems-level":[280],"proposal;":[283],"empirical":[284],"validation":[285],"left":[287],"future":[289],"work.":[290],"Keywords:":[291],"storage\u2013compute":[292],"disaggregation;":[293],"VRAM":[294],"wall;":[295],"chunk":[296],"loading;":[297],"temporal":[298],"addressing;":[299],"state":[300],"snapshot;":[301],"edge":[302],"computing;":[303],"compliance;":[305],"low-cost":[306],"AI":[307]},"counts_by_year":[],"updated_date":"2026-11-01T11:40:01.658920","created_date":"2026-10-31T00:00:00"}