{"id":"https://openalex.org/W7141092135","doi":"https://doi.org/10.1109/icce67443.2026.11449769","title":"On-Device-First Hybrid LLM Inference on AI-PCs: Closing the Enterprise GenAI Divide","display_name":"On-Device-First Hybrid LLM Inference on AI-PCs: Closing the Enterprise GenAI Divide","publication_year":2026,"publication_date":"2026-02-03","ids":{"openalex":"https://openalex.org/W7141092135","doi":"https://doi.org/10.1109/icce67443.2026.11449769"},"language":null,"primary_location":{"id":"doi:10.1109/icce67443.2026.11449769","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icce67443.2026.11449769","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE International Conference on Consumer Electronics (ICCE)","raw_type":"proceedings-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/A5130776915","display_name":"Sultana Begum","orcid":null},"institutions":[{"id":"https://openalex.org/I4210158342","display_name":"Intel (United Kingdom)","ror":"https://ror.org/058cxws58","country_code":"GB","type":"company","lineage":["https://openalex.org/I1343180700","https://openalex.org/I4210158342"]}],"countries":["GB"],"is_corresponding":true,"raw_author_name":"Sultana Begum","raw_affiliation_strings":["Intel Corporation"],"affiliations":[{"raw_affiliation_string":"Intel Corporation","institution_ids":["https://openalex.org/I4210158342"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130797941","display_name":"Kris Fleming","orcid":null},"institutions":[{"id":"https://openalex.org/I4210158342","display_name":"Intel (United Kingdom)","ror":"https://ror.org/058cxws58","country_code":"GB","type":"company","lineage":["https://openalex.org/I1343180700","https://openalex.org/I4210158342"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Kris Fleming","raw_affiliation_strings":["Intel Corporation"],"affiliations":[{"raw_affiliation_string":"Intel Corporation","institution_ids":["https://openalex.org/I4210158342"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5057263481","display_name":"Todd Lewellen","orcid":null},"institutions":[{"id":"https://openalex.org/I4210158342","display_name":"Intel (United Kingdom)","ror":"https://ror.org/058cxws58","country_code":"GB","type":"company","lineage":["https://openalex.org/I1343180700","https://openalex.org/I4210158342"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Todd Lewellen","raw_affiliation_strings":["Intel Corporation"],"affiliations":[{"raw_affiliation_string":"Intel Corporation","institution_ids":["https://openalex.org/I4210158342"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5130776915"],"corresponding_institution_ids":["https://openalex.org/I4210158342"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.91989048,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"6"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.2840999960899353,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.2840999960899353,"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/T10883","display_name":"Ethics and Social Impacts of AI","score":0.052299998700618744,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T13643","display_name":"Artificial Intelligence in Law","score":0.03359999880194664,"subfield":{"id":"https://openalex.org/subfields/3320","display_name":"Political Science and International Relations"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/closing","display_name":"Closing (real estate)","score":0.48410001397132874},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.383899986743927},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.3610999882221222},{"id":"https://openalex.org/keywords/class","display_name":"Class (philosophy)","score":0.2973000109195709},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.2766000032424927}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5939000248908997},{"id":"https://openalex.org/C2778775528","wikidata":"https://www.wikidata.org/wiki/Q5135432","display_name":"Closing (real estate)","level":2,"score":0.48410001397132874},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40790000557899475},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.383899986743927},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3610999882221222},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.2973000109195709},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.2766000032424927},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.2476000040769577},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2264000028371811},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.22509999573230743}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icce67443.2026.11449769","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icce67443.2026.11449769","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE International Conference on Consumer Electronics (ICCE)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":10,"referenced_works":["https://openalex.org/W4387321091","https://openalex.org/W4399534541","https://openalex.org/W4402667093","https://openalex.org/W4409313050","https://openalex.org/W4409365821","https://openalex.org/W4413299510","https://openalex.org/W4413308156","https://openalex.org/W4415428439","https://openalex.org/W4416035376","https://openalex.org/W4416873642"],"related_works":[],"abstract_inverted_index":{"AI-PCs":[0],"and":[1,9,32,48,62,84,119,124,127,132,140,149,157,162],"modern":[2],"PCs":[3],"equipped":[4],"with":[5,112],"capable":[6],"CPUs,":[7],"GPUs,":[8],"NPUs":[10],"now":[11],"make":[12,153],"on-device":[13,103],"inference":[14],"for":[15,67,91],"small":[16],"language":[17],"models":[18,40,88,114],"(SLMs)":[19],"practical":[20,53,184],"across":[21],"many":[22],"enterprise":[23],"workloads.":[24],"This":[25,59],"enables":[26],"assistants":[27],"that":[28,109,152,167,177],"are":[29],"low-latency,":[30],"privacy-preserving,":[31],"resilient":[33],"to":[34,54,86,130,186,196],"network":[35],"conditions,":[36],"while":[37],"frontier":[38],"cloud":[39,87],"still":[41],"offer":[42,182],"higher":[43],"accuracy,":[44],"richer":[45],"tool":[46],"use,":[47],"stronger":[49],"multimodality":[50],"than":[51],"is":[52,89],"run":[55],"on":[56,76,144],"every":[57],"endpoint.":[58],"paper":[60],"surveys":[61],"organizes":[63],"the":[64,77,188],"design":[65,106,160],"space":[66],"an":[68,74],"on-device-first":[69],"hybrid":[70,180],"architecture":[71],"in":[72,100,191],"which":[73,192],"SLM":[75],"AI-PC":[78],"handles":[79],"most":[80],"requests":[81],"by":[82],"default,":[83],"escalation":[85],"reserved":[90],"difficult,":[92],"tool-heavy,":[93],"or":[94],"long-context":[95],"tasks.":[96],"We":[97,175],"synthesize":[98],"advances":[99],"compact":[101],"local":[102],"optimized":[104],"model":[105],"including":[107],"architectures":[108],"mix":[110],"self-attention":[111],"state-space":[113],"(SSMs),":[115],"parameter-efficient":[116],"adaptation,":[117],"multimodal":[118],"structure-aware":[120],"retrieval-augmented":[121],"generation":[122],"(RAG),":[123],"long-horizon":[125],"memory,":[126],"connect":[128],"these":[129],"hardware":[131],"serving":[133],"optimizations":[134],"such":[135,178],"as":[136],"quantization,":[137],"mixed":[138],"precision,":[139],"KV-cache":[141],"management.":[142],"Building":[143],"this,":[145],"we":[146],"propose":[147],"routing":[148],"reporting":[150],"patterns":[151],"latency,":[154],"energy,":[155],"cost,":[156],"privacy":[158],"explicit":[159],"parameters,":[161],"outline":[163],"a":[164,183],"total-cost-of-ownership":[165],"view":[166],"clarifies":[168],"when":[169],"AI-PC\u2013centric":[170],"deployments":[171],"outperform":[172],"all-cloud":[173],"strategies.":[174],"argue":[176],"edge-first":[179],"designs":[181],"path":[185],"narrowing":[187],"\"GenAI":[189],"Divide,\"":[190],"impressive":[193],"pilots":[194],"fail":[195],"translate":[197],"into":[198],"durable,":[199],"production-grade":[200],"business":[201],"impact.":[202]},"counts_by_year":[],"updated_date":"2026-03-29T06:01:01.467347","created_date":"2026-03-28T00:00:00"}