{"id":"https://openalex.org/W4411486062","doi":"https://doi.org/10.1145/3695053.3731092","title":"LIA: A Single-GPU LLM Inference Acceleration with Cooperative AMX-Enabled CPU-GPU Computation and CXL Offloading","display_name":"LIA: A Single-GPU LLM Inference Acceleration with Cooperative AMX-Enabled CPU-GPU Computation and CXL Offloading","publication_year":2025,"publication_date":"2025-06-20","ids":{"openalex":"https://openalex.org/W4411486062","doi":"https://doi.org/10.1145/3695053.3731092"},"language":"en","primary_location":{"id":"doi:10.1145/3695053.3731092","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3695053.3731092","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3695053.3731092","source":null,"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 52nd Annual International Symposium on Computer Architecture","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3695053.3731092","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5050020235","display_name":"Hyungyo Kim","orcid":"https://orcid.org/0000-0002-1157-0127"},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hyungyo Kim","raw_affiliation_strings":["University of Illinois at Urbana-Champaign, Champaign, Illinois, USA"],"raw_orcid":"https://orcid.org/0000-0002-1157-0127","affiliations":[{"raw_affiliation_string":"University of Illinois at Urbana-Champaign, Champaign, Illinois, USA","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030337371","display_name":"Nachuan Wang","orcid":"https://orcid.org/0009-0000-1431-0067"},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nachuan Wang","raw_affiliation_strings":["University of Illinois at Urbana-Champaign, Champaign, Illinois, USA"],"raw_orcid":"https://orcid.org/0009-0000-1431-0067","affiliations":[{"raw_affiliation_string":"University of Illinois at Urbana-Champaign, Champaign, Illinois, USA","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Qirong Xia","orcid":"https://orcid.org/0009-0008-1643-0554"},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qirong Xia","raw_affiliation_strings":["University of Illinois at Urbana-Champaign, Champaign, USA"],"raw_orcid":"https://orcid.org/0009-0008-1643-0554","affiliations":[{"raw_affiliation_string":"University of Illinois at Urbana-Champaign, Champaign, USA","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033906953","display_name":"Jinghan Huang","orcid":"https://orcid.org/0009-0009-2314-2734"},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jinghan Huang","raw_affiliation_strings":["University of Illinois at Urbana-Champaign, Champaign, USA"],"raw_orcid":"https://orcid.org/0009-0009-2314-2734","affiliations":[{"raw_affiliation_string":"University of Illinois at Urbana-Champaign, Champaign, USA","institution_ids":["https://openalex.org/I157725225"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070172290","display_name":"Amir Yazdanbakhsh","orcid":"https://orcid.org/0000-0001-8199-7671"},"institutions":[{"id":"https://openalex.org/I1291425158","display_name":"Google (United States)","ror":"https://ror.org/00njsd438","country_code":"US","type":"company","lineage":["https://openalex.org/I1291425158","https://openalex.org/I4210128969"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Amir Yazdanbakhsh","raw_affiliation_strings":["Google DeepMind, Mountain View, USA"],"raw_orcid":"https://orcid.org/0000-0001-8199-7671","affiliations":[{"raw_affiliation_string":"Google DeepMind, Mountain View, USA","institution_ids":["https://openalex.org/I1291425158"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5037648751","display_name":"Nam Sung Kim","orcid":"https://orcid.org/0000-0002-0442-5634"},"institutions":[{"id":"https://openalex.org/I157725225","display_name":"University of Illinois Urbana-Champaign","ror":"https://ror.org/047426m28","country_code":"US","type":"education","lineage":["https://openalex.org/I157725225"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Nam Sung Kim","raw_affiliation_strings":["University of Illinois at Urbana-Champaign, Champaign, USA"],"raw_orcid":"https://orcid.org/0000-0002-0442-5634","affiliations":[{"raw_affiliation_string":"University of Illinois at Urbana-Champaign, Champaign, USA","institution_ids":["https://openalex.org/I157725225"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":15.3559,"has_fulltext":true,"cited_by_count":11,"citation_normalized_percentile":{"value":0.98884512,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"544","last_page":"558"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9923999905586243,"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":0.9923999905586243,"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/T11986","display_name":"Scientific Computing and Data Management","score":0.9890999794006348,"subfield":{"id":"https://openalex.org/subfields/1802","display_name":"Information Systems and Management"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T11181","display_name":"Advanced Data Storage Technologies","score":0.9800999760627747,"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/computer-science","display_name":"Computer science","score":0.8319913744926453},{"id":"https://openalex.org/keywords/acceleration","display_name":"Acceleration","score":0.6562016606330872},{"id":"https://openalex.org/keywords/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.6272756457328796},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.568195641040802},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5362403392791748},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.46880796551704407},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.32342392206192017},{"id":"https://openalex.org/keywords/computer-graphics","display_name":"Computer graphics (images)","score":0.26168859004974365},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.09595373272895813},{"id":"https://openalex.org/keywords/graphics","display_name":"Graphics","score":0.06530070304870605}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8319913744926453},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.6562016606330872},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.6272756457328796},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.568195641040802},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5362403392791748},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.46880796551704407},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.32342392206192017},{"id":"https://openalex.org/C121684516","wikidata":"https://www.wikidata.org/wiki/Q7600677","display_name":"Computer graphics (images)","level":1,"score":0.26168859004974365},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.09595373272895813},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.06530070304870605},{"id":"https://openalex.org/C74650414","wikidata":"https://www.wikidata.org/wiki/Q11397","display_name":"Classical mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3695053.3731092","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3695053.3731092","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3695053.3731092","source":null,"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 52nd Annual International Symposium on Computer Architecture","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3695053.3731092","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3695053.3731092","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3695053.3731092","source":null,"license":"other-oa","license_id":"https://openalex.org/licenses/other-oa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 52nd Annual International Symposium on Computer Architecture","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.46000000834465027,"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7"}],"awards":[{"id":"https://openalex.org/G1452086972","display_name":null,"funder_award_id":"JUMP 2.0","funder_id":"https://openalex.org/F4320306087","funder_display_name":"Semiconductor Research Corporation"},{"id":"https://openalex.org/G7298062262","display_name":null,"funder_award_id":"RS-2024-00456287","funder_id":"https://openalex.org/F4320328359","funder_display_name":"Ministry of Science and ICT, South Korea"}],"funders":[{"id":"https://openalex.org/F4320306087","display_name":"Semiconductor Research Corporation","ror":"https://ror.org/047z4n946"},{"id":"https://openalex.org/F4320328359","display_name":"Ministry of Science and ICT, South Korea","ror":"https://ror.org/01wpjm123"},{"id":"https://openalex.org/F4320332180","display_name":"Defense Advanced Research Projects Agency","ror":"https://ror.org/02caytj08"},{"id":"https://openalex.org/F4320335489","display_name":"Institute for Information and Communications Technology Promotion","ror":"https://ror.org/01g0hqq23"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4411486062.pdf","grobid_xml":"https://content.openalex.org/works/W4411486062.grobid-xml"},"referenced_works_count":12,"referenced_works":["https://openalex.org/W2888482885","https://openalex.org/W3004304303","https://openalex.org/W3098136301","https://openalex.org/W3204998121","https://openalex.org/W3205803342","https://openalex.org/W4312783897","https://openalex.org/W4321636575","https://openalex.org/W4361194512","https://openalex.org/W4382246105","https://openalex.org/W4385571011","https://openalex.org/W4394998968","https://openalex.org/W4401211704"],"related_works":["https://openalex.org/W1963859303","https://openalex.org/W2364044215","https://openalex.org/W2389600408","https://openalex.org/W240129890","https://openalex.org/W3048701459","https://openalex.org/W2149078538","https://openalex.org/W2370314112","https://openalex.org/W1912958759","https://openalex.org/W2005148983","https://openalex.org/W2792081825"],"abstract_inverted_index":{"The":[0],"limited":[1],"memory":[2,126,129,181],"capacity":[3],"of":[4,55],"single":[5,143],"GPUs":[6],"constrains":[7],"large":[8],"language":[9],"model":[10],"(LLM)":[11],"inference,":[12],"necessitating":[13],"cost-prohibitive":[14],"multi-GPU":[15],"deployments":[16],"or":[17],"frequent":[18],"performance-limiting":[19],"CPU-GPU":[20,75,100],"transfers":[21,76],"over":[22,176],"slow":[23],"PCIe.In":[24],"this":[25],"work,":[26],"we":[27,88],"first":[28],"benchmark":[29],"recent":[30,64],"Intel":[31],"CPUs":[32],"with":[33,127,141,182],"Advanced":[34],"Matrix":[35],"Extensions":[36],"(AMX),":[37],"including":[38],"4th":[39],"generation":[40,45],"(Sapphire":[41],"Rapids)":[42,47,139],"and":[43,58,77,102,113,154],"6th":[44],"(Granite":[46,138],"Xeon":[48],"Scalable":[49],"Processors,":[50],"demonstrating":[51],"matrix":[52],"multiplication":[53],"throughput":[54,79,158,174],"20":[56],"TFLOPS":[57],"40":[59],"TFLOPS,":[60],"respectivelycomparable":[61],"to":[62,72,82,108,130,149,160],"some":[63],"GPUs.These":[65],"findings":[66],"unlock":[67],"more":[68],"extensive":[69],"computation":[70,101,107],"offloading":[71,164,169],"CPUs,":[73,109],"reducing":[74],"alleviating":[78],"bottlenecks":[80],"compared":[81,159],"priorgeneration":[83],"CPUs.Building":[84],"on":[85],"these":[86],"insights,":[87],"design":[89],"LIA,":[90],"a":[91,118,142,183],"single-GPU":[92,163],"LLM":[93],"inference":[94],"acceleration":[95],"framework":[96,115],"leveraging":[97],"cooperative":[98],"AMX-enabled":[99],"CXL":[103,125,168],"offloading.LIA":[104],"systematically":[105],"offloads":[106],"optimizing":[110],"both":[111],"latency":[112,153],"throughput.The":[114],"also":[116],"introduces":[117],"memoryoffloading":[119],"policy":[120],"that":[121],"seamlessly":[122],"integrates":[123],"affordable":[124],"DDR":[128,180],"enhance":[131],"performance":[132],"in":[133,186],"throughput-driven":[134],"tasks.On":[135],"Saphhire":[136],"Rapids":[137],"systems":[140],"H100":[144],"GPU,":[145],"LIA":[146,166,177],"achieves":[147],"up":[148],"5.1":[150],"(19)":[151],"lower":[152],"3.7":[155],"(5.1)":[156],"higher":[157],"the":[161],"latest":[162],"framework.Furthermore,":[165],"deploying":[167],"yields":[170],"an":[171],"additional":[172],"1.5":[173],"improvement":[175],"using":[178],"only":[179],"1.8":[184],"increase":[185],"maximum":[187],"batch":[188],"size":[189],"(9001.6K).":[190]},"counts_by_year":[{"year":2026,"cited_by_count":9},{"year":2025,"cited_by_count":2}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}