{"id":"https://openalex.org/W4410492668","doi":"https://doi.org/10.1109/cicc63670.2025.10983849","title":"A 28nm 3.14 TFLOP/W BF16 LLM Fine-Tuning Processor with Asymmetric Quantization Computing for AI PC","display_name":"A 28nm 3.14 TFLOP/W BF16 LLM Fine-Tuning Processor with Asymmetric Quantization Computing for AI PC","publication_year":2025,"publication_date":"2025-04-13","ids":{"openalex":"https://openalex.org/W4410492668","doi":"https://doi.org/10.1109/cicc63670.2025.10983849"},"language":"en","primary_location":{"id":"doi:10.1109/cicc63670.2025.10983849","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cicc63670.2025.10983849","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Custom Integrated Circuits Conference (CICC)","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/A5039208284","display_name":"Xinyuan Lin","orcid":"https://orcid.org/0009-0002-6881-5348"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Xinyuan Lin","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Leran Huang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210114105","display_name":"Tsinghua\u2013Berkeley Shenzhen Institute","ror":"https://ror.org/02hhwwz98","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210114105","https://openalex.org/I95457486","https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Leran Huang","raw_affiliation_strings":["Tsinghua Shenzhen International Graduate School,Shenzhen,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua Shenzhen International Graduate School,Shenzhen,China","institution_ids":["https://openalex.org/I4210114105"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5018178773","display_name":"Chenhan Wei","orcid":"https://orcid.org/0000-0003-2683-9835"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chenhan Wei","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5042337656","display_name":"Wenbin Jia","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wenbin Jia","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101280807","display_name":"Hedi Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hedi Wang","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023029393","display_name":"Wenxun Wang","orcid":"https://orcid.org/0009-0007-1999-6441"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wenxun Wang","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5088075152","display_name":"Hongyang Jia","orcid":"https://orcid.org/0000-0001-8692-1860"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongyang Jia","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103867707","display_name":"Huazhong Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huazhong Yang","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5045721867","display_name":"Yongpan Liu","orcid":"https://orcid.org/0000-0002-4892-2309"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yongpan Liu","raw_affiliation_strings":["Tsinghua University,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5039208284"],"corresponding_institution_ids":["https://openalex.org/I99065089"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.04813507,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"3"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.954200029373169,"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/T10320","display_name":"Neural Networks and Applications","score":0.954200029373169,"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/T12611","display_name":"Neural Networks and Reservoir Computing","score":0.9455000162124634,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6756525635719299},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.5668200254440308},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5266779661178589},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.32242220640182495},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.20711541175842285}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6756525635719299},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.5668200254440308},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5266779661178589},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.32242220640182495},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.20711541175842285}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/cicc63670.2025.10983849","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cicc63670.2025.10983849","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE Custom Integrated Circuits Conference (CICC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.7599999904632568,"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy"}],"awards":[{"id":"https://openalex.org/G2021969373","display_name":null,"funder_award_id":"92267203","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G5685581577","display_name":null,"funder_award_id":"2021ZD0114402","funder_id":"https://openalex.org/F4320329860","funder_display_name":"National Science and Technology Major Project"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320329860","display_name":"National Science and Technology Major Project","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":9,"referenced_works":["https://openalex.org/W3135859967","https://openalex.org/W3172967059","https://openalex.org/W4360606939","https://openalex.org/W4376134052","https://openalex.org/W4378509449","https://openalex.org/W4401879629","https://openalex.org/W6846164622","https://openalex.org/W6848451824","https://openalex.org/W6856931071"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"The":[0],"powerful":[1],"capabilities":[2],"of":[3,46,76,157],"large":[4],"language":[5],"models":[6],"(LLMs)":[7],"enable":[8],"them":[9],"to":[10],"function":[11],"as":[12,36],"personal":[13],"digital":[14],"assistants.":[15],"To":[16],"ensure":[17],"user":[18],"privacy,":[19],"personalized":[20],"fine-tuning":[21,92,107],"can":[22],"be":[23],"conducted":[24],"locally":[25],"on":[26,94],"memoryconstrained":[27],"AI":[28],"PCs":[29],"using":[30],"Parameter-Efficient":[31],"Fine-Tuning":[32],"(PEFT)":[33],"algorithms,":[34],"such":[35],"QLoRA[1]":[37],"and":[38,59,130,142,151],"QA-LORA[2].":[39],"Figure":[40],"1":[41],"illustrates":[42],"the":[43,70,77,104,155],"computation":[44],"flow":[45],"QLoRA:":[47],"BF16":[48,60],"input":[49],"activations":[50],"undergo":[51],"matrix":[52],"multiplication":[53],"with":[54,168],"4-bit":[55,144],"quantized":[56,145],"pre-trained":[57],"weights":[58],"adapter":[61],"weights.":[62],"In":[63],"this":[64],"flow,":[65],"asymmetric":[66,90,111,114],"quantization":[67,115],"MACs":[68],"represent":[69],"primary":[71],"bottleneck,":[72],"consuming":[73],"approximately":[74],"97%":[75],"computational":[78],"load.":[79],"However,":[80],"current":[81],"neural":[82],"processing":[83],"units":[84],"(NPUs)":[85],"offer":[86],"limited":[87,169],"support":[88,124,137],"for":[89,106,110,138],"computation:":[91],"Llama2-13B":[93],"an":[95],"RTX":[96],"3090":[97],"takes":[98],"over":[99],"25":[100],"hours.":[101],"This":[102],"highlights":[103],"need":[105],"processors":[108],"optimized":[109],"quantization.":[112],"Yet,":[113],"presents":[116],"hardware":[117],"design":[118],"challenges:":[119],"1)":[120],"Existing":[121],"NPUs":[122,134],"primarily":[123],"symmetric":[125],"formats,":[126],"introducing":[127],"conversion":[128],"overhead":[129,167],"inefficiencies;":[131],"2)":[132],"Current":[133],"lack":[135],"efficient":[136],"low-precision":[139],"data":[140],"transposition;":[141],"3)":[143],"QLoRA":[146],"encounters":[147],"high":[148],"external":[149],"access":[150],"storage":[152],"demands,":[153],"while":[154],"use":[156],"2:4":[158],"sparsity":[159],"in":[160],"low-bit":[161],"LLM":[162],"finetuning[3]":[163],"incurs":[164],"substantial":[165],"bitmask":[166],"benefits.":[170]},"counts_by_year":[],"updated_date":"2026-04-21T08:09:41.155169","created_date":"2025-10-10T00:00:00"}