{"id":"https://openalex.org/W4414197630","doi":"https://doi.org/10.1109/dac63849.2025.11132978","title":"BBAL: A Bidirectional Block Floating Point-Based Quantisation Accelerator for Large Language Models","display_name":"BBAL: A Bidirectional Block Floating Point-Based Quantisation Accelerator for Large Language Models","publication_year":2025,"publication_date":"2025-06-22","ids":{"openalex":"https://openalex.org/W4414197630","doi":"https://doi.org/10.1109/dac63849.2025.11132978"},"language":"en","primary_location":{"id":"doi:10.1109/dac63849.2025.11132978","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac63849.2025.11132978","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 62nd ACM/IEEE Design Automation Conference (DAC)","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/A5030201424","display_name":"Xiaomeng Han","orcid":null},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Xiaomeng Han","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102783410","display_name":"Yuan Cheng","orcid":"https://orcid.org/0000-0003-4173-293X"},"institutions":[{"id":"https://openalex.org/I4210091958","display_name":"Hanoi Open University","ror":"https://ror.org/00gjkwr85","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210091958"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Yuan Cheng","raw_affiliation_strings":["Houmo AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI","institution_ids":["https://openalex.org/I4210091958"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100378697","display_name":"Jing Wang","orcid":"https://orcid.org/0000-0003-3653-7013"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Jing Wang","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101922705","display_name":"Junyang Lu","orcid":"https://orcid.org/0000-0003-0100-5535"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Junyang Lu","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100395405","display_name":"Hui Wang","orcid":"https://orcid.org/0000-0002-6400-0881"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Hui Wang","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021980728","display_name":"Xingzheng Zhang","orcid":null},"institutions":[{"id":"https://openalex.org/I149240348","display_name":"Jilin Normal University","ror":"https://ror.org/00xtsag93","country_code":"CN","type":"education","lineage":["https://openalex.org/I149240348"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"X.x. Zhang","raw_affiliation_strings":["Jilin Normal University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Jilin Normal University","institution_ids":["https://openalex.org/I149240348"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101773973","display_name":"Ning Xu","orcid":"https://orcid.org/0000-0001-8910-0937"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Ning Xu","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100917814","display_name":"Dawei Yang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210091958","display_name":"Hanoi Open University","ror":"https://ror.org/00gjkwr85","country_code":"VN","type":"education","lineage":["https://openalex.org/I4210091958"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Dawei Yang","raw_affiliation_strings":["Houmo AI"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Houmo AI","institution_ids":["https://openalex.org/I4210091958"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5101769415","display_name":"Zhe Jiang","orcid":"https://orcid.org/0000-0003-3108-0632"},"institutions":[{"id":"https://openalex.org/I4210090971","display_name":"Southeast University","ror":"https://ror.org/00cf0ab87","country_code":"BD","type":"education","lineage":["https://openalex.org/I4210090971"]}],"countries":["BD"],"is_corresponding":false,"raw_author_name":"Zhe Jiang","raw_affiliation_strings":["Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Southeast University,National Center of Technology Innovation for EDA, School of Integrated Circuits","institution_ids":["https://openalex.org/I4210090971"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":9,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":7.0351,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.96760991,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":91,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.8084999918937683,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.8084999918937683,"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/T10901","display_name":"Advanced Data Compression Techniques","score":0.7544000148773193,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10824","display_name":"Image Retrieval and Classification Techniques","score":0.7098000049591064,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/block","display_name":"Block (permutation group theory)","score":0.6873000264167786},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6398000121116638},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.6064000129699707},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.5092999935150146},{"id":"https://openalex.org/keywords/block-size","display_name":"Block size","score":0.4966999888420105},{"id":"https://openalex.org/keywords/point","display_name":"Point (geometry)","score":0.4652000069618225},{"id":"https://openalex.org/keywords/nonlinear-system","display_name":"Nonlinear system","score":0.374099999666214},{"id":"https://openalex.org/keywords/enhanced-data-rates-for-gsm-evolution","display_name":"Enhanced Data Rates for GSM Evolution","score":0.3328000009059906}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7232999801635742},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.6873000264167786},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6398000121116638},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.6064000129699707},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.5092999935150146},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.5073000192642212},{"id":"https://openalex.org/C41431624","wikidata":"https://www.wikidata.org/wiki/Q1053357","display_name":"Block size","level":3,"score":0.4966999888420105},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.4652000069618225},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3880000114440918},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.374099999666214},{"id":"https://openalex.org/C162307627","wikidata":"https://www.wikidata.org/wiki/Q204833","display_name":"Enhanced Data Rates for GSM Evolution","level":2,"score":0.3328000009059906},{"id":"https://openalex.org/C2779679103","wikidata":"https://www.wikidata.org/wiki/Q5251805","display_name":"Degradation (telecommunications)","level":2,"score":0.3206999897956848},{"id":"https://openalex.org/C176649486","wikidata":"https://www.wikidata.org/wiki/Q2308807","display_name":"Memory management","level":3,"score":0.3001999855041504},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.2799000144004822},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.2782000005245209},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.27799999713897705},{"id":"https://openalex.org/C138236772","wikidata":"https://www.wikidata.org/wiki/Q25098575","display_name":"Edge device","level":3,"score":0.2734000086784363},{"id":"https://openalex.org/C61445026","wikidata":"https://www.wikidata.org/wiki/Q217608","display_name":"Fixed point","level":2,"score":0.27250000834465027},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.2718999981880188},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.2526000142097473}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/dac63849.2025.11132978","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac63849.2025.11132978","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 62nd ACM/IEEE Design Automation Conference (DAC)","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":14,"referenced_works":["https://openalex.org/W1983394510","https://openalex.org/W2123736966","https://openalex.org/W2963679555","https://openalex.org/W2981849318","https://openalex.org/W3177265267","https://openalex.org/W3183327882","https://openalex.org/W3205807913","https://openalex.org/W3211525823","https://openalex.org/W3216547854","https://openalex.org/W4366341968","https://openalex.org/W4378364815","https://openalex.org/W4389519100","https://openalex.org/W4390889788","https://openalex.org/W4404133703"],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs),":[3],"with":[4,132],"their":[5],"billions":[6],"of":[7,58,72,92],"parameters,":[8],"pose":[9],"substantial":[10],"challenges":[11],"for":[12,119],"deployment":[13],"on":[14,129],"edge":[15],"devices,":[16],"straining":[17],"both":[18],"memory":[19,30],"capacity":[20],"and":[21,31,60,67,156],"computational":[22,32],"resources.":[23],"Block":[24,82,114],"Floating":[25,83,115],"Point":[26,84],"(BFP)":[27],"quantisation":[28,65,101],"reduces":[29,89],"overhead":[33],"by":[34],"converting":[35],"high-overhead":[36],"floating":[37],"point":[38,43],"operations":[39],"into":[40],"low-bit":[41],"fixed":[42],"operations.":[44],"However,":[45],"BFP":[46],"requires":[47],"aligning":[48],"all":[49],"data":[50,86],"to":[51,149],"the":[52,70,90,94,105],"maximum":[53,95],"exponent,":[54,98],"which":[55,88],"causes":[56],"loss":[57],"small":[59],"moderate":[61],"values,":[62],"resulting":[63],"in":[64,69,107,146],"error":[66],"degradation":[68],"accuracy":[71,147],"LLMs.":[73],"To":[74],"address":[75],"this":[76],"issue,":[77],"we":[78,109],"propose":[79],"a":[80,111,124,143,157,162],"Bidirectional":[81,113],"(BBFP)":[85],"format,":[87],"probability":[91],"selecting":[93],"as":[96],"shared":[97],"thereby":[99],"reducing":[100],"error.":[102],"By":[103],"utilizing":[104],"features":[106],"BBFP,":[108,130],"present":[110],"full-stack":[112],"Point-Based":[116],"Quantisation":[117],"Accelerator":[118],"LLMs":[120],"(BBAL),":[121],"primarily":[122],"comprising":[123],"processing":[125],"element":[126],"array":[127],"based":[128],"paired":[131],"proposed":[133],"cost-effective":[134],"nonlinear":[135],"computation":[136],"unit.":[137],"Experimental":[138],"results":[139],"show":[140],"BBAL":[141],"achieves":[142],"22%":[144],"improvement":[145,160],"compared":[148],"an":[150],"outlier-aware":[151],"accelerator":[152,164],"at":[153,165],"similar":[154,166],"efficiency,":[155],"40%":[158],"efficiency":[159],"over":[161],"BFP-based":[163],"accuracy.":[167]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":1}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}