{"id":"https://openalex.org/W4404308449","doi":"https://doi.org/10.1145/3676536.3676747","title":"MCUBERT: Memory-Efficient BERT Inference on Commodity Microcontrollers","display_name":"MCUBERT: Memory-Efficient BERT Inference on Commodity Microcontrollers","publication_year":2024,"publication_date":"2024-10-27","ids":{"openalex":"https://openalex.org/W4404308449","doi":"https://doi.org/10.1145/3676536.3676747"},"language":"en","primary_location":{"id":"doi:10.1145/3676536.3676747","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3676536.3676747","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3676536.3676747","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 43rd IEEE/ACM International Conference on Computer-Aided Design","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3676536.3676747","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5114950458","display_name":"Zebin Yang","orcid":"https://orcid.org/0009-0007-0672-6389"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]},{"id":"https://openalex.org/I4210100255","display_name":"Beijing Academy of Artificial Intelligence","ror":"https://ror.org/016a74861","country_code":"CN","type":"other","lineage":["https://openalex.org/I4210100255"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zebin Yang","raw_affiliation_strings":["Institute for Artificial Intelligence, Peking University, Beijing, China","School of Integrated Circuits, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0007-0672-6389","affiliations":[{"raw_affiliation_string":"Institute for Artificial Intelligence, Peking University, Beijing, China","institution_ids":["https://openalex.org/I4210100255","https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Integrated Circuits, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004459947","display_name":"Renze Chen","orcid":"https://orcid.org/0000-0001-5938-7965"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Renze Chen","raw_affiliation_strings":["School of Computer Science, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0001-5938-7965","affiliations":[{"raw_affiliation_string":"School of Computer Science, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074532565","display_name":"Taiqiang Wu","orcid":"https://orcid.org/0000-0002-3664-3513"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]},{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Taiqiang Wu","raw_affiliation_strings":["The University of Hong Kong, Hong Kong, Hong Kong"],"raw_orcid":"https://orcid.org/0000-0002-3664-3513","affiliations":[{"raw_affiliation_string":"The University of Hong Kong, Hong Kong, Hong Kong","institution_ids":["https://openalex.org/I177725633","https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5072120057","display_name":"Ngai Wong","orcid":"https://orcid.org/0000-0002-3026-0108"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]},{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Ngai Wong","raw_affiliation_strings":["The University of Hong Kong, Hong Kong, Hong Kong"],"raw_orcid":"https://orcid.org/0000-0002-3026-0108","affiliations":[{"raw_affiliation_string":"The University of Hong Kong, Hong Kong, Hong Kong","institution_ids":["https://openalex.org/I177725633","https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100604860","display_name":"Yun Liang","orcid":"https://orcid.org/0000-0002-9076-7998"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]},{"id":"https://openalex.org/I4210165198","display_name":"Beijing Advanced Sciences and Innovation Center","ror":"https://ror.org/05qm21180","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210165198"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yun Liang","raw_affiliation_strings":["Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","School of Integrated Circuits, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-9076-7998","affiliations":[{"raw_affiliation_string":"Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","institution_ids":["https://openalex.org/I4210165198"]},{"raw_affiliation_string":"School of Integrated Circuits, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5002760019","display_name":"Runsheng Wang","orcid":"https://orcid.org/0000-0002-7514-0767"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]},{"id":"https://openalex.org/I4210165198","display_name":"Beijing Advanced Sciences and Innovation Center","ror":"https://ror.org/05qm21180","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210165198"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Runsheng Wang","raw_affiliation_strings":["Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","Institute of Electronic Design Automation, Peking University, Wuxi, Jiangsu, China","School of Integrated Circuits, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-7514-0767","affiliations":[{"raw_affiliation_string":"Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","institution_ids":["https://openalex.org/I4210165198"]},{"raw_affiliation_string":"Institute of Electronic Design Automation, Peking University, Wuxi, Jiangsu, China","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Integrated Circuits, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062886480","display_name":"Ru Huang","orcid":"https://orcid.org/0000-0002-8146-4821"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]},{"id":"https://openalex.org/I4210165198","display_name":"Beijing Advanced Sciences and Innovation Center","ror":"https://ror.org/05qm21180","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210165198"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ru Huang","raw_affiliation_strings":["Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","Institute of Electronic Design Automation, Peking University, Wuxi, Jiangsu, China","School of Integrated Circuits, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-8146-4821","affiliations":[{"raw_affiliation_string":"Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","institution_ids":["https://openalex.org/I4210165198"]},{"raw_affiliation_string":"Institute of Electronic Design Automation, Peking University, Wuxi, Jiangsu, China","institution_ids":["https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Integrated Circuits, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100457502","display_name":"Meng Li","orcid":"https://orcid.org/0000-0002-7212-2264"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]},{"id":"https://openalex.org/I4210100255","display_name":"Beijing Academy of Artificial Intelligence","ror":"https://ror.org/016a74861","country_code":"CN","type":"other","lineage":["https://openalex.org/I4210100255"]},{"id":"https://openalex.org/I4210165198","display_name":"Beijing Advanced Sciences and Innovation Center","ror":"https://ror.org/05qm21180","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210165198"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Meng Li","raw_affiliation_strings":["Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","Institute for Artificial Intelligence, Peking University, Beijing, China","School of Integrated Circuits, Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-7212-2264","affiliations":[{"raw_affiliation_string":"Beijing Advanced Innovation Center for Integrated Circuits, Beijing, China","institution_ids":["https://openalex.org/I4210165198"]},{"raw_affiliation_string":"Institute for Artificial Intelligence, Peking University, Beijing, China","institution_ids":["https://openalex.org/I4210100255","https://openalex.org/I20231570"]},{"raw_affiliation_string":"School of Integrated Circuits, Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5100457502"],"corresponding_institution_ids":["https://openalex.org/I20231570","https://openalex.org/I4210100255","https://openalex.org/I4210165198"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":3,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"9"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12254","display_name":"Machine Learning in Bioinformatics","score":0.9970999956130981,"subfield":{"id":"https://openalex.org/subfields/1312","display_name":"Molecular Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},"topics":[{"id":"https://openalex.org/T12254","display_name":"Machine Learning in Bioinformatics","score":0.9970999956130981,"subfield":{"id":"https://openalex.org/subfields/1312","display_name":"Molecular Biology"},"field":{"id":"https://openalex.org/fields/13","display_name":"Biochemistry, Genetics and Molecular Biology"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9887999892234802,"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/T10627","display_name":"Advanced Image and Video Retrieval Techniques","score":0.9882000088691711,"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/computer-science","display_name":"Computer science","score":0.7971510887145996},{"id":"https://openalex.org/keywords/microcontroller","display_name":"Microcontroller","score":0.6711372137069702},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.5766744613647461},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.48188501596450806},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.478302538394928},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.46544942259788513},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.42955508828163147},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.39715132117271423},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.1408209502696991},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.07711192965507507}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7971510887145996},{"id":"https://openalex.org/C173018170","wikidata":"https://www.wikidata.org/wiki/Q165678","display_name":"Microcontroller","level":2,"score":0.6711372137069702},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.5766744613647461},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.48188501596450806},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.478302538394928},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.46544942259788513},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.42955508828163147},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.39715132117271423},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.1408209502696991},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.07711192965507507},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3676536.3676747","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3676536.3676747","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3676536.3676747","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 43rd IEEE/ACM International Conference on Computer-Aided Design","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2410.17957","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2410.17957","pdf_url":"https://arxiv.org/pdf/2410.17957","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"doi:10.1145/3676536.3676747","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3676536.3676747","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3676536.3676747","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 43rd IEEE/ACM International Conference on Computer-Aided Design","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1953188521","display_name":null,"funder_award_id":"B18001","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G4817879587","display_name":null,"funder_award_id":"62125401","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G6847781164","display_name":null,"funder_award_id":"B18001","funder_id":"https://openalex.org/F4320327912","funder_display_name":"Higher Education Discipline Innovation Project"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320327912","display_name":"Higher Education Discipline Innovation Project","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4404308449.pdf","grobid_xml":"https://content.openalex.org/works/W4404308449.grobid-xml"},"referenced_works_count":20,"referenced_works":["https://openalex.org/W2947515409","https://openalex.org/W2970106668","https://openalex.org/W2996854111","https://openalex.org/W2998183051","https://openalex.org/W3038827379","https://openalex.org/W3044604993","https://openalex.org/W3081168214","https://openalex.org/W3093348380","https://openalex.org/W3094136258","https://openalex.org/W3130716829","https://openalex.org/W3167266074","https://openalex.org/W3197527602","https://openalex.org/W3202586005","https://openalex.org/W4226033575","https://openalex.org/W4281758439","https://openalex.org/W4287639545","https://openalex.org/W4287748555","https://openalex.org/W4288346093","https://openalex.org/W4384705353","https://openalex.org/W6780827055"],"related_works":["https://openalex.org/W2595172197","https://openalex.org/W2084856301","https://openalex.org/W2127970246","https://openalex.org/W2885125400","https://openalex.org/W1989889224","https://openalex.org/W4382618745","https://openalex.org/W1973775000","https://openalex.org/W2748922771","https://openalex.org/W1987128138","https://openalex.org/W1981130628"],"abstract_inverted_index":{"In":[0],"this":[1],"paper,":[2],"we":[3,42,65,85],"propose":[4,43,67],"MCUBERT":[5,101,123,133],"to":[6,28],"enable":[7],"language":[8],"models":[9,137],"like":[10],"BERT":[11,35,136],"on":[12,52,93,138],"tiny":[13,34],"microcontroller":[14],"units":[15],"(MCUs)":[16],"through":[17],"network":[18,40],"and":[19,78,108,112,114,120,141],"scheduling":[20,72],"co-optimization.":[21],"We":[22],"observe":[23],"the":[24,29,39,61,88,103,115,130],"embedding":[25,57],"table":[26],"contributes":[27],"major":[30],"storage":[31],"bottleneck":[32],"for":[33,56],"models.":[36],"Hence,":[37],"at":[38],"level,":[41],"an":[44],"MCU-aware":[45],"two-stage":[46],"neural":[47],"architecture":[48],"search":[49],"algorithm":[50],"based":[51],"clustered":[53],"low-rank":[54],"approximation":[55],"compression.":[58],"To":[59],"reduce":[60],"inference":[62],"memory":[63,117],"requirements,":[64],"further":[66],"a":[68],"novel":[69],"fine-grained":[70],"MCU-friendly":[71],"strategy.":[73],"Through":[74],"careful":[75],"computation":[76],"tiling":[77],"re-ordering":[79],"as":[80,82],"well":[81],"kernel":[83],"design,":[84],"drastically":[86],"increase":[87],"input":[89],"sequence":[90],"lengths":[91],"supported":[92],"MCUs":[94,140],"without":[95],"any":[96],"latency":[97,127],"or":[98],"accuracy":[99],"penalty.":[100],"reduces":[102],"parameter":[104],"size":[105],"of":[106,151],"BERT-tiny":[107],"BERT-mini":[109],"by":[110,118],"5.7\u00d7":[111],"3.0\u00d7":[113],"execution":[116],"3.5\u00d7":[119],"4.3\u00d7,":[121],"respectively.":[122],"also":[124],"achieves":[125],"1.5\u00d7":[126],"reduction.":[128],"For":[129],"first":[131],"time,":[132],"enables":[134],"lightweight":[135],"commodity":[139],"processing":[142],"more":[143],"than":[144,149],"512":[145],"tokens":[146],"with":[147],"less":[148],"256KB":[150],"memory.":[152]},"counts_by_year":[{"year":2026,"cited_by_count":3}],"updated_date":"2026-07-16T13:24:37.021932","created_date":"2025-10-10T00:00:00"}
