{"id":"https://openalex.org/W4400233352","doi":"https://doi.org/10.1109/iscas58744.2024.10558053","title":"A 28nm Energy-Area-Efficient Row-based pipelined Training Accelerator with Mixed FXP4/FP16 for On-Device Transfer Learning","display_name":"A 28nm Energy-Area-Efficient Row-based pipelined Training Accelerator with Mixed FXP4/FP16 for On-Device Transfer Learning","publication_year":2024,"publication_date":"2024-05-19","ids":{"openalex":"https://openalex.org/W4400233352","doi":"https://doi.org/10.1109/iscas58744.2024.10558053"},"language":"en","primary_location":{"id":"doi:10.1109/iscas58744.2024.10558053","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas58744.2024.10558053","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5002275070","display_name":"Wei Lu","orcid":"https://orcid.org/0000-0003-4769-9261"},"institutions":[{"id":"https://openalex.org/I148366613","display_name":"National Yang Ming Chiao Tung University","ror":"https://ror.org/00se2k293","country_code":"TW","type":"education","lineage":["https://openalex.org/I148366613"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Wei Lu","raw_affiliation_strings":["National Yang Ming Chiao Tung University,Institute of Electronics,Taiwan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National Yang Ming Chiao Tung University,Institute of Electronics,Taiwan","institution_ids":["https://openalex.org/I148366613"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100317090","display_name":"Han-Hsiang Pei","orcid":null},"institutions":[{"id":"https://openalex.org/I148366613","display_name":"National Yang Ming Chiao Tung University","ror":"https://ror.org/00se2k293","country_code":"TW","type":"education","lineage":["https://openalex.org/I148366613"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Han-Hsiang Pei","raw_affiliation_strings":["National Yang Ming Chiao Tung University,International College of Semiconductor Technology,Taiwan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National Yang Ming Chiao Tung University,International College of Semiconductor Technology,Taiwan","institution_ids":["https://openalex.org/I148366613"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100598240","display_name":"Jheng-Rong Yu","orcid":null},"institutions":[{"id":"https://openalex.org/I148366613","display_name":"National Yang Ming Chiao Tung University","ror":"https://ror.org/00se2k293","country_code":"TW","type":"education","lineage":["https://openalex.org/I148366613"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Jheng-Rong Yu","raw_affiliation_strings":["National Yang Ming Chiao Tung University,International College of Semiconductor Technology,Taiwan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National Yang Ming Chiao Tung University,International College of Semiconductor Technology,Taiwan","institution_ids":["https://openalex.org/I148366613"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021615416","display_name":"Hung-Ming Chen","orcid":"https://orcid.org/0000-0001-8173-3131"},"institutions":[{"id":"https://openalex.org/I148366613","display_name":"National Yang Ming Chiao Tung University","ror":"https://ror.org/00se2k293","country_code":"TW","type":"education","lineage":["https://openalex.org/I148366613"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Hung-Ming Chen","raw_affiliation_strings":["National Yang Ming Chiao Tung University,Institute of Electronics,Taiwan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National Yang Ming Chiao Tung University,Institute of Electronics,Taiwan","institution_ids":["https://openalex.org/I148366613"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5066525875","display_name":"Po-Tsang Huang","orcid":"https://orcid.org/0000-0001-8679-2755"},"institutions":[{"id":"https://openalex.org/I148366613","display_name":"National Yang Ming Chiao Tung University","ror":"https://ror.org/00se2k293","country_code":"TW","type":"education","lineage":["https://openalex.org/I148366613"]}],"countries":["TW"],"is_corresponding":false,"raw_author_name":"Po-Tsang Huang","raw_affiliation_strings":["National Yang Ming Chiao Tung University,International College of Semiconductor Technology,Taiwan"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"National Yang Ming Chiao Tung University,International College of Semiconductor Technology,Taiwan","institution_ids":["https://openalex.org/I148366613"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I148366613"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9670000076293945,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9670000076293945,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11992","display_name":"CCD and CMOS Imaging Sensors","score":0.9174000024795532,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7372307777404785},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.5307915806770325},{"id":"https://openalex.org/keywords/transfer","display_name":"Transfer (computing)","score":0.5012741088867188},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4551633596420288},{"id":"https://openalex.org/keywords/transfer-of-learning","display_name":"Transfer of learning","score":0.44497764110565186},{"id":"https://openalex.org/keywords/energy","display_name":"Energy (signal processing)","score":0.43063458800315857},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.41317862272262573},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.33937400579452515},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.31873753666877747},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.08882170915603638}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7372307777404785},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.5307915806770325},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.5012741088867188},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4551633596420288},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.44497764110565186},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.43063458800315857},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.41317862272262573},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.33937400579452515},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.31873753666877747},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.08882170915603638},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iscas58744.2024.10558053","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iscas58744.2024.10558053","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE International Symposium on Circuits and Systems (ISCAS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","score":0.8600000143051147,"id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":15,"referenced_works":["https://openalex.org/W2108598243","https://openalex.org/W2194775991","https://openalex.org/W2289252105","https://openalex.org/W2773339846","https://openalex.org/W2792607576","https://openalex.org/W2884367402","https://openalex.org/W3007895486","https://openalex.org/W3015729306","https://openalex.org/W3015764133","https://openalex.org/W3048842964","https://openalex.org/W3172051607","https://openalex.org/W4241715658","https://openalex.org/W4285264221","https://openalex.org/W6682778277","https://openalex.org/W6797463349"],"related_works":["https://openalex.org/W3201126466","https://openalex.org/W4282827391","https://openalex.org/W3165580226","https://openalex.org/W4386828785","https://openalex.org/W3135401135","https://openalex.org/W2886688859","https://openalex.org/W3133164560","https://openalex.org/W2758348317","https://openalex.org/W2604564350","https://openalex.org/W2977607533"],"abstract_inverted_index":{"Training":[0],"deep":[1],"convolutional":[2,97,134],"neural":[3],"networks":[4],"(DNNs)":[5],"requires":[6],"significantly":[7],"more":[8,34],"computational":[9],"capacity,":[10],"complex":[11],"dataflow,":[12],"memory":[13],"accesses,":[14],"and":[15,36,46,68,84,99,149,157,172,195,203,220],"data":[16,170],"movement":[17],"among":[18],"processing":[19],"elements":[20],"(PEs),":[21],"as":[22,24],"well":[23],"higher":[25],"bit":[26],"precision":[27],"for":[28,71,96,133,145,176],"back":[29],"propagation":[30],"(BP),":[31],"which":[32],"demands":[33],"power":[35],"area":[37,47,69,205,218],"overhead":[38],"than":[39],"DNN":[40,58],"inference.":[41],"For":[42],"mobile/edge":[43],"devices,":[44],"energy":[45,67,198,224],"efficiency":[48,70,199,206,219],"are":[49],"critical":[50],"concerns.":[51],"This":[52],"research":[53],"proposes":[54],"a":[55,137],"row-based":[56,138],"pipelined":[57,140],"training":[59,141],"accelerator":[60,142,186],"that":[61,111],"employs":[62],"three":[63],"techniques":[64],"to":[65,153,168,181],"improve":[66],"resource-constrained":[72],"edge/mobile":[73],"devices.":[74],"The":[75,88,107,161,184],"first":[76],"technique":[77,90],"involves":[78,91],"freezing":[79],"weight":[80],"updates":[81],"in":[82],"convolution":[83,113],"batch":[85,104,150],"normalization":[86,105,151],"layers.":[87,106,135],"second":[89],"decomposing":[92],"the":[93,101,129,155,174,190,217,223],"simulated":[94],"quantization":[95],"layers":[98,146,152],"reorganizing":[100],"operations":[102,114],"of":[103,200,207],"mathematical":[108],"demonstration":[109],"shows":[110],"FP":[112,131,158],"can":[115,127],"be":[116],"completed":[117],"using":[118,189],"fixed":[119],"point":[120],"(FXP)":[121],"calculations.":[122],"FXP":[123,156],"MACs":[124,132],"with":[125,214],"dequantizer":[126],"replace":[128],"original":[130],"Additionally,":[136],"FXP/FP":[139],"is":[143],"designed":[144],"pipeline,":[147],"convolution,":[148],"increase":[154],"resource":[159],"utilization.":[160],"third":[162],"method":[163],"uses":[164],"multi-bank":[165],"buffer":[166],"management":[167],"prevent":[169],"conflicts":[171],"reduce":[173],"need":[175],"on-chip":[177],"buffers":[178],"by":[179],"up":[180],"3.5":[182],"times.":[183],"proposed":[185],"was":[187],"implemented":[188],"TSMC":[191],"28nm":[192],"CMOS":[193],"process":[194],"achieved":[196],"an":[197,204],"2.19":[201],"TFLOPS/W":[202],"85.32":[208],"GFLOPS/mm2.":[209],"It":[210],"outperforms":[211],"state-of-the-art":[212],"works":[213],"6.8":[215],"times":[216,222],"3.7":[221],"efficiency.":[225]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":3}],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}