{"id":"https://openalex.org/W4210823275","doi":"https://doi.org/10.1109/tnnls.2022.3145850","title":"ETA: An Efficient Training Accelerator for DNNs Based on Hardware-Algorithm Co-Optimization","display_name":"ETA: An Efficient Training Accelerator for DNNs Based on Hardware-Algorithm Co-Optimization","publication_year":2022,"publication_date":"2022-02-08","ids":{"openalex":"https://openalex.org/W4210823275","doi":"https://doi.org/10.1109/tnnls.2022.3145850","pmid":"https://pubmed.ncbi.nlm.nih.gov/35133969"},"language":"en","primary_location":{"id":"doi:10.1109/tnnls.2022.3145850","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tnnls.2022.3145850","pdf_url":null,"source":{"id":"https://openalex.org/S4210175523","display_name":"IEEE Transactions on Neural Networks and Learning Systems","issn_l":"2162-237X","issn":["2162-237X","2162-2388"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Neural Networks and Learning Systems","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","pubmed"],"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/A5059107361","display_name":"Jinming Lu","orcid":"https://orcid.org/0000-0002-7134-6514"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jinming Lu","raw_affiliation_strings":["School of Electronic Science and Engineering, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-7134-6514","affiliations":[{"raw_affiliation_string":"School of Electronic Science and Engineering, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029039080","display_name":"Chao Ni","orcid":"https://orcid.org/0000-0002-5139-4607"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chao Ni","raw_affiliation_strings":["School of Electronic Science and Engineering, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-5139-4607","affiliations":[{"raw_affiliation_string":"School of Electronic Science and Engineering, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100696999","display_name":"Zhongfeng Wang","orcid":"https://orcid.org/0000-0002-7227-4786"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhongfeng Wang","raw_affiliation_strings":["School of Electronic Science and Engineering, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-7227-4786","affiliations":[{"raw_affiliation_string":"School of Electronic Science and Engineering, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":2.8429,"has_fulltext":false,"cited_by_count":28,"citation_normalized_percentile":{"value":0.91894946,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":99},"biblio":{"volume":"34","issue":"10","first_page":"7660","last_page":"7674"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":1.0,"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"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":1.0,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9983999729156494,"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/T12676","display_name":"Machine Learning and ELM","score":0.9958999752998352,"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":0.7954055070877075},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.7828962802886963},{"id":"https://openalex.org/keywords/normalization","display_name":"Normalization (sociology)","score":0.6167151927947998},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.538214385509491},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.47591835260391235},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.466019868850708},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.4658981263637543},{"id":"https://openalex.org/keywords/hardware-acceleration","display_name":"Hardware acceleration","score":0.46471795439720154},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4526706039905548},{"id":"https://openalex.org/keywords/computational-complexity-theory","display_name":"Computational complexity theory","score":0.417648583650589},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.41230136156082153},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.36880767345428467},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.34642815589904785},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.2861069440841675},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2184252142906189}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7954055070877075},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.7828962802886963},{"id":"https://openalex.org/C136886441","wikidata":"https://www.wikidata.org/wiki/Q926129","display_name":"Normalization (sociology)","level":2,"score":0.6167151927947998},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.538214385509491},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.47591835260391235},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.466019868850708},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.4658981263637543},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.46471795439720154},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4526706039905548},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.417648583650589},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.41230136156082153},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.36880767345428467},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.34642815589904785},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2861069440841675},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2184252142906189},{"id":"https://openalex.org/C144024400","wikidata":"https://www.wikidata.org/wiki/Q21201","display_name":"Sociology","level":0,"score":0.0},{"id":"https://openalex.org/C19165224","wikidata":"https://www.wikidata.org/wiki/Q23404","display_name":"Anthropology","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/tnnls.2022.3145850","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tnnls.2022.3145850","pdf_url":null,"source":{"id":"https://openalex.org/S4210175523","display_name":"IEEE Transactions on Neural Networks and Learning Systems","issn_l":"2162-237X","issn":["2162-237X","2162-2388"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Neural Networks and Learning Systems","raw_type":"journal-article"},{"id":"pmid:35133969","is_oa":false,"landing_page_url":"https://pubmed.ncbi.nlm.nih.gov/35133969","pdf_url":null,"source":{"id":"https://openalex.org/S4306525036","display_name":"PubMed","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1299303238","host_organization_name":"National Institutes of Health","host_organization_lineage":["https://openalex.org/I1299303238"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE transactions on neural networks and learning systems","raw_type":null}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7","score":0.800000011920929}],"awards":[{"id":"https://openalex.org/G1941337620","display_name":null,"funder_award_id":"021014380065","funder_id":"https://openalex.org/F4320335787","funder_display_name":"Fundamental Research Funds for the Central Universities"},{"id":"https://openalex.org/G4112626697","display_name":null,"funder_award_id":"61774082","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320335787","display_name":"Fundamental Research Funds for the Central Universities","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":70,"referenced_works":["https://openalex.org/W1836465849","https://openalex.org/W1841592590","https://openalex.org/W2183341477","https://openalex.org/W2193413348","https://openalex.org/W2194775991","https://openalex.org/W2276486856","https://openalex.org/W2289252105","https://openalex.org/W2291160084","https://openalex.org/W2342840547","https://openalex.org/W2416197829","https://openalex.org/W2469490737","https://openalex.org/W2557355796","https://openalex.org/W2604319603","https://openalex.org/W2743322459","https://openalex.org/W2763421725","https://openalex.org/W2768993447","https://openalex.org/W2787094505","https://openalex.org/W2788388592","https://openalex.org/W2807571484","https://openalex.org/W2890068895","https://openalex.org/W2896457183","https://openalex.org/W2917518248","https://openalex.org/W2949335192","https://openalex.org/W2963163009","https://openalex.org/W2963855133","https://openalex.org/W2988559961","https://openalex.org/W2999599167","https://openalex.org/W3003315945","https://openalex.org/W3016829100","https://openalex.org/W3021654819","https://openalex.org/W3034733338","https://openalex.org/W3034940165","https://openalex.org/W3035016149","https://openalex.org/W3035232708","https://openalex.org/W3035460915","https://openalex.org/W3040850704","https://openalex.org/W3044934451","https://openalex.org/W3083199783","https://openalex.org/W3089022192","https://openalex.org/W3091592563","https://openalex.org/W3093799822","https://openalex.org/W3102122551","https://openalex.org/W3105966348","https://openalex.org/W3109212549","https://openalex.org/W3112412839","https://openalex.org/W3114479342","https://openalex.org/W3128120039","https://openalex.org/W3130875094","https://openalex.org/W3138562265","https://openalex.org/W3166859509","https://openalex.org/W4212788319","https://openalex.org/W4288346545","https://openalex.org/W6638667902","https://openalex.org/W6638783484","https://openalex.org/W6687566353","https://openalex.org/W6696798448","https://openalex.org/W6704559304","https://openalex.org/W6720242923","https://openalex.org/W6730248441","https://openalex.org/W6742278618","https://openalex.org/W6745245109","https://openalex.org/W6755207826","https://openalex.org/W6761694480","https://openalex.org/W6763653508","https://openalex.org/W6780724972","https://openalex.org/W6780997268","https://openalex.org/W6783246024","https://openalex.org/W6784309755","https://openalex.org/W6789240164","https://openalex.org/W6790866784"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W2146343568","https://openalex.org/W2532502681","https://openalex.org/W4319952061","https://openalex.org/W4280636456","https://openalex.org/W4388913998","https://openalex.org/W4310584535","https://openalex.org/W4295935044","https://openalex.org/W3159906349"],"abstract_inverted_index":{"Recently,":[0],"the":[1,28,100,110,201,241],"efficient":[2,52],"training":[3,30,53,69,125,198],"of":[4,38,190,236,249,256],"deep":[5],"neural":[6],"networks":[7,217],"(DNNs)":[8],"on":[9,56,215,224,258],"resource-constrained":[10],"platforms":[11],"has":[12],"attracted":[13],"increasing":[14],"attention":[15],"for":[16],"protecting":[17],"user":[18],"privacy.":[19],"However,":[20],"it":[21],"is":[22,71,117,165],"still":[23],"a":[24,35,63,86,93,112,135,140,161,181,187,247],"severe":[25],"challenge":[26],"since":[27],"DNN":[29],"involves":[31],"intensive":[32],"computations":[33],"and":[34,92,103,139,146,153,174,184,212,220,231,251],"large":[36],"amount":[37,189],"data":[39,90,147],"access.":[40],"To":[41],"deal":[42],"with":[43,80,240],"these":[44],"issues,":[45],"in":[46,84,180,234],"this":[47],"work,":[48],"we":[49],"implement":[50],"an":[51,252],"accelerator":[54],"(ETA)":[55],"field-programmable":[57],"gate":[58],"array":[59],"(FPGA)":[60],"by":[61],"adopting":[62],"hardware-algorithm":[64],"co-optimization":[65],"approach.":[66],"A":[67],"novel":[68],"scheme":[70,199],"proposed":[72],"to":[73,119,167],"effectively":[74],"train":[75],"DNNs":[76],"using":[77],"8-bit":[78],"precision":[79],"arbitrary":[81],"batch":[82],"sizes,":[83],"which":[85,178],"compact":[87],"but":[88],"powerful":[89],"format":[91],"hardware-oriented":[94],"normalization":[95,173],"layer":[96],"are":[97,106,156],"introduced.":[98],"Thus":[99],"computational":[101,122,144],"complexity":[102],"memory":[104,154],"accesses":[105,155],"significantly":[107],"reduced.":[108,158],"In":[109,159],"ETA,":[111],"reconfigurable":[113],"processing":[114],"element":[115],"(PE)":[116],"designed":[118],"support":[120],"various":[121],"patterns":[123],"during":[124],"while":[126],"avoiding":[127],"redundant":[128],"calculations":[129],"from":[130],"nonunit-stride":[131],"convolutional":[132],"layers.":[133],"With":[134],"flexible":[136],"network-on-chip":[137],"(NoC)":[138],"hierarchical":[141],"PE":[142],"array,":[143],"parallelism":[145],"reuse":[148],"can":[149],"be":[150],"fully":[151],"exploited,":[152],"further":[157],"addition,":[160],"unified":[162],"computing":[163],"core":[164],"developed":[166],"execute":[168],"auxiliary":[169],"layers":[170],"such":[171],"as":[172],"weight":[175],"update":[176],"(WU),":[177],"works":[179],"time-multiplexed":[182],"manner":[183],"consumes":[185],"only":[186],"small":[188],"hardware":[191],"resources.":[192],"The":[193],"experiments":[194],"show":[195],"that":[196],"our":[197,222,244],"achieves":[200,228],"state-of-the-art":[202],"accuracy":[203],"across":[204],"multiple":[205],"models,":[206],"including":[207],"CIFAR-VGG16,":[208],"CIFAR-ResNet20,":[209,219],"CIFAR-InceptionV3,":[210],"ResNet18,":[211],"ResNet50.":[213],"Evaluated":[214],"three":[216],"(CIFAR-VGG16,":[218],"ResNet18),":[221],"ETA":[223],"Xilinx":[225],"VC709":[226],"FPGA":[227],"610.98,":[229],"658.64,":[230],"811.24":[232],"GOPS":[233],"terms":[235],"throughput,":[237],"respectively.":[238],"Compared":[239],"prior":[242],"art,":[243],"design":[245],"demonstrates":[246],"speedup":[248],"3.65\u00d7":[250],"energy":[253],"efficiency":[254],"improvement":[255],"8.54\u00d7":[257],"CIFAR-ResNet20.":[259]},"counts_by_year":[{"year":2025,"cited_by_count":12},{"year":2024,"cited_by_count":5},{"year":2023,"cited_by_count":9},{"year":2022,"cited_by_count":2}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}