{"id":"https://openalex.org/W4312872667","doi":"https://doi.org/10.1109/tvlsi.2022.3226185","title":"Multiple-Mode-Supporting Floating-Point FMA Unit for Deep Learning Processors","display_name":"Multiple-Mode-Supporting Floating-Point FMA Unit for Deep Learning Processors","publication_year":2022,"publication_date":"2022-12-14","ids":{"openalex":"https://openalex.org/W4312872667","doi":"https://doi.org/10.1109/tvlsi.2022.3226185"},"language":"en","primary_location":{"id":"doi:10.1109/tvlsi.2022.3226185","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tvlsi.2022.3226185","pdf_url":null,"source":{"id":"https://openalex.org/S37538908","display_name":"IEEE Transactions on Very Large Scale Integration (VLSI) Systems","issn_l":"1063-8210","issn":["1063-8210","1557-9999"],"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 Very Large Scale Integration (VLSI) Systems","raw_type":"journal-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/A5019476529","display_name":"Hongbing Tan","orcid":"https://orcid.org/0000-0003-4184-4173"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongbing Tan","raw_affiliation_strings":["College of Computer, National University of Defense Technology, Changsha, China"],"raw_orcid":"https://orcid.org/0000-0003-4184-4173","affiliations":[{"raw_affiliation_string":"College of Computer, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5042525314","display_name":"Gan Tong","orcid":"https://orcid.org/0000-0002-4660-6150"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Gan Tong","raw_affiliation_strings":["College of Computer, National University of Defense Technology, Changsha, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"College of Computer, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5044840341","display_name":"Libo Huang","orcid":"https://orcid.org/0000-0001-7878-3998"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Libo Huang","raw_affiliation_strings":["College of Computer, National University of Defense Technology, Changsha, China"],"raw_orcid":"https://orcid.org/0000-0001-7878-3998","affiliations":[{"raw_affiliation_string":"College of Computer, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100648910","display_name":"Liquan Xiao","orcid":"https://orcid.org/0000-0002-3285-2625"},"institutions":[{"id":"https://openalex.org/I170215575","display_name":"National University of Defense Technology","ror":"https://ror.org/05d2yfz11","country_code":"CN","type":"education","lineage":["https://openalex.org/I170215575"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Liquan Xiao","raw_affiliation_strings":["College of Computer, National University of Defense Technology, Changsha, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"College of Computer, National University of Defense Technology, Changsha, China","institution_ids":["https://openalex.org/I170215575"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5023506057","display_name":"Nong Xiao","orcid":"https://orcid.org/0000-0002-2166-977X"},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Nong Xiao","raw_affiliation_strings":["School of Data and Computer Science, Sun Yat-sen University, Guangzhou, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Data and Computer Science, Sun Yat-sen University, Guangzhou, China","institution_ids":["https://openalex.org/I157773358"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":4.1998,"has_fulltext":false,"cited_by_count":20,"citation_normalized_percentile":{"value":0.95090486,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":99},"biblio":{"volume":"31","issue":"2","first_page":"253","last_page":"266"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9987999796867371,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9987999796867371,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T11697","display_name":"Numerical Methods and Algorithms","score":0.9977999925613403,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.9941999912261963,"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/datapath","display_name":"Datapath","score":0.6929861903190613},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6335840821266174},{"id":"https://openalex.org/keywords/floating-point-unit","display_name":"Floating-point unit","score":0.5693913698196411},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.5656823515892029},{"id":"https://openalex.org/keywords/operand","display_name":"Operand","score":0.5611948370933533},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.5307883024215698},{"id":"https://openalex.org/keywords/single-precision-floating-point-format","display_name":"Single-precision floating-point format","score":0.44231724739074707},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.4358804225921631},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4346621334552765},{"id":"https://openalex.org/keywords/dot-product","display_name":"Dot product","score":0.43032991886138916},{"id":"https://openalex.org/keywords/multiplication","display_name":"Multiplication (music)","score":0.4240914583206177},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4168543815612793},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.36815834045410156},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.31683850288391113},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.2564508318901062},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2540726065635681},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.15148961544036865}],"concepts":[{"id":"https://openalex.org/C2781198647","wikidata":"https://www.wikidata.org/wiki/Q1633673","display_name":"Datapath","level":2,"score":0.6929861903190613},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6335840821266174},{"id":"https://openalex.org/C110305270","wikidata":"https://www.wikidata.org/wiki/Q733507","display_name":"Floating-point unit","level":3,"score":0.5693913698196411},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.5656823515892029},{"id":"https://openalex.org/C55526617","wikidata":"https://www.wikidata.org/wiki/Q719375","display_name":"Operand","level":2,"score":0.5611948370933533},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.5307883024215698},{"id":"https://openalex.org/C133095886","wikidata":"https://www.wikidata.org/wiki/Q1307173","display_name":"Single-precision floating-point format","level":3,"score":0.44231724739074707},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.4358804225921631},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4346621334552765},{"id":"https://openalex.org/C32900221","wikidata":"https://www.wikidata.org/wiki/Q181365","display_name":"Dot product","level":2,"score":0.43032991886138916},{"id":"https://openalex.org/C2780595030","wikidata":"https://www.wikidata.org/wiki/Q3860309","display_name":"Multiplication (music)","level":2,"score":0.4240914583206177},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4168543815612793},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.36815834045410156},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.31683850288391113},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2564508318901062},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2540726065635681},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.15148961544036865},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C555944384","wikidata":"https://www.wikidata.org/wiki/Q249","display_name":"Wireless","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tvlsi.2022.3226185","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tvlsi.2022.3226185","pdf_url":null,"source":{"id":"https://openalex.org/S37538908","display_name":"IEEE Transactions on Very Large Scale Integration (VLSI) Systems","issn_l":"1063-8210","issn":["1063-8210","1557-9999"],"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 Very Large Scale Integration (VLSI) Systems","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G4459438753","display_name":null,"funder_award_id":"62272475","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G7974704810","display_name":null,"funder_award_id":"618722374","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"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":38,"referenced_works":["https://openalex.org/W1980697457","https://openalex.org/W1999085092","https://openalex.org/W2008424999","https://openalex.org/W2026445983","https://openalex.org/W2082444966","https://openalex.org/W2142464516","https://openalex.org/W2144921006","https://openalex.org/W2147479230","https://openalex.org/W2194775991","https://openalex.org/W2224617858","https://openalex.org/W2259472270","https://openalex.org/W2260663238","https://openalex.org/W2275025625","https://openalex.org/W2497255926","https://openalex.org/W2541452119","https://openalex.org/W2604319603","https://openalex.org/W2606722458","https://openalex.org/W2618530766","https://openalex.org/W2750822049","https://openalex.org/W2763421725","https://openalex.org/W2781383697","https://openalex.org/W2801795932","https://openalex.org/W2914973130","https://openalex.org/W2952046647","https://openalex.org/W2981022124","https://openalex.org/W2982219368","https://openalex.org/W2983490771","https://openalex.org/W3021029305","https://openalex.org/W3190062760","https://openalex.org/W3217237164","https://openalex.org/W4221160294","https://openalex.org/W4233996382","https://openalex.org/W4288346545","https://openalex.org/W6638783484","https://openalex.org/W6692521979","https://openalex.org/W6692563993","https://openalex.org/W6745245109","https://openalex.org/W6763653508"],"related_works":["https://openalex.org/W2150733028","https://openalex.org/W3137832335","https://openalex.org/W2008844697","https://openalex.org/W2186297658","https://openalex.org/W2123623986","https://openalex.org/W2001966371","https://openalex.org/W2020888328","https://openalex.org/W2124576060","https://openalex.org/W2102729529","https://openalex.org/W2037018251"],"abstract_inverted_index":{"In":[0,53,74,134],"this":[1,147],"article,":[2],"a":[3,105],"new":[4],"multiple-mode":[5],"floating-point":[6,213],"fused":[7,86],"multiply\u2013add":[8],"(FMA)":[9],"unit":[10],"is":[11,117,143],"proposed":[12,18,207],"for":[13],"deep":[14,50,169],"learning":[15,51],"processors.":[16],"The":[17,174],"design":[19,176,208],"supports":[20,209],"three":[21],"functional":[22],"modes\u2014normal":[23],"FMA":[24,27,31,56,59,77,81,113,122,137,151,175,204],"mode,":[25,28,57,82,114],"mixed":[26,76,112],"and":[29,42,79,88,183,193,197,215,220],"dual":[30,80,121],"mode\u2014and":[32],"four":[33],"types":[34,211],"of":[35,49,97,129,168,212],"precision\u2014single-precision":[36],"(SP),":[37],"half-precision":[38],"(HP),":[39],"BFloat16":[40],"(BF16),":[41],"TensorFloat-32":[43],"(TF32)\u2014based":[44],"on":[45],"the":[46,54,75,85,89,95,111,115,120,126,130,162,166,181,201,206],"practical":[47],"requirements":[48],"applications.":[52],"normal":[55],"conventional":[58],"operations,":[60,68,84,138],"one":[61],"SP":[62],"operation":[63,214],"or":[64],"two":[65],"parallel":[66],"HP":[67],"are":[69,91],"performed":[70],"every":[71],"clock":[72],"cycle.":[73],"mode":[78,123],"mixed-precision":[83],"multiply\u2013accumulate":[87],"dot-product,":[90],"implemented,":[92],"respectively.":[93],"Specifically,":[94],"product":[96],"lower":[98,157,221],"precision":[99,107,140,150,158],"multiplication":[100],"can":[101,153],"be":[102,154],"accumulated":[103],"to":[104,125,136,161,186],"higher":[106,149,218],"addend.":[108],"Compared":[109,199],"with":[110,200,217],"throughput":[116,192,219],"doubled":[118],"in":[119,146,165],"due":[124],"full":[127],"utilization":[128],"multiplier":[131],"operand":[132],"bandwidth.":[133],"addition":[135],"numerical":[139],"conversion":[141],"(NPCvt)":[142],"also":[144],"supported":[145],"work:":[148],"results":[152],"converted":[155],"into":[156],"numbers,":[159],"corresponding":[160],"datatype":[163],"transform":[164],"datapath":[167],"neural":[170],"network":[171],"(DNN)":[172],"training.":[173],"presented":[177],"herein":[178],"uses":[179],"both":[180],"segmentation":[182],"reusing":[184],"methods":[185],"trade":[187],"off":[188],"performance,":[189],"such":[190],"as":[191],"latency,":[194],"against":[195],"area,":[196],"power.":[198],"state-of-the-art":[202],"multiple-precision":[203],"unit,":[205],"more":[210],"NPCvt,":[216],"hardware":[222],"overhead.":[223]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":5},{"year":2024,"cited_by_count":7},{"year":2023,"cited_by_count":6}],"updated_date":"2026-06-19T17:40:00.097472","created_date":"2025-10-10T00:00:00"}