{"id":"https://openalex.org/W4323345742","doi":"https://doi.org/10.1109/tcad.2023.3253045","title":"ApproxTrain: Fast Simulation of Approximate Multipliers for DNN Training and Inference","display_name":"ApproxTrain: Fast Simulation of Approximate Multipliers for DNN Training and Inference","publication_year":2023,"publication_date":"2023-03-06","ids":{"openalex":"https://openalex.org/W4323345742","doi":"https://doi.org/10.1109/tcad.2023.3253045"},"language":"en","primary_location":{"id":"doi:10.1109/tcad.2023.3253045","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcad.2023.3253045","pdf_url":null,"source":{"id":"https://openalex.org/S100835903","display_name":"IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems","issn_l":"0278-0070","issn":["0278-0070","1937-4151"],"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 Computer-Aided Design of Integrated Circuits and 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/A5086185655","display_name":"Jing Gong","orcid":"https://orcid.org/0000-0002-5820-8985"},"institutions":[{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":true,"raw_author_name":"Jing Gong","raw_affiliation_strings":["School of Computer Science and Engineering, UNSW Sydney, Sydney, NSW, Australia"],"affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, UNSW Sydney, Sydney, NSW, Australia","institution_ids":["https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055160617","display_name":"Hassaan Saadat","orcid":"https://orcid.org/0000-0003-3691-4130"},"institutions":[{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Hassaan Saadat","raw_affiliation_strings":["School of Electrical Engineering and Telecommunications, UNSW Sydney, Sydney, NSW, Australia"],"affiliations":[{"raw_affiliation_string":"School of Electrical Engineering and Telecommunications, UNSW Sydney, Sydney, NSW, Australia","institution_ids":["https://openalex.org/I31746571"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050266267","display_name":"Hasindu Gamaarachchi","orcid":"https://orcid.org/0000-0002-9034-9905"},"institutions":[{"id":"https://openalex.org/I31746571","display_name":"UNSW Sydney","ror":"https://ror.org/03r8z3t63","country_code":"AU","type":"education","lineage":["https://openalex.org/I31746571"]},{"id":"https://openalex.org/I136750679","display_name":"Garvan Institute of Medical Research","ror":"https://ror.org/01b3dvp57","country_code":"AU","type":"nonprofit","lineage":["https://openalex.org/I136750679","https://openalex.org/I4210087284"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Hasindu Gamaarachchi","raw_affiliation_strings":["School of Computer Science and Engineering, UNSW Sydney, Sydney, NSW, Australia","Kinghorn Centre for Clinical Genomics, Garvan Institute of Medical Research, Darlinghurst, NSW, Australia"],"affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, UNSW Sydney, Sydney, NSW, Australia","institution_ids":["https://openalex.org/I31746571"]},{"raw_affiliation_string":"Kinghorn Centre for Clinical Genomics, Garvan Institute of Medical Research, Darlinghurst, NSW, Australia","institution_ids":["https://openalex.org/I136750679"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5035736924","display_name":"Haris Javaid","orcid":"https://orcid.org/0009-0008-3472-0803"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Haris Javaid","raw_affiliation_strings":["Adaptive, Embedded and AI Group, AMD, Singapore, Singapore"],"affiliations":[{"raw_affiliation_string":"Adaptive, Embedded and AI Group, AMD, Singapore, Singapore","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100600905","display_name":"Xiaobo Sharon Hu","orcid":"https://orcid.org/0000-0002-6636-9738"},"institutions":[{"id":"https://openalex.org/I107639228","display_name":"University of Notre Dame","ror":"https://ror.org/00mkhxb43","country_code":"US","type":"education","lineage":["https://openalex.org/I107639228"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xiaobo Sharon Hu","raw_affiliation_strings":["Xiaobo Sharon Hu with the Department of Computer Science and Engineering, University of Notre Dame, Notre Dame, IN, USA"],"affiliations":[{"raw_affiliation_string":"Xiaobo Sharon Hu with the Department of Computer Science and Engineering, University of Notre Dame, Notre Dame, IN, USA","institution_ids":["https://openalex.org/I107639228"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5030042327","display_name":"Sri Parameswaran","orcid":"https://orcid.org/0000-0003-0435-9080"},"institutions":[{"id":"https://openalex.org/I129604602","display_name":"University of Sydney","ror":"https://ror.org/0384j8v12","country_code":"AU","type":"education","lineage":["https://openalex.org/I129604602"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Sri Parameswaran","raw_affiliation_strings":["School of Electrical and Information Engineering, University of Sydney, Sydney, NSW, Australia"],"affiliations":[{"raw_affiliation_string":"School of Electrical and Information Engineering, University of Sydney, Sydney, NSW, Australia","institution_ids":["https://openalex.org/I129604602"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5086185655"],"corresponding_institution_ids":["https://openalex.org/I31746571"],"apc_list":null,"apc_paid":null,"fwci":2.4563,"has_fulltext":false,"cited_by_count":20,"citation_normalized_percentile":{"value":0.90829324,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":89,"max":100},"biblio":{"volume":"42","issue":"11","first_page":"3505","last_page":"3518"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9994000196456909,"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":0.9994000196456909,"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/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.9975000023841858,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.996999979019165,"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":0.7324298024177551},{"id":"https://openalex.org/keywords/multiplier","display_name":"Multiplier (economics)","score":0.6979196071624756},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.579552412033081},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5776054263114929},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.5358755588531494},{"id":"https://openalex.org/keywords/lookup-table","display_name":"Lookup table","score":0.49845194816589355},{"id":"https://openalex.org/keywords/lagrange-multiplier","display_name":"Lagrange multiplier","score":0.4778834879398346},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.46329841017723083},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.46240073442459106},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.44419369101524353},{"id":"https://openalex.org/keywords/deep-neural-networks","display_name":"Deep neural networks","score":0.44084253907203674},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3515014946460724},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.32927483320236206},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3290562331676483},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.24307593703269958},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.15145516395568848}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7324298024177551},{"id":"https://openalex.org/C124584101","wikidata":"https://www.wikidata.org/wiki/Q1053266","display_name":"Multiplier (economics)","level":2,"score":0.6979196071624756},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.579552412033081},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5776054263114929},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.5358755588531494},{"id":"https://openalex.org/C134835016","wikidata":"https://www.wikidata.org/wiki/Q690265","display_name":"Lookup table","level":2,"score":0.49845194816589355},{"id":"https://openalex.org/C73684929","wikidata":"https://www.wikidata.org/wiki/Q598870","display_name":"Lagrange multiplier","level":2,"score":0.4778834879398346},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.46329841017723083},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.46240073442459106},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.44419369101524353},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.44084253907203674},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3515014946460724},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.32927483320236206},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3290562331676483},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.24307593703269958},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.15145516395568848},{"id":"https://openalex.org/C139719470","wikidata":"https://www.wikidata.org/wiki/Q39680","display_name":"Macroeconomics","level":1,"score":0.0},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C50522688","wikidata":"https://www.wikidata.org/wiki/Q189833","display_name":"Economic growth","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tcad.2023.3253045","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcad.2023.3253045","pdf_url":null,"source":{"id":"https://openalex.org/S100835903","display_name":"IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems","issn_l":"0278-0070","issn":["0278-0070","1937-4151"],"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 Computer-Aided Design of Integrated Circuits and Systems","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Decent work and economic growth","score":0.5099999904632568,"id":"https://metadata.un.org/sdg/8"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":37,"referenced_works":["https://openalex.org/W1530262073","https://openalex.org/W2007339694","https://openalex.org/W2037227137","https://openalex.org/W2048266589","https://openalex.org/W2108598243","https://openalex.org/W2111013824","https://openalex.org/W2112796928","https://openalex.org/W2119144962","https://openalex.org/W2194775991","https://openalex.org/W2265166184","https://openalex.org/W2763421725","https://openalex.org/W2790285612","https://openalex.org/W2798081680","https://openalex.org/W2809371234","https://openalex.org/W2897107698","https://openalex.org/W2900613774","https://openalex.org/W2964085268","https://openalex.org/W2997288154","https://openalex.org/W3007372856","https://openalex.org/W3036417437","https://openalex.org/W3036553373","https://openalex.org/W3046653923","https://openalex.org/W3093799822","https://openalex.org/W3113151582","https://openalex.org/W3118608800","https://openalex.org/W3128386784","https://openalex.org/W3176145847","https://openalex.org/W4287757601","https://openalex.org/W4302296459","https://openalex.org/W4385245566","https://openalex.org/W6631660994","https://openalex.org/W6677580257","https://openalex.org/W6739901393","https://openalex.org/W6745245109","https://openalex.org/W6783644704","https://openalex.org/W6785440099","https://openalex.org/W6964735071"],"related_works":["https://openalex.org/W3062287","https://openalex.org/W2380390332","https://openalex.org/W2742145873","https://openalex.org/W4245975140","https://openalex.org/W2062253548","https://openalex.org/W4225414539","https://openalex.org/W4289522463","https://openalex.org/W1977763331","https://openalex.org/W4318483369","https://openalex.org/W2002560966"],"abstract_inverted_index":{"Edge":[0],"training":[1,42,63,90,200,241],"of":[2,62,88,110,119,127,159,180,198],"deep":[3],"neural":[4,213],"networks":[5],"(DNNs)":[6],"is":[7,16,46,75,98,151,247],"a":[8,59,107,111,136,148],"desirable":[9],"goal":[10],"for":[11,67,204],"continuous":[12],"learning;":[13],"however,":[14,41],"it":[15],"hindered":[17],"by":[18,24,134],"the":[19,120,125,128,131,173,178,193,244,269],"enormous":[20],"computational":[21],"power":[22],"required":[23],"training.":[25],"Hardware":[26],"approximate":[27,44,54,73,95,121,139,160,183,202,237],"multipliers":[28,45,55,74,162,203],"have":[29],"shown":[30],"their":[31],"effectiveness":[32],"in":[33,37,185,225,240],"gaining":[34],"resource":[35],"efficiency":[36],"DNN":[38,57,69,89,112,199],"inference":[39,92],"accelerators;":[40],"with":[43,53,115,201,265],"largely":[47],"unexplored.":[48],"To":[49],"build":[50],"resource-efficient":[51],"accelerators":[52],"supporting":[56],"training,":[58],"thorough":[60],"evaluation":[61,87],"convergence":[64,194,220],"and":[65,71,91,104,168,195,212,222,231,242],"accuracy":[66,196,227],"different":[68,72],"architectures":[70],"needed.":[76],"This":[77],"article":[78],"presents":[79],"ApproxTrain,":[80],"an":[81],"open-source":[82],"framework":[83],"that":[84],"allows":[85],"fast":[86],"using":[93,135],"simulated":[94],"multipliers.":[96,233],"ApproxTrain":[97,165,190,246],"as":[99,101],"user-friendly":[100],"TensorFlow":[102,174,271],"(TF)":[103],"requires":[105],"only":[106,275],"high-level":[108],"description":[109],"architecture":[113],"along":[114],"C/C++":[116,156],"functional":[117,157],"models":[118,158],"multiplier.":[122],"We":[123,188],"improve":[124],"speed":[126],"simulation":[129],"at":[130],"multiplier":[132,142,184,238],"level":[133],"novel":[137,149],"LUT-based":[138],"floating-point":[140],"(FP)":[141],"simulator":[143],"on":[144,259,273],"GPU":[145],"(AMSim).":[146],"Additionally,":[147],"flow":[150],"presented":[152],"to":[153,176,191,229,235],"seamlessly":[154],"convert":[155],"FP":[161],"into":[163,172],"AMSim.":[164],"leverages":[166],"CUDA":[167],"efficiently":[169],"integrates":[170],"AMSim":[171],"library":[175],"overcome":[177],"absence":[179],"native":[181,266],"hardware":[182,267],"commercial":[186],"GPUs.":[187],"use":[189],"evaluate":[192],"performance":[197],"three":[205],"application":[206],"domains:":[207],"image":[208],"classification,":[209],"object":[210],"detection,":[211],"machine":[214],"translation.":[215],"The":[216],"evaluations":[217],"demonstrate":[218],"similar":[219],"behavior":[221],"negligible":[223],"change":[224],"test":[226],"compared":[228],"FP32":[230],"Bfloat16":[232],"Compared":[234],"CPU-based":[236],"simulations":[239],"inference,":[243],"GPU-accelerated":[245],"more":[248],"than":[249,284],"<inline-formula":[250,276],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[251,277],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">":[252,278],"<tex-math":[253,279],"notation=\"LaTeX\">$2500\\times":[254],"$":[255,281],"</tex-math></inline-formula>":[256,282],"faster.":[257],"Based":[258],"highly":[260],"optimized":[261],"closed-source":[262],"cuDNN/cuBLAS":[263],"libraries":[264],"multipliers,":[268],"original":[270],"is,":[272],"average,":[274],"notation=\"LaTeX\">$8\\times":[280],"faster":[283],"ApproxTrain.":[285]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":11},{"year":2024,"cited_by_count":6},{"year":2023,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}