{"id":"https://openalex.org/W3198405160","doi":"https://doi.org/10.1109/tcad.2021.3109857","title":"NEAT: Nonlinearity Aware Training for Accurate, Energy-Efficient, and Robust Implementation of Neural Networks on 1T-1R Crossbars","display_name":"NEAT: Nonlinearity Aware Training for Accurate, Energy-Efficient, and Robust Implementation of Neural Networks on 1T-1R Crossbars","publication_year":2021,"publication_date":"2021-09-03","ids":{"openalex":"https://openalex.org/W3198405160","doi":"https://doi.org/10.1109/tcad.2021.3109857","mag":"3198405160"},"language":"en","primary_location":{"id":"doi:10.1109/tcad.2021.3109857","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcad.2021.3109857","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/A5004629816","display_name":"Abhiroop Bhattacharjee","orcid":"https://orcid.org/0000-0002-7721-271X"},"institutions":[{"id":"https://openalex.org/I32971472","display_name":"Yale University","ror":"https://ror.org/03v76x132","country_code":"US","type":"education","lineage":["https://openalex.org/I32971472"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Abhiroop Bhattacharjee","raw_affiliation_strings":["Department of Electrical Engineering, Yale University, New Haven, CT, USA"],"raw_orcid":"https://orcid.org/0000-0002-7721-271X","affiliations":[{"raw_affiliation_string":"Department of Electrical Engineering, Yale University, New Haven, CT, USA","institution_ids":["https://openalex.org/I32971472"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5015699875","display_name":"Lakshya Bhatnagar","orcid":"https://orcid.org/0000-0002-4152-1916"},"institutions":[{"id":"https://openalex.org/I32971472","display_name":"Yale University","ror":"https://ror.org/03v76x132","country_code":"US","type":"education","lineage":["https://openalex.org/I32971472"]},{"id":"https://openalex.org/I68891433","display_name":"Indian Institute of Technology Delhi","ror":"https://ror.org/049tgcd06","country_code":"IN","type":"education","lineage":["https://openalex.org/I68891433"]}],"countries":["IN","US"],"is_corresponding":false,"raw_author_name":"Lakshya Bhatnagar","raw_affiliation_strings":["Department of Electrical Engineering, Yale University, New Haven, CT, USA","Department of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, India"],"raw_orcid":"https://orcid.org/0000-0001-6083-9017","affiliations":[{"raw_affiliation_string":"Department of Electrical Engineering, Yale University, New Haven, CT, USA","institution_ids":["https://openalex.org/I32971472"]},{"raw_affiliation_string":"Department of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, India","institution_ids":["https://openalex.org/I68891433"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086829943","display_name":"Youngeun Kim","orcid":"https://orcid.org/0000-0001-9905-9528"},"institutions":[{"id":"https://openalex.org/I32971472","display_name":"Yale University","ror":"https://ror.org/03v76x132","country_code":"US","type":"education","lineage":["https://openalex.org/I32971472"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Youngeun Kim","raw_affiliation_strings":["Department of Electrical Engineering, Yale University, New Haven, CT, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Electrical Engineering, Yale University, New Haven, CT, USA","institution_ids":["https://openalex.org/I32971472"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5050310538","display_name":"Priyadarshini Panda","orcid":"https://orcid.org/0000-0002-4167-6782"},"institutions":[{"id":"https://openalex.org/I32971472","display_name":"Yale University","ror":"https://ror.org/03v76x132","country_code":"US","type":"education","lineage":["https://openalex.org/I32971472"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Priyadarshini Panda","raw_affiliation_strings":["Department of Electrical Engineering, Yale University, New Haven, CT, USA"],"raw_orcid":"https://orcid.org/0000-0002-4167-6782","affiliations":[{"raw_affiliation_string":"Department of Electrical Engineering, Yale University, New Haven, CT, USA","institution_ids":["https://openalex.org/I32971472"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":1.9297,"has_fulltext":false,"cited_by_count":29,"citation_normalized_percentile":{"value":0.85988736,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":98,"max":99},"biblio":{"volume":"41","issue":"8","first_page":"2625","last_page":"2637"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":1.0,"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":1.0,"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.9997000098228455,"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/T10472","display_name":"Semiconductor materials and devices","score":0.9987999796867371,"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/crossbar-switch","display_name":"Crossbar switch","score":0.8026245832443237},{"id":"https://openalex.org/keywords/nonlinear-system","display_name":"Nonlinear system","score":0.6723686456680298},{"id":"https://openalex.org/keywords/parasitic-extraction","display_name":"Parasitic extraction","score":0.6479929685592651},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6261746287345886},{"id":"https://openalex.org/keywords/neuromorphic-engineering","display_name":"Neuromorphic engineering","score":0.561286211013794},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5517044067382812},{"id":"https://openalex.org/keywords/transistor","display_name":"Transistor","score":0.5073348879814148},{"id":"https://openalex.org/keywords/interconnection","display_name":"Interconnection","score":0.502166748046875},{"id":"https://openalex.org/keywords/topology","display_name":"Topology (electrical circuits)","score":0.493854820728302},{"id":"https://openalex.org/keywords/relay","display_name":"Relay","score":0.4311813712120056},{"id":"https://openalex.org/keywords/energy","display_name":"Energy (signal processing)","score":0.41868507862091064},{"id":"https://openalex.org/keywords/spice","display_name":"Spice","score":0.4165268540382385},{"id":"https://openalex.org/keywords/electronic-engineering","display_name":"Electronic engineering","score":0.380770206451416},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.3381667733192444},{"id":"https://openalex.org/keywords/voltage","display_name":"Voltage","score":0.30082905292510986},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2522766590118408},{"id":"https://openalex.org/keywords/electrical-engineering","display_name":"Electrical engineering","score":0.22931638360023499},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.14543399214744568},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.14504826068878174},{"id":"https://openalex.org/keywords/physics","display_name":"Physics","score":0.14117693901062012},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.104633629322052}],"concepts":[{"id":"https://openalex.org/C29984679","wikidata":"https://www.wikidata.org/wiki/Q1929149","display_name":"Crossbar switch","level":2,"score":0.8026245832443237},{"id":"https://openalex.org/C158622935","wikidata":"https://www.wikidata.org/wiki/Q660848","display_name":"Nonlinear system","level":2,"score":0.6723686456680298},{"id":"https://openalex.org/C159818811","wikidata":"https://www.wikidata.org/wiki/Q7135947","display_name":"Parasitic extraction","level":2,"score":0.6479929685592651},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6261746287345886},{"id":"https://openalex.org/C151927369","wikidata":"https://www.wikidata.org/wiki/Q1981312","display_name":"Neuromorphic engineering","level":3,"score":0.561286211013794},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5517044067382812},{"id":"https://openalex.org/C172385210","wikidata":"https://www.wikidata.org/wiki/Q5339","display_name":"Transistor","level":3,"score":0.5073348879814148},{"id":"https://openalex.org/C123745756","wikidata":"https://www.wikidata.org/wiki/Q1665949","display_name":"Interconnection","level":2,"score":0.502166748046875},{"id":"https://openalex.org/C184720557","wikidata":"https://www.wikidata.org/wiki/Q7825049","display_name":"Topology (electrical circuits)","level":2,"score":0.493854820728302},{"id":"https://openalex.org/C2778156585","wikidata":"https://www.wikidata.org/wiki/Q174053","display_name":"Relay","level":3,"score":0.4311813712120056},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.41868507862091064},{"id":"https://openalex.org/C2780077345","wikidata":"https://www.wikidata.org/wiki/Q16891888","display_name":"Spice","level":2,"score":0.4165268540382385},{"id":"https://openalex.org/C24326235","wikidata":"https://www.wikidata.org/wiki/Q126095","display_name":"Electronic engineering","level":1,"score":0.380770206451416},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3381667733192444},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.30082905292510986},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2522766590118408},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.22931638360023499},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.14543399214744568},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.14504826068878174},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.14117693901062012},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.104633629322052},{"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/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.0},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tcad.2021.3109857","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tcad.2021.3109857","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":[{"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy","score":0.9100000262260437}],"awards":[{"id":"https://openalex.org/G3974418817","display_name":"CRII: SHF: Efficiency-Aware Robust Implementation of Neural Networks with Algorithm-Hardware Co-design","funder_award_id":"1947826","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":41,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W1945616565","https://openalex.org/W2056507634","https://openalex.org/W2064814573","https://openalex.org/W2194775991","https://openalex.org/W2591675147","https://openalex.org/W2603766943","https://openalex.org/W2612375349","https://openalex.org/W2624863671","https://openalex.org/W2766179883","https://openalex.org/W2796625795","https://openalex.org/W2798170643","https://openalex.org/W2913104037","https://openalex.org/W2913848079","https://openalex.org/W2946047477","https://openalex.org/W2951055820","https://openalex.org/W2963389226","https://openalex.org/W2966199719","https://openalex.org/W3017820626","https://openalex.org/W3020429010","https://openalex.org/W3033519076","https://openalex.org/W3080490461","https://openalex.org/W3091835145","https://openalex.org/W3092585568","https://openalex.org/W3104263540","https://openalex.org/W3118608800","https://openalex.org/W3183458034","https://openalex.org/W4236709213","https://openalex.org/W4245731639","https://openalex.org/W4293846201","https://openalex.org/W4311409824","https://openalex.org/W4383466480","https://openalex.org/W6637373629","https://openalex.org/W6640425456","https://openalex.org/W6729756640","https://openalex.org/W6737043114","https://openalex.org/W6739868092","https://openalex.org/W6758975236","https://openalex.org/W6776346441","https://openalex.org/W6781967490","https://openalex.org/W6787972765"],"related_works":["https://openalex.org/W2986579802","https://openalex.org/W4385753159","https://openalex.org/W4387251107","https://openalex.org/W2620105511","https://openalex.org/W2612269878","https://openalex.org/W2991631906","https://openalex.org/W1998175862","https://openalex.org/W4297664933","https://openalex.org/W2068476500","https://openalex.org/W2125822386"],"abstract_inverted_index":{"In":[0,146,361],"this":[1,147,266],"era":[2],"of":[3,10,28,97,104,139,186,202,211,258,303,347,363],"IoT,":[4],"energy-efficient":[5,26],"and":[6,37,54,164,167,251,295,392],"adversarially":[7,344],"secure":[8],"implementation":[9],"deep":[11,29],"neural":[12,212],"networks":[13,213,324],"(DNNs)":[14],"on":[15,265,293,325,367,401],"hardware":[16,31],"has":[17,80,246],"become":[18],"imperative.":[19],"Memristive":[20],"crossbars":[21,157,333,352],"have":[22],"emerged":[23],"as":[24,161],"an":[25,223],"component":[27],"learning":[30],"accelerators":[32],"due":[33],"to":[34,83,128,143,176,214,260,281,299,342,408],"their":[35,59],"compact":[36],"efficient":[38],"matrix-vector":[39],"multiplication":[40],"(MVM)":[41],"implementation.":[42],"However,":[43,90],"they":[44],"suffer":[45],"from":[46,119],"nonidealities":[47],"(such":[48],"as,":[49],"interconnect":[50,165],"parasitics,":[51],"device":[52,162],"variations,":[53],"sneak":[55,85],"paths)":[56],"introduced":[57],"by":[58,221,237,276],"circuit":[60],"topology":[61],"that":[62,93,243,339],"degrades":[63],"computational":[64],"accuracy.":[65],"A":[66],"1T-1R":[67,106,116,156,195,283,326,332,351],"synapse,":[68],"adding":[69],"a":[70,88,135,169,247,364],"transistor":[71,99,259],"(1T)":[72],"in":[73,87,109,154,252,374,382,406],"series":[74],"with":[75,311,334],"the":[76,94,98,101,105,112,120,152,178,184,194,198,203,209,217,232,238,282,301,331],"memristive":[77],"synapse":[78],"(1R),":[79],"been":[81],"proposed":[82],"mitigate":[84],"paths":[86],"crossbar.":[89],"we":[91,149,181,207,241,268,289,329,370],"observe":[92],"nonlinear":[95],"characteristics":[96],"affect":[100],"overall":[102],"conductance":[103],"cell":[107,196],"which":[108,189],"turn":[110,253],"affects":[111],"MVM":[113],"operation.":[114,263],"This":[115],"nonlinearity":[117,123,153,171],"arising":[118],"input":[121],"voltage-dependent":[122],"is":[124],"not":[125],"only":[126],"difficult":[127],"model":[129],"or":[130],"formulate,":[131],"but":[132],"also":[133],"causes":[134],"drastic":[136],"performance":[137],"degradation":[138],"DNNs":[140,348,355],"when":[141,321],"mapped":[142,192,356],"such":[144,160],"crossbars.":[145,327,360],"article,":[148],"first":[150,182],"analyses":[151],"ideal":[155],"(excluding":[158],"nonidealities,":[159],"variations":[163],"parasitics)":[166],"propose":[168],"novel":[170],"aware":[172],"training":[173,225,229],"(NEAT)":[174],"method":[175,390],"address":[177],"nonlinearities.":[179],"Specifically,":[180],"identify":[183],"range":[185,220],"network":[187,366],"weights,":[188],"can":[190],"be":[191],"into":[193],"within":[197,216],"linear":[199,218,262],"operating":[200,219],"region":[201],"transistor.":[204],"After":[205],"that,":[206],"regularize":[208],"weights":[210],"exist":[215],"using":[222],"iterative":[224,228],"algorithm.":[226],"Our":[227],"significantly":[230],"recovers":[231],"classification":[233,274],"accuracy":[234,275,315,376],"drop":[235],"caused":[236],"nonlinearity.":[239],"Moreover,":[240],"find":[242],"each":[244],"layer":[245],"different":[248,255,286],"weight":[249],"distribution":[250],"requires":[254],"gate":[256,319],"voltage":[257],"guarantee":[261],"Based":[264],"observation,":[267],"achieve":[269],"energy":[270],"efficiency":[271],"while":[272],"preserving":[273],"applying":[277],"heterogeneous":[278],"gate-voltage":[279],"control":[280],"cells":[284],"across":[285],"layers.":[287],"Finally,":[288],"conduct":[290],"various":[291,335],"experiments":[292],"CIFAR10":[294],"CIFAR100":[296,368],"benchmark":[297],"datasets":[298],"demonstrate":[300],"effectiveness":[302],"our":[304],"NEAT.":[305],"Overall,":[306],"NEAT":[307,340,400],"yields$\\sim":[308],"20\\%$energy":[309],"gain":[310],"less":[312],"than":[313,353],"1%":[314],"loss":[316],"(with":[317],"homogeneous":[318],"control)":[320],"mapping":[322],"ResNet18":[323],"Thereafter,":[328],"integrate":[330],"nonidealities.":[336],"We":[337],"show":[338],"leads":[341],"more":[343],"robust":[345],"mappings":[346],"onto":[349,358],"nonideal":[350,402],"standard":[354,409],"directly":[357],"1R":[359],"case":[362],"VGG11":[365],"dataset,":[369],"obtain$\\sim":[371],"17$%":[372],"improvement":[373],"clean":[375],"and$\\sim":[377],"2$%\u20138%":[378],"&$\\sim":[379],"5$%\u20136%":[380],"improvements":[381],"adversarial":[383,397],"accuracies,":[384],"respectively,":[385],"for":[386],"fast":[387],"gradient":[388,394],"sign":[389],"(FGSM)":[391],"projected":[393],"descent":[395],"(PGD)-based":[396],"attacks":[398],"via":[399],"64$\\times":[403],"$64":[404],"crossbars,":[405],"comparison":[407],"DNNs.":[410]},"counts_by_year":[{"year":2025,"cited_by_count":5},{"year":2024,"cited_by_count":8},{"year":2023,"cited_by_count":7},{"year":2022,"cited_by_count":9}],"updated_date":"2026-07-29T09:40:50.615796","created_date":"2025-10-10T00:00:00"}