{"id":"https://openalex.org/W2786618695","doi":"https://doi.org/10.1109/fpt.2017.8280150","title":"FPGA-based training of convolutional neural networks with a reduced precision floating-point library","display_name":"FPGA-based training of convolutional neural networks with a reduced precision floating-point library","publication_year":2017,"publication_date":"2017-12-01","ids":{"openalex":"https://openalex.org/W2786618695","doi":"https://doi.org/10.1109/fpt.2017.8280150","mag":"2786618695"},"language":"en","primary_location":{"id":"doi:10.1109/fpt.2017.8280150","is_oa":false,"landing_page_url":"https://doi.org/10.1109/fpt.2017.8280150","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2017 International Conference on Field Programmable Technology (ICFPT)","raw_type":"proceedings-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/A5015845004","display_name":"Roberto DiCecco","orcid":null},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":true,"raw_author_name":"Roberto DiCecco","raw_affiliation_strings":["Department of Electrical and Computer Engineering, University of Toronto, Ontario, Canada"],"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering, University of Toronto, Ontario, Canada","institution_ids":["https://openalex.org/I185261750"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5075221489","display_name":"Lin Sun","orcid":"https://orcid.org/0000-0002-2466-4632"},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Lin Sun","raw_affiliation_strings":["Department of Electrical and Computer Engineering, University of Toronto, Ontario, Canada"],"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering, University of Toronto, Ontario, Canada","institution_ids":["https://openalex.org/I185261750"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5035437657","display_name":"Paul Chow","orcid":"https://orcid.org/0000-0002-0523-7117"},"institutions":[{"id":"https://openalex.org/I185261750","display_name":"University of Toronto","ror":"https://ror.org/03dbr7087","country_code":"CA","type":"education","lineage":["https://openalex.org/I185261750"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Paul Chow","raw_affiliation_strings":["Department of Electrical and Computer Engineering, University of Toronto, Ontario, Canada"],"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering, University of Toronto, Ontario, Canada","institution_ids":["https://openalex.org/I185261750"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5015845004"],"corresponding_institution_ids":["https://openalex.org/I185261750"],"apc_list":null,"apc_paid":null,"fwci":0.7282,"has_fulltext":false,"cited_by_count":18,"citation_normalized_percentile":{"value":0.81376628,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"239","last_page":"242"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9919000267982483,"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":0.9919000267982483,"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/T10363","display_name":"Low-power high-performance VLSI design","score":0.9879999756813049,"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/T11206","display_name":"Model Reduction and Neural Networks","score":0.9846000075340271,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/mnist-database","display_name":"MNIST database","score":0.8345468044281006},{"id":"https://openalex.org/keywords/adder","display_name":"Adder","score":0.7999699115753174},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.7819315791130066},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7790865302085876},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.7626649737358093},{"id":"https://openalex.org/keywords/floating-point","display_name":"Floating point","score":0.6884604096412659},{"id":"https://openalex.org/keywords/double-precision-floating-point-format","display_name":"Double-precision floating-point format","score":0.6120578646659851},{"id":"https://openalex.org/keywords/lookup-table","display_name":"Lookup table","score":0.6021733283996582},{"id":"https://openalex.org/keywords/ieee-floating-point","display_name":"IEEE floating point","score":0.5731098651885986},{"id":"https://openalex.org/keywords/single-precision-floating-point-format","display_name":"Single-precision floating-point format","score":0.5714144110679626},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.46221938729286194},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.41327089071273804},{"id":"https://openalex.org/keywords/bandwidth","display_name":"Bandwidth (computing)","score":0.4101066589355469},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.4075523912906647},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.40411192178726196},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3379806876182556},{"id":"https://openalex.org/keywords/latency","display_name":"Latency (audio)","score":0.11330622434616089}],"concepts":[{"id":"https://openalex.org/C190502265","wikidata":"https://www.wikidata.org/wiki/Q17069496","display_name":"MNIST database","level":3,"score":0.8345468044281006},{"id":"https://openalex.org/C164620267","wikidata":"https://www.wikidata.org/wiki/Q376953","display_name":"Adder","level":3,"score":0.7999699115753174},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.7819315791130066},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7790865302085876},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.7626649737358093},{"id":"https://openalex.org/C84211073","wikidata":"https://www.wikidata.org/wiki/Q117879","display_name":"Floating point","level":2,"score":0.6884604096412659},{"id":"https://openalex.org/C35912277","wikidata":"https://www.wikidata.org/wiki/Q1243369","display_name":"Double-precision floating-point format","level":3,"score":0.6120578646659851},{"id":"https://openalex.org/C134835016","wikidata":"https://www.wikidata.org/wiki/Q690265","display_name":"Lookup table","level":2,"score":0.6021733283996582},{"id":"https://openalex.org/C137231763","wikidata":"https://www.wikidata.org/wiki/Q828287","display_name":"IEEE floating point","level":3,"score":0.5731098651885986},{"id":"https://openalex.org/C133095886","wikidata":"https://www.wikidata.org/wiki/Q1307173","display_name":"Single-precision floating-point format","level":3,"score":0.5714144110679626},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.46221938729286194},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41327089071273804},{"id":"https://openalex.org/C2776257435","wikidata":"https://www.wikidata.org/wiki/Q1576430","display_name":"Bandwidth (computing)","level":2,"score":0.4101066589355469},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.4075523912906647},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.40411192178726196},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3379806876182556},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.11330622434616089},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","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}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/fpt.2017.8280150","is_oa":false,"landing_page_url":"https://doi.org/10.1109/fpt.2017.8280150","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2017 International Conference on Field Programmable Technology (ICFPT)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":23,"referenced_works":["https://openalex.org/W1667652561","https://openalex.org/W1841592590","https://openalex.org/W1983157164","https://openalex.org/W2097117768","https://openalex.org/W2112796928","https://openalex.org/W2114508814","https://openalex.org/W2163605009","https://openalex.org/W2483618896","https://openalex.org/W2520083297","https://openalex.org/W2557283755","https://openalex.org/W2574797063","https://openalex.org/W2962987932","https://openalex.org/W2963374099","https://openalex.org/W3118608800","https://openalex.org/W4302296459","https://openalex.org/W4387007199","https://openalex.org/W6637151318","https://openalex.org/W6638783484","https://openalex.org/W6674914833","https://openalex.org/W6677192272","https://openalex.org/W6684191040","https://openalex.org/W6787972765","https://openalex.org/W6856891780"],"related_works":["https://openalex.org/W1564887326","https://openalex.org/W3215589575","https://openalex.org/W2000092506","https://openalex.org/W2596474508","https://openalex.org/W2587538765","https://openalex.org/W2169016399","https://openalex.org/W2730658480","https://openalex.org/W2116803521","https://openalex.org/W1573821047","https://openalex.org/W2565957982"],"abstract_inverted_index":{"Convolutional":[0],"Neural":[1],"Networks":[2],"(CNNs)":[3],"have":[4,8],"been":[5],"shown":[6],"to":[7,95,98,114,155],"high":[9,26,35],"accuracy":[10,27,112],"for":[11,40,75,84,117,129,135,140,145,150],"classification":[12,42],"tasks":[13],"in":[14,20],"numerous":[15],"applications,":[16],"which":[17,82],"has":[18],"resulted":[19],"their":[21],"widespread":[22],"adoption.":[23],"However,":[24],"the":[25,32,62,118,130,136,146,151,160],"of":[28,34,104,109,143,163],"CNNs":[29],"comes":[30],"at":[31],"cost":[33],"compute":[36],"and":[37,43,77,106,120,133,148],"bandwidth":[38],"requirements":[39],"both":[41],"training.":[44],"In":[45],"this":[46],"work":[47],"we":[48,69],"discuss":[49],"an":[50,101],"FPGA-based":[51],"CNN":[52],"training":[53],"engine:":[54],"FCTE,":[55],"implemented":[56,79],"using":[57,80,127,159],"High-Level":[58],"Synthesis":[59],"(HLS),":[60],"targeting":[61],"Xilinx":[63],"Kintex":[64],"Ultrascale":[65],"XCKU115":[66],"device.":[67],"Furthermore,":[68],"detail":[70],"custom-precision":[71],"floating-point":[72,116],"(CPFP)":[73],"cores":[74,91],"multiplication":[76],"addition":[78],"HLS,":[81],"allows":[83],"reduced":[85],"area":[86],"utilization.":[87],"We":[88],"use":[89],"these":[90],"with":[92],"our":[93],"engine":[94],"train":[96],"networks":[97],"demonstrate":[99],"that":[100],"exponent":[102],"width":[103,108],"6":[105],"mantissa":[107],"5":[110],"achieves":[111],"comparable":[113],"single-precision":[115],"MNIST":[119],"CIFAR-10":[121],"datasets.":[122],"These":[123],"results":[124],"are":[125],"achieved":[126],"round-to-zero":[128],"CPFP":[131,137],"multipliers":[132,147],"round-to-nearest":[134],"adders,":[138],"allowing":[139],"LUT":[141],"savings":[142],"32.6%":[144],"21.7%":[149],"adders":[152],"when":[153],"compared":[154],"half-precision":[156],"floating-point,":[157],"while":[158],"same":[161],"number":[162],"DSPs.":[164]},"counts_by_year":[{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":1},{"year":2022,"cited_by_count":3},{"year":2021,"cited_by_count":5},{"year":2020,"cited_by_count":1},{"year":2019,"cited_by_count":5},{"year":2018,"cited_by_count":2}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}