{"id":"https://openalex.org/W2917710201","doi":"https://doi.org/10.1145/3289602.3293964","title":"A Fine-Grained Sparse Accelerator for Multi-Precision DNN","display_name":"A Fine-Grained Sparse Accelerator for Multi-Precision DNN","publication_year":2019,"publication_date":"2019-02-20","ids":{"openalex":"https://openalex.org/W2917710201","doi":"https://doi.org/10.1145/3289602.3293964","mag":"2917710201"},"language":"en","primary_location":{"id":"doi:10.1145/3289602.3293964","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3289602.3293964","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2019 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays","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/A5026831784","display_name":"Shulin Zeng","orcid":"https://orcid.org/0000-0002-1030-3748"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Shulin Zeng","raw_affiliation_strings":["Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021398655","display_name":"Yujun Lin","orcid":"https://orcid.org/0000-0001-6314-1722"},"institutions":[{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yujun Lin","raw_affiliation_strings":["Massachusetts Institute of Technology, Cambridge, MA, USA"],"affiliations":[{"raw_affiliation_string":"Massachusetts Institute of Technology, Cambridge, MA, USA","institution_ids":["https://openalex.org/I63966007"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101870874","display_name":"Shuang Liang","orcid":"https://orcid.org/0000-0003-0881-5549"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shuang Liang","raw_affiliation_strings":["Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017382359","display_name":"Junlong Kang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Junlong Kang","raw_affiliation_strings":["Xilinx, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Xilinx, Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100636145","display_name":"Dongliang Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dongliang Xie","raw_affiliation_strings":["Xilinx, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Xilinx, Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5009175141","display_name":"Yi Shan","orcid":"https://orcid.org/0000-0003-2646-8835"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yi Shan","raw_affiliation_strings":["Xilinx, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Xilinx, Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070926896","display_name":"Song Han","orcid":"https://orcid.org/0000-0002-4186-7618"},"institutions":[{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Song Han","raw_affiliation_strings":["Massachusetts Institute of Technology, Cambridge, MA, USA"],"affiliations":[{"raw_affiliation_string":"Massachusetts Institute of Technology, Cambridge, MA, USA","institution_ids":["https://openalex.org/I63966007"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017851308","display_name":"Yu Wang","orcid":"https://orcid.org/0000-0001-7763-4261"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yu Wang","raw_affiliation_strings":["Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5023755254","display_name":"Huazhong Yang","orcid":"https://orcid.org/0000-0003-2421-353X"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Huazhong Yang","raw_affiliation_strings":["Tsinghua University, Beijing, China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing, China","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5026831784"],"corresponding_institution_ids":["https://openalex.org/I99065089"],"apc_list":null,"apc_paid":null,"fwci":0.5601,"has_fulltext":false,"cited_by_count":10,"citation_normalized_percentile":{"value":0.7374908,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"185","last_page":"185"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","score":0.9485999941825867,"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"}},"topics":[{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","score":0.9485999941825867,"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/T10320","display_name":"Neural Networks and Applications","score":0.9277999997138977,"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/T14444","display_name":"Power Systems and Renewable Energy","score":0.9035000205039978,"subfield":{"id":"https://openalex.org/subfields/2102","display_name":"Energy Engineering and Power Technology"},"field":{"id":"https://openalex.org/fields/21","display_name":"Energy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8125641345977783},{"id":"https://openalex.org/keywords/datapath","display_name":"Datapath","score":0.6075942516326904},{"id":"https://openalex.org/keywords/hardware-acceleration","display_name":"Hardware acceleration","score":0.551096498966217},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.5458460450172424},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.534206748008728},{"id":"https://openalex.org/keywords/multiplication","display_name":"Multiplication (music)","score":0.5243383049964905},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.5113139152526855},{"id":"https://openalex.org/keywords/recurrent-neural-network","display_name":"Recurrent neural network","score":0.5004239082336426},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.497835636138916},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4922451078891754},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.43935713171958923},{"id":"https://openalex.org/keywords/network-architecture","display_name":"Network architecture","score":0.4314202070236206},{"id":"https://openalex.org/keywords/sparse-matrix","display_name":"Sparse matrix","score":0.4257136583328247},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.36494532227516174},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.34161511063575745},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.33923929929733276},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.2611130475997925}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8125641345977783},{"id":"https://openalex.org/C2781198647","wikidata":"https://www.wikidata.org/wiki/Q1633673","display_name":"Datapath","level":2,"score":0.6075942516326904},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.551096498966217},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.5458460450172424},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.534206748008728},{"id":"https://openalex.org/C2780595030","wikidata":"https://www.wikidata.org/wiki/Q3860309","display_name":"Multiplication (music)","level":2,"score":0.5243383049964905},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.5113139152526855},{"id":"https://openalex.org/C147168706","wikidata":"https://www.wikidata.org/wiki/Q1457734","display_name":"Recurrent neural network","level":3,"score":0.5004239082336426},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.497835636138916},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4922451078891754},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.43935713171958923},{"id":"https://openalex.org/C193415008","wikidata":"https://www.wikidata.org/wiki/Q639681","display_name":"Network architecture","level":2,"score":0.4314202070236206},{"id":"https://openalex.org/C56372850","wikidata":"https://www.wikidata.org/wiki/Q1050404","display_name":"Sparse matrix","level":3,"score":0.4257136583328247},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.36494532227516174},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.34161511063575745},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33923929929733276},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2611130475997925},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.0},{"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/C24890656","wikidata":"https://www.wikidata.org/wiki/Q82811","display_name":"Acoustics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3289602.3293964","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3289602.3293964","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2019 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.8999999761581421,"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W2109699519","https://openalex.org/W2006568360","https://openalex.org/W102726818","https://openalex.org/W4233616027","https://openalex.org/W2059591361","https://openalex.org/W970262775","https://openalex.org/W4244724753","https://openalex.org/W2535673728","https://openalex.org/W1972081536","https://openalex.org/W2537803004"],"abstract_inverted_index":{"Neural":[0],"Networks":[1],"(NNs)":[2],"have":[3,166],"made":[4],"a":[5,15,155,167],"significant":[6],"breakthrough":[7],"in":[8,171],"many":[9],"fields,":[10],"while":[11],"they":[12],"also":[13],"pose":[14],"great":[16],"challenge":[17],"to":[18,46,78,139,193],"hardware":[19,43,82,178],"platforms":[20],"since":[21],"the":[22,51,58,62,69,96,107,120,123,128,141,147,163,172,176,186,220,227],"state-of-the-art":[23,221],"neural":[24,74,85,133,143],"networks":[25,75],"are":[26],"both":[27],"communicational-":[28],"and":[29,38,81,103,116,131,189,200,236],"computational-intensive.":[30],"Researchers":[31],"proposed":[32,164,181],"model":[33],"compression":[34],"algorithms":[35],"using":[36,146],"sparsification":[37],"quantization,":[39],"along":[40],"with":[41,158,175,232],"specific":[42],"architecture":[44,70,165,183],"designs,":[45],"accelerate":[47],"various":[48],"applications.":[49,87],"However,":[50],"irregularity":[52],"of":[53,64,99,109,122],"memory":[54],"access":[55],"caused":[56],"by":[57],"sparsity":[59],"severely":[60],"damages":[61],"regularity":[63],"intensive":[65],"computation":[66,97],"loops.":[67],"Therefore,":[68],"design":[71],"for":[72,84,198,205],"sparse":[73,110,113],"is":[76,219],"crucial":[77],"better":[79,168],"software":[80],"co-design":[83],"network":[86,130,134,144,231],"To":[88],"face":[89],"these":[90],"challenges,":[91],"this":[92],"paper":[93,153],"first":[94],"analyzes":[95],"patterns":[98],"different":[100],"NN":[101],"structures":[102],"unify":[104],"them":[105],"into":[106],"form":[108],"matrix-vector":[111],"multiplication,":[112,115],"matrix-matrix":[114],"element-wise":[117],"multiplication.":[118],"On":[119],"basis":[121],"EIE":[124],"which":[125,161],"supports":[126],"only":[127],"fully-connected":[129],"recurrent":[132,229],"(RNN),":[135],"we":[136],"expand":[137],"it":[138,218],"support":[140],"convolution":[142],"(CNN)":[145],"input":[148],"vector":[149],"transform":[150],"unit.":[151],"This":[152],"designs":[154],"multi-precision":[156],"multiplier":[157],"supporting":[159,223],"datapath,":[160],"makes":[162],"acceleration":[169],"effect":[170],"low-bit":[173],"quantization":[174,209],"same":[177],"architecture.":[179],"The":[180],"accelerator":[182,222],"can":[184],"achieve":[185],"equivalent":[187],"performance":[188,235],"energy":[190,239],"efficiency":[191,240],"up":[192],"574.2":[194],"GOPS,":[195,202],"42.8":[196],"GOPS/W":[197,204,238],"CNN":[199],"110.4":[201],"8.24":[203],"RNN":[206],"under":[207,241],"4-bit":[208,242],"on":[210],"Xilinx":[211],"XCKU115":[212],"FPGA":[213],"running":[214],"at":[215],"200MHz.":[216],"And":[217],"CNN-RNN-based":[224],"models":[225],"like":[226],"long-term":[228],"convolutional":[230],"571.1":[233],"GOPS":[234],"42.6":[237],"data":[243],"format.":[244]},"counts_by_year":[{"year":2025,"cited_by_count":6},{"year":2022,"cited_by_count":1},{"year":2021,"cited_by_count":1},{"year":2020,"cited_by_count":2}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}