{"id":"https://openalex.org/W4200413266","doi":"https://doi.org/10.1109/asicon52560.2021.9620448","title":"Exploiting Dynamic Bit Sparsity in Activation for Deep Neural Network Acceleration","display_name":"Exploiting Dynamic Bit Sparsity in Activation for Deep Neural Network Acceleration","publication_year":2021,"publication_date":"2021-10-26","ids":{"openalex":"https://openalex.org/W4200413266","doi":"https://doi.org/10.1109/asicon52560.2021.9620448"},"language":"en","primary_location":{"id":"doi:10.1109/asicon52560.2021.9620448","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asicon52560.2021.9620448","pdf_url":null,"source":{"id":"https://openalex.org/S4363607945","display_name":"2021 IEEE 14th International Conference on ASIC (ASICON)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 IEEE 14th International Conference on ASIC (ASICON)","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/A5080944993","display_name":"Yongshuai Sun","orcid":null},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Yongshuai Sun","raw_affiliation_strings":["Shanghai Jiao Tong University, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030001296","display_name":"Mengyu Guo","orcid":"https://orcid.org/0000-0002-9953-0073"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Mengyu Guo","raw_affiliation_strings":["Shanghai Jiao Tong University, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081047186","display_name":"Dacheng Liang","orcid":"https://orcid.org/0000-0002-8898-3771"},"institutions":[{"id":"https://openalex.org/I4210153410","display_name":"China Nerin Engineering (China)","ror":"https://ror.org/04th5gf24","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153410"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dacheng Liang","raw_affiliation_strings":["Biren Research, China"],"affiliations":[{"raw_affiliation_string":"Biren Research, China","institution_ids":["https://openalex.org/I4210153410"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062976895","display_name":"Shan Tang","orcid":"https://orcid.org/0000-0003-4483-1108"},"institutions":[{"id":"https://openalex.org/I4210153410","display_name":"China Nerin Engineering (China)","ror":"https://ror.org/04th5gf24","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210153410"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shan Tang","raw_affiliation_strings":["Biren Research, China"],"affiliations":[{"raw_affiliation_string":"Biren Research, China","institution_ids":["https://openalex.org/I4210153410"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5045693138","display_name":"Naifeng Jing","orcid":"https://orcid.org/0000-0001-8417-5796"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Naifeng Jing","raw_affiliation_strings":["Shanghai Jiao Tong University, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University, China","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5080944993"],"corresponding_institution_ids":["https://openalex.org/I183067930"],"apc_list":null,"apc_paid":null,"fwci":0.0657,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.35342976,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"4"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9984999895095825,"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":0.9984999895095825,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9948999881744385,"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/T12303","display_name":"Tensor decomposition and applications","score":0.9936000108718872,"subfield":{"id":"https://openalex.org/subfields/2605","display_name":"Computational Mathematics"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.8695812821388245},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8393067717552185},{"id":"https://openalex.org/keywords/convolutional-neural-network","display_name":"Convolutional neural network","score":0.6535859107971191},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.6361970901489258},{"id":"https://openalex.org/keywords/synchronization","display_name":"Synchronization (alternating current)","score":0.570286214351654},{"id":"https://openalex.org/keywords/acceleration","display_name":"Acceleration","score":0.5691986680030823},{"id":"https://openalex.org/keywords/hardware-acceleration","display_name":"Hardware acceleration","score":0.5248382687568665},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.49772194027900696},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4936691224575043},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4386706054210663},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.433843731880188},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.41961801052093506},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2425146996974945},{"id":"https://openalex.org/keywords/channel","display_name":"Channel (broadcasting)","score":0.12816768884658813},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.10713541507720947}],"concepts":[{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.8695812821388245},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8393067717552185},{"id":"https://openalex.org/C81363708","wikidata":"https://www.wikidata.org/wiki/Q17084460","display_name":"Convolutional neural network","level":2,"score":0.6535859107971191},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.6361970901489258},{"id":"https://openalex.org/C2778562939","wikidata":"https://www.wikidata.org/wiki/Q1298791","display_name":"Synchronization (alternating current)","level":3,"score":0.570286214351654},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.5691986680030823},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.5248382687568665},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.49772194027900696},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4936691224575043},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4386706054210663},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.433843731880188},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.41961801052093506},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2425146996974945},{"id":"https://openalex.org/C127162648","wikidata":"https://www.wikidata.org/wiki/Q16858953","display_name":"Channel (broadcasting)","level":2,"score":0.12816768884658813},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.10713541507720947},{"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/C74650414","wikidata":"https://www.wikidata.org/wiki/Q11397","display_name":"Classical mechanics","level":1,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/asicon52560.2021.9620448","is_oa":false,"landing_page_url":"https://doi.org/10.1109/asicon52560.2021.9620448","pdf_url":null,"source":{"id":"https://openalex.org/S4363607945","display_name":"2021 IEEE 14th International Conference on ASIC (ASICON)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 IEEE 14th International Conference on ASIC (ASICON)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":13,"referenced_works":["https://openalex.org/W2048266589","https://openalex.org/W2156387975","https://openalex.org/W2560217098","https://openalex.org/W2612076670","https://openalex.org/W2804032941","https://openalex.org/W2931118404","https://openalex.org/W2949870694","https://openalex.org/W3036939803","https://openalex.org/W3043504674","https://openalex.org/W6662587704","https://openalex.org/W6682889407","https://openalex.org/W6730470330","https://openalex.org/W6751349269"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W98480971","https://openalex.org/W2150291671","https://openalex.org/W2027972911","https://openalex.org/W2146343568","https://openalex.org/W2218038495","https://openalex.org/W2518118925","https://openalex.org/W2532502681","https://openalex.org/W3159273459"],"abstract_inverted_index":{"Data":[0],"sparsity":[1,17,40],"is":[2,110],"important":[3],"in":[4,19,24,46,84,94,104,131,159],"accelerating":[5],"deep":[6],"neural":[7],"networks":[8],"(DNNs).":[9],"However,":[10],"besides":[11],"the":[12,15,38,54,65,70,75,81,92,95,118,151,176],"zeroed":[13,44,82],"values,":[14],"bit":[16,39],"especially":[18],"activations":[20,93],"are":[21],"oftentimes":[22],"missing":[23],"conventional":[25],"DNN":[26,34],"accelerators.":[27],"In":[28,173],"this":[29,49],"paper,":[30],"we":[31,51,87],"present":[32],"a":[33],"accelerator":[35,120],"to":[36,63],"exploit":[37],"by":[41,74],"dynamically":[42],"skipping":[43],"bits":[45,83,103],"activations.":[47],"To":[48,68],"goal,":[50],"first":[52],"substitute":[53],"multiply-and-accumulate":[55],"(MAC)":[56],"units":[57,62,124],"with":[58,121,146,156,170],"more":[59],"serial":[60],"shift-and-accumulate":[61],"sustain":[64],"computing":[66],"parallelism.":[67],"prevent":[69],"low":[71],"efficiency":[72,152],"caused":[73],"random":[76],"number":[77],"and":[78,108,126,185],"positions":[79],"of":[80,153,175],"different":[85,105,114],"activations,":[86],"propose":[88],"activation-grouping,":[89],"so":[90],"that":[91],"same":[96],"group":[97],"can":[98,143,180],"be":[99],"computed":[100],"on":[101,134,167,189],"non-zero":[102],"channels":[106],"freely,":[107],"synchronization":[109],"only":[111],"needed":[112],"between":[113],"groups.":[115],"We":[116,149],"implement":[117],"proposed":[119],"16":[122,127],"process":[123],"(PU)":[125],"processing":[128],"elements":[129],"(PE)":[130],"each":[132],"PU":[133],"FPGA":[135],"built":[136],"upon":[137],"VTA":[138,171],"(Versatile":[139],"Tensor":[140],"Accelerator)":[141],"which":[142,162],"integrate":[144],"seamlessly":[145],"TVM":[147],"compilation.":[148],"evaluate":[150],"our":[154],"design":[155],"convolutional":[157],"layers":[158],"resnet18":[160],"respectively,":[161],"achieves":[163],"over":[164,182,186],"3.2x":[165],"speedup":[166,184],"average":[168],"compared":[169],"design.":[172],"terms":[174],"whole":[177],"network,":[178],"it":[179],"achieve":[181],"2.26x":[183],"2.0x":[187],"improvement":[188],"area":[190],"efficiency.":[191]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2024,"cited_by_count":1}],"updated_date":"2026-04-02T15:55:50.835912","created_date":"2025-10-10T00:00:00"}