{"id":"https://openalex.org/W4280582299","doi":"https://doi.org/10.23919/date54114.2022.9774692","title":"DTQAtten: Leveraging Dynamic Token-based Quantization for Efficient Attention Architecture","display_name":"DTQAtten: Leveraging Dynamic Token-based Quantization for Efficient Attention Architecture","publication_year":2022,"publication_date":"2022-03-14","ids":{"openalex":"https://openalex.org/W4280582299","doi":"https://doi.org/10.23919/date54114.2022.9774692"},"language":"en","primary_location":{"id":"doi:10.23919/date54114.2022.9774692","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date54114.2022.9774692","pdf_url":null,"source":{"id":"https://openalex.org/S4363607924","display_name":"2022 Design, Automation & Test in Europe Conference & Exhibition (DATE)","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":"2022 Design, Automation & Test in Europe Conference & Exhibition (DATE)","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/A5022287798","display_name":"Tao Yang","orcid":"https://orcid.org/0000-0001-8588-9483"},"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":"Tao Yang","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100683853","display_name":"Dongyue Li","orcid":"https://orcid.org/0000-0001-5781-2366"},"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":"Dongyue Li","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067161373","display_name":"Zhuoran Song","orcid":"https://orcid.org/0000-0002-6494-4786"},"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":"Zhuoran Song","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101870121","display_name":"Yilong Zhao","orcid":"https://orcid.org/0000-0002-4888-9027"},"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":"Yilong Zhao","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017670541","display_name":"Fangxin Liu","orcid":"https://orcid.org/0000-0002-8769-293X"},"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":"Fangxin Liu","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107269495","display_name":"Zongwu Wang","orcid":"https://orcid.org/0009-0003-2157-4927"},"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":"Zongwu Wang","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036755436","display_name":"Zhezhi He","orcid":"https://orcid.org/0000-0002-6357-236X"},"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":"Zhezhi He","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5053801300","display_name":"Li Jiang","orcid":"https://orcid.org/0000-0002-7353-8798"},"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"]},{"id":"https://openalex.org/I4210122302","display_name":"ShangHai JiAi Genetics & IVF Institute","ror":"https://ror.org/02rgbry52","country_code":"CN","type":"healthcare","lineage":["https://openalex.org/I4210122302"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Li Jiang","raw_affiliation_strings":["Shanghai Jiao Tong University,Shanghai,China","MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong University","Shanghai Qi Zhi Institute, Shanghai, China","Shanghai Jiao Tong University, Shanghai, China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"Shanghai Qi Zhi Institute, Shanghai, China","institution_ids":["https://openalex.org/I4210122302"]},{"raw_affiliation_string":"Shanghai Jiao Tong University, Shanghai, China","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5022287798"],"corresponding_institution_ids":["https://openalex.org/I183067930"],"apc_list":null,"apc_paid":null,"fwci":1.4933,"has_fulltext":false,"cited_by_count":25,"citation_normalized_percentile":{"value":0.88090845,"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":"700","last_page":"705"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9998000264167786,"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":0.9998000264167786,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.9984999895095825,"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/T10028","display_name":"Topic Modeling","score":0.9980000257492065,"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":0.8657089471817017},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.857098400592804},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.6099798083305359},{"id":"https://openalex.org/keywords/edge-device","display_name":"Edge device","score":0.5694154500961304},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.5532693266868591},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5508579611778259},{"id":"https://openalex.org/keywords/hardware-acceleration","display_name":"Hardware acceleration","score":0.48437294363975525},{"id":"https://openalex.org/keywords/efficient-energy-use","display_name":"Efficient energy use","score":0.47542908787727356},{"id":"https://openalex.org/keywords/memory-footprint","display_name":"Memory footprint","score":0.47082072496414185},{"id":"https://openalex.org/keywords/energy-consumption","display_name":"Energy consumption","score":0.44949954748153687},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.42206212878227234},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.38016462326049805},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.3138052821159363},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3041002154350281},{"id":"https://openalex.org/keywords/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.26854783296585083},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.264651358127594},{"id":"https://openalex.org/keywords/cloud-computing","display_name":"Cloud computing","score":0.2182348072528839}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8657089471817017},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.857098400592804},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.6099798083305359},{"id":"https://openalex.org/C138236772","wikidata":"https://www.wikidata.org/wiki/Q25098575","display_name":"Edge device","level":3,"score":0.5694154500961304},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.5532693266868591},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5508579611778259},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.48437294363975525},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.47542908787727356},{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.47082072496414185},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.44949954748153687},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.42206212878227234},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.38016462326049805},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.3138052821159363},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3041002154350281},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.26854783296585083},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.264651358127594},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.2182348072528839},{"id":"https://openalex.org/C86803240","wikidata":"https://www.wikidata.org/wiki/Q420","display_name":"Biology","level":0,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.0},{"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/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.0},{"id":"https://openalex.org/C18903297","wikidata":"https://www.wikidata.org/wiki/Q7150","display_name":"Ecology","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.23919/date54114.2022.9774692","is_oa":false,"landing_page_url":"https://doi.org/10.23919/date54114.2022.9774692","pdf_url":null,"source":{"id":"https://openalex.org/S4363607924","display_name":"2022 Design, Automation & Test in Europe Conference & Exhibition (DATE)","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":"2022 Design, Automation & Test in Europe Conference & Exhibition (DATE)","raw_type":"proceedings-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/G1303213773","display_name":null,"funder_award_id":"2018YFB1403400","funder_id":"https://openalex.org/F4320335777","funder_display_name":"National Key Research and Development Program of China"},{"id":"https://openalex.org/G744300498","display_name":null,"funder_award_id":"61834006","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320335777","display_name":"National Key Research and Development Program of China","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":20,"referenced_works":["https://openalex.org/W2131241448","https://openalex.org/W2289252105","https://openalex.org/W2606722458","https://openalex.org/W2896457183","https://openalex.org/W2949591530","https://openalex.org/W2998183051","https://openalex.org/W3019166713","https://openalex.org/W3035083896","https://openalex.org/W3035332806","https://openalex.org/W3043504674","https://openalex.org/W3108833523","https://openalex.org/W3159727696","https://openalex.org/W3170233084","https://openalex.org/W3177265267","https://openalex.org/W3178941450","https://openalex.org/W3213528054","https://openalex.org/W4385245566","https://openalex.org/W6678911119","https://openalex.org/W6739901393","https://openalex.org/W6788001715"],"related_works":["https://openalex.org/W4221139464","https://openalex.org/W4289827464","https://openalex.org/W2532502681","https://openalex.org/W4295943704","https://openalex.org/W3208617247","https://openalex.org/W4380681266","https://openalex.org/W3046471834","https://openalex.org/W3210256422","https://openalex.org/W3128751578","https://openalex.org/W3192190837"],"abstract_inverted_index":{"Models":[0],"based":[1],"on":[2,179],"the":[3,59,62,102,108,116,133,136,146,159,189,209],"attention":[4,138,211],"mechanism,":[5],"i.e.":[6],"transformers,":[7],"have":[8],"shown":[9],"extraordinary":[10],"performance":[11],"in":[12,69,101,162,197,203],"Natural":[13],"Language":[14],"Processing":[15],"(NLP)":[16],"tasks.":[17,182],"However,":[18],"their":[19],"memory":[20],"footprint,":[21],"inference":[22,32],"latency,":[23],"and":[24,51,104,151,177,201,221],"power":[25],"consumption":[26],"are":[27,99],"still":[28],"prohibitive":[29],"for":[30],"efficient":[31],"at":[33,37,217],"edge":[34],"devices,":[35],"even":[36],"data":[38],"centers.":[39],"To":[40],"tackle":[41],"this":[42],"issue,":[43],"we":[44,87],"present":[45,56],"an":[46,153],"algorithm-architecture":[47],"co-design":[48],"with":[49,81,145,170,208],"dynamic":[50,117],"mixed-precision":[52,118],"quantization,":[53],"DTQAtten.":[54],"We":[55,140,167],"empirically":[57],"that":[58,92,186],"tolerance":[60],"to":[61,67,77,115,157],"noise":[63],"varies":[64],"from":[65],"token":[66,68],"attention-based":[70,172],"NLP":[71,173],"models.":[72],"This":[73],"finding":[74],"leads":[75],"us":[76],"quantize":[78],"different":[79],"tokens":[80,96,119],"mixed":[82],"levels":[83],"of":[84,110,135,199,205],"bits.":[85],"Thus,":[86],"design":[88,142],"a":[89],"compression":[90],"framework":[91],"(i)":[93],"dynamically":[94],"quantizes":[95],"while":[97],"they":[98],"forwarded":[100],"models":[103],"(ii)":[105],"jointly":[106],"determines":[107],"ratio":[109],"each":[111],"precision.":[112],"Moreover,":[113],"due":[114],"caused":[120],"by":[121,195],"our":[122,143,214],"framework,":[123],"previous":[124,190],"matrix-multiplication":[125],"accelerators":[126],"(e.g.":[127],"systolic":[128,148],"array)":[129],"cannot":[130],"effectively":[131],"exploit":[132],"benefit":[134],"compressed":[137],"computation.":[139],"thus":[141],"accelerator":[144,193,212],"variable-speed":[147],"array":[149],"(VSSA)":[150],"propose":[152],"effective":[154],"optimization":[155],"strategy":[156],"alleviate":[158],"pipeline-stall":[160],"problem":[161],"VSSA":[163],"without":[164],"hardware":[165],"overhead.":[166],"conduct":[168],"experiments":[169],"existing":[171],"models,":[174],"including":[175],"BERT":[176],"GPT-2":[178],"various":[180],"language":[181],"Our":[183],"results":[184],"show":[185],"DTQAtten":[187,215],"outperforms":[188],"neural":[191],"network":[192],"Eyeriss":[194],"13.12\u00d7":[196],"terms":[198,204],"speedup":[200,220],"3.8\u00d7":[202],"energy-saving.":[206],"Compared":[207],"state-of-the-art":[210],"SpAtten,":[213],"achieves":[216],"least":[218],"2.65\u00d7":[219],"3.38\u00d7":[222],"energy":[223],"efficiency":[224],"improvement.":[225]},"counts_by_year":[{"year":2025,"cited_by_count":12},{"year":2024,"cited_by_count":7},{"year":2023,"cited_by_count":5},{"year":2022,"cited_by_count":1}],"updated_date":"2026-03-28T08:17:26.163206","created_date":"2025-10-10T00:00:00"}