{"id":"https://openalex.org/W4404057359","doi":"https://doi.org/10.1109/socc62300.2024.10737793","title":"Exploring Approximation and Dataflow Co-Optimization for Scalable Transformer Inference Architecture on the Edge","display_name":"Exploring Approximation and Dataflow Co-Optimization for Scalable Transformer Inference Architecture on the Edge","publication_year":2024,"publication_date":"2024-09-16","ids":{"openalex":"https://openalex.org/W4404057359","doi":"https://doi.org/10.1109/socc62300.2024.10737793"},"language":"en","primary_location":{"id":"doi:10.1109/socc62300.2024.10737793","is_oa":false,"landing_page_url":"https://doi.org/10.1109/socc62300.2024.10737793","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE 37th International System-on-Chip Conference (SOCC)","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/A5100317841","display_name":"He Liu","orcid":"https://orcid.org/0000-0001-9715-2606"},"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":"Liu He","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100733456","display_name":"Yujin Wang","orcid":"https://orcid.org/0000-0001-6587-1438"},"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":"Yujin Wang","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114189100","display_name":"Zongle Huang","orcid":"https://orcid.org/0009-0002-4557-3163"},"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":"Zongle Huang","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027809257","display_name":"Shupei Fan","orcid":"https://orcid.org/0009-0004-5314-3002"},"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":"Shupei Fan","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101867082","display_name":"Chen Tang","orcid":"https://orcid.org/0000-0001-6830-4113"},"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":"Chen Tang","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100699727","display_name":"Shuyuan Zhang","orcid":"https://orcid.org/0000-0001-7302-3359"},"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":"Shuyuan Zhang","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023201867","display_name":"Luchang Lei","orcid":"https://orcid.org/0000-0002-1293-7891"},"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":"Luchang Lei","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","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,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045721867","display_name":"Yongpan Liu","orcid":"https://orcid.org/0000-0002-4892-2309"},"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":"Yongpan Liu","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5088075152","display_name":"Hongyang Jia","orcid":"https://orcid.org/0000-0001-8692-1860"},"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":"Hongyang Jia","raw_affiliation_strings":["Tsinghua University,Department of Electronic Engineering,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University,Department of Electronic Engineering,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":10,"corresponding_author_ids":["https://openalex.org/A5100317841"],"corresponding_institution_ids":["https://openalex.org/I99065089"],"apc_list":null,"apc_paid":null,"fwci":0.3626,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.68490933,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"6"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10320","display_name":"Neural Networks and Applications","score":0.8665000200271606,"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/T10320","display_name":"Neural Networks and Applications","score":0.8665000200271606,"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/T10876","display_name":"Fault Detection and Control Systems","score":0.8226000070571899,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems 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/T11343","display_name":"Power Transformer Diagnostics and Insulation","score":0.8046000003814697,"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/dataflow","display_name":"Dataflow","score":0.7641716003417969},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7161199450492859},{"id":"https://openalex.org/keywords/dataflow-architecture","display_name":"Dataflow architecture","score":0.7090399861335754},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6987740397453308},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.6361860036849976},{"id":"https://openalex.org/keywords/architecture","display_name":"Architecture","score":0.5816187262535095},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.5147016644477844},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4111586809158325},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.3961324989795685},{"id":"https://openalex.org/keywords/theoretical-computer-science","display_name":"Theoretical computer science","score":0.34484773874282837},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2330990433692932},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.06454390287399292}],"concepts":[{"id":"https://openalex.org/C96324660","wikidata":"https://www.wikidata.org/wiki/Q205446","display_name":"Dataflow","level":2,"score":0.7641716003417969},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7161199450492859},{"id":"https://openalex.org/C176727019","wikidata":"https://www.wikidata.org/wiki/Q1172415","display_name":"Dataflow architecture","level":3,"score":0.7090399861335754},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6987740397453308},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6361860036849976},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.5816187262535095},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.5147016644477844},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4111586809158325},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.3961324989795685},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.34484773874282837},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2330990433692932},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.06454390287399292},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","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/C142362112","wikidata":"https://www.wikidata.org/wiki/Q735","display_name":"Art","level":0,"score":0.0},{"id":"https://openalex.org/C153349607","wikidata":"https://www.wikidata.org/wiki/Q36649","display_name":"Visual arts","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/socc62300.2024.10737793","is_oa":false,"landing_page_url":"https://doi.org/10.1109/socc62300.2024.10737793","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE 37th International System-on-Chip Conference (SOCC)","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":29,"referenced_works":["https://openalex.org/W2108598243","https://openalex.org/W2896457183","https://openalex.org/W2910396952","https://openalex.org/W2923014074","https://openalex.org/W2965373594","https://openalex.org/W3090555870","https://openalex.org/W3094502228","https://openalex.org/W3138516171","https://openalex.org/W3159727696","https://openalex.org/W3168124404","https://openalex.org/W3211525823","https://openalex.org/W4214686755","https://openalex.org/W4281758439","https://openalex.org/W4281773951","https://openalex.org/W4285601701","https://openalex.org/W4312894406","https://openalex.org/W4319996342","https://openalex.org/W4322718191","https://openalex.org/W4388327470","https://openalex.org/W4389166736","https://openalex.org/W6755207826","https://openalex.org/W6766673545","https://openalex.org/W6788001715","https://openalex.org/W6788135285","https://openalex.org/W6796494063","https://openalex.org/W6838322825","https://openalex.org/W6838956374","https://openalex.org/W6850625674","https://openalex.org/W6857972656"],"related_works":["https://openalex.org/W2564598376","https://openalex.org/W1484403103","https://openalex.org/W2584408851","https://openalex.org/W2115158825","https://openalex.org/W2101960124","https://openalex.org/W4377693460","https://openalex.org/W2033683327","https://openalex.org/W2783505431","https://openalex.org/W2521947294","https://openalex.org/W4236419692"],"abstract_inverted_index":{"Transformer-based":[0,46],"neural":[1],"networks":[2],"(NNs)":[3],"prevail":[4],"in":[5,36,72,79,85,101],"today\u2019s":[6],"artificial":[7],"intelligence":[8],"applications,":[9],"including":[10],"autonomous":[11],"driving,":[12],"natural":[13],"language":[14],"processing":[15],"and":[16,22,33,50,114,136,152],"generative":[17],"modeling,":[18],"showing":[19],"superior":[20],"accuracy":[21],"generalization":[23],"over":[24,156],"traditional":[25],"deep-learning":[26],"models.":[27],"However,":[28],"the":[29,37,42,54,95,99,130,133,139,157],"quadratic":[30],"scaling":[31,84],"computation":[32],"complex":[34],"dataflow":[35],"self-attention":[38],"set":[39],"challenges":[40,108],"to":[41,94,149],"efficient":[43],"deployment":[44],"of":[45,56,65,98,132],"NNs":[47],"on":[48],"edge":[49],"edge-server":[51],"devices,":[52],"where":[53],"latency":[55],"single-batch":[57],"inference":[58,81,134],"is":[59,91],"a":[60,118],"critical":[61],"concern.":[62],"The":[63],"lack":[64],"data":[66],"parallelism":[67,78],"necessitates":[68],"exploring":[69],"more":[70,75],"dimensions":[71],"tensor":[73],"parallelism,":[74],"specifically,":[76],"sequence":[77],"transformer":[80],"for":[82,125],"strong":[83],"domain-specific":[86],"accelerator":[87],"(DSA)":[88],"design,":[89],"which":[90,128],"non-trivial":[92],"due":[93],"temporal":[96],"dependency":[97],"max-finding":[100],"softmax":[102,126],"operators.":[103],"This":[104],"work":[105],"formulates":[106],"these":[107],"into":[109],"an":[110],"on-chip":[111],"buffering":[112,141],"problem,":[113],"then":[115],"puts":[116],"forward":[117],"hardware-software":[119],"co-design":[120],"approach":[121],"exploiting":[122],"max-findingfree":[123],"approximation":[124],"operators,":[127],"removes":[129],"blocking":[131],"pipeline":[135],"thus":[137],"alleviates":[138],"onchip":[140],"pressure.":[142],"An":[143],"example":[144],"architecture":[145],"design":[146],"shows":[147],"up":[148],"$2.83":[150],"\\times$":[151,154],"$28.02":[153],"speedup,":[155],"baseline":[158],"DSA":[159],"designs":[160],"respectively,":[161],"with":[162],"negligible":[163],"algorithmic":[164],"performance":[165],"loss.":[166]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2025-12-19T19:40:27.379048","created_date":"2025-10-10T00:00:00"}