{"id":"https://openalex.org/W3214917546","doi":"https://doi.org/10.1109/isocc53507.2021.9613929","title":"Understanding and Reducing Weight-Load Overhead of Systolic Deep Learning Accelerators","display_name":"Understanding and Reducing Weight-Load Overhead of Systolic Deep Learning Accelerators","publication_year":2021,"publication_date":"2021-10-06","ids":{"openalex":"https://openalex.org/W3214917546","doi":"https://doi.org/10.1109/isocc53507.2021.9613929","mag":"3214917546"},"language":"en","primary_location":{"id":"doi:10.1109/isocc53507.2021.9613929","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isocc53507.2021.9613929","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 18th International SoC Design Conference (ISOCC)","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/A5013576423","display_name":"JinWon Joo","orcid":null},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"JinWon Joo","raw_affiliation_strings":["Electronic Engineering, Hanyang University"],"affiliations":[{"raw_affiliation_string":"Electronic Engineering, Hanyang University","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064813707","display_name":"Minyong Yoon","orcid":"https://orcid.org/0009-0003-0529-8184"},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Minyong Yoon","raw_affiliation_strings":["Electronic Engineering, Hanyang University"],"affiliations":[{"raw_affiliation_string":"Electronic Engineering, Hanyang University","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101935712","display_name":"Jungwook Choi","orcid":"https://orcid.org/0000-0002-5691-4771"},"institutions":[{"id":"https://openalex.org/I4575257","display_name":"Hanyang University","ror":"https://ror.org/046865y68","country_code":"KR","type":"education","lineage":["https://openalex.org/I4575257"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jungwook Choi","raw_affiliation_strings":["Electronic Engineering, Hanyang University"],"affiliations":[{"raw_affiliation_string":"Electronic Engineering, Hanyang University","institution_ids":["https://openalex.org/I4575257"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5006712350","display_name":"Mingu Kang","orcid":"https://orcid.org/0000-0001-8104-5136"},"institutions":[{"id":"https://openalex.org/I36258959","display_name":"University of California, San Diego","ror":"https://ror.org/0168r3w48","country_code":"US","type":"education","lineage":["https://openalex.org/I36258959"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mingu Kang","raw_affiliation_strings":["Electrical and Computer Engineering, University of California, San Diego"],"affiliations":[{"raw_affiliation_string":"Electrical and Computer Engineering, University of California, San Diego","institution_ids":["https://openalex.org/I36258959"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062501954","display_name":"Jong-Geon Lee","orcid":"https://orcid.org/0000-0001-9339-3277"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"JongGeon Lee","raw_affiliation_strings":["Samsung Electronics"],"affiliations":[{"raw_affiliation_string":"Samsung Electronics","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081625166","display_name":"Jinin So","orcid":"https://orcid.org/0000-0002-7569-3505"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"JinIn So","raw_affiliation_strings":["Samsung Electronics"],"affiliations":[{"raw_affiliation_string":"Samsung Electronics","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054514507","display_name":"Ilkwon Yun","orcid":"https://orcid.org/0000-0002-1641-4529"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"IlKwon Yun","raw_affiliation_strings":["Samsung Electronics"],"affiliations":[{"raw_affiliation_string":"Samsung Electronics","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089034192","display_name":"Yongsuk Kwon","orcid":"https://orcid.org/0000-0002-1956-4629"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Yongsuk Kwon","raw_affiliation_strings":["Samsung Electronics"],"affiliations":[{"raw_affiliation_string":"Samsung Electronics","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100719028","display_name":"Kyung-Soo Kim","orcid":"https://orcid.org/0000-0002-8927-1530"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"KyungSoo Kim","raw_affiliation_strings":["Samsung Electronics"],"affiliations":[{"raw_affiliation_string":"Samsung Electronics","institution_ids":["https://openalex.org/I2250650973"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":9,"corresponding_author_ids":["https://openalex.org/A5013576423"],"corresponding_institution_ids":["https://openalex.org/I4575257"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.18365415,"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":"413","last_page":"414"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9998000264167786,"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"}},"topics":[{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9998000264167786,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9994999766349792,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9988999962806702,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.8217770457267761},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7359373569488525},{"id":"https://openalex.org/keywords/systolic-array","display_name":"Systolic array","score":0.6836028695106506},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.6761330962181091},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.660588800907135},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5281994938850403},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.5277332067489624},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.492544025182724},{"id":"https://openalex.org/keywords/energy","display_name":"Energy (signal processing)","score":0.4757111668586731},{"id":"https://openalex.org/keywords/register-file","display_name":"Register file","score":0.4656525254249573},{"id":"https://openalex.org/keywords/path","display_name":"Path (computing)","score":0.41195064783096313},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.3324490189552307},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.28427204489707947},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.27248263359069824},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.1806633174419403},{"id":"https://openalex.org/keywords/instruction-set","display_name":"Instruction set","score":0.14045560359954834},{"id":"https://openalex.org/keywords/very-large-scale-integration","display_name":"Very-large-scale integration","score":0.10170647501945496},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.09611830115318298}],"concepts":[{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.8217770457267761},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7359373569488525},{"id":"https://openalex.org/C150741067","wikidata":"https://www.wikidata.org/wiki/Q2377218","display_name":"Systolic array","level":3,"score":0.6836028695106506},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6761330962181091},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.660588800907135},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5281994938850403},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.5277332067489624},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.492544025182724},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.4757111668586731},{"id":"https://openalex.org/C117280010","wikidata":"https://www.wikidata.org/wiki/Q180944","display_name":"Register file","level":3,"score":0.4656525254249573},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.41195064783096313},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3324490189552307},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.28427204489707947},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.27248263359069824},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.1806633174419403},{"id":"https://openalex.org/C202491316","wikidata":"https://www.wikidata.org/wiki/Q272683","display_name":"Instruction set","level":2,"score":0.14045560359954834},{"id":"https://openalex.org/C14580979","wikidata":"https://www.wikidata.org/wiki/Q876049","display_name":"Very-large-scale integration","level":2,"score":0.10170647501945496},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.09611830115318298},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/isocc53507.2021.9613929","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isocc53507.2021.9613929","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2021 18th International SoC Design Conference (ISOCC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","score":0.9100000262260437,"display_name":"Affordable and clean energy"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320322120","display_name":"National Research Foundation of Korea","ror":"https://ror.org/013aysd81"},{"id":"https://openalex.org/F4320332195","display_name":"Samsung","ror":"https://ror.org/04w3jy968"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":8,"referenced_works":["https://openalex.org/W2163605009","https://openalex.org/W2606722458","https://openalex.org/W2947737663","https://openalex.org/W3005538384","https://openalex.org/W3172967059","https://openalex.org/W6684191040","https://openalex.org/W6763737044","https://openalex.org/W6796875897"],"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/W1966837078","https://openalex.org/W4319431385","https://openalex.org/W3214917546","https://openalex.org/W4283703276"],"abstract_inverted_index":{"As":[0],"an":[1],"energy-efficient":[2,29],"computing":[3],"engine":[4],"for":[5,85],"deep":[6,20,89],"neural":[7],"network":[8],"inference,":[9],"2D":[10],"systolic":[11,32],"array":[12],"architectures":[13],"have":[14],"been":[15],"widely":[16],"adopted":[17],"in":[18],"modern":[19],"learning":[21,90],"accelerators.":[22],"However,":[23],"despite":[24],"high":[25],"compute":[26],"density":[27],"and":[28,80,99],"data":[30,40,60],"passing,":[31],"accelerators":[33],"suffer":[34],"a":[35,51,55,69,82],"non-trivial":[36],"overhead":[37,49,98],"of":[38,58],"loading":[39,48],"stationed":[41],"inside":[42],"their":[43],"local":[44],"register":[45],"file.":[46],"This":[47,66],"becomes":[50],"critical":[52],"issue":[53],"when":[54],"frequent":[56],"reload":[57],"stationary":[59],"(e.g.,":[61],"weight":[62],"parameters)":[63],"is":[64],"required.":[65],"paper":[67],"proposes":[68],"simple":[70],"yet":[71],"practical":[72],"SW-HW":[73],"co-optimization":[74],"that":[75],"reverses":[76],"the":[77,92,96],"weight-load":[78,97],"order":[79],"adds":[81],"dedicated":[83],"path":[84],"weight-load.":[86],"On":[87],"diverse":[88],"applications,":[91],"proposed":[93],"method":[94],"reduces":[95],"achieves":[100],"up":[101],"to":[102],"1.8\u00d7":[103],"speedup":[104],"with":[105],"40%":[106],"energy":[107],"savings.":[108]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}