{"id":"https://openalex.org/W4385444817","doi":"https://doi.org/10.1109/tc.2023.3299030","title":"Enabling Fine-Grained Spatial Multitasking on Systolic-Array NPUs Using Dataflow Mirroring","display_name":"Enabling Fine-Grained Spatial Multitasking on Systolic-Array NPUs Using Dataflow Mirroring","publication_year":2023,"publication_date":"2023-08-01","ids":{"openalex":"https://openalex.org/W4385444817","doi":"https://doi.org/10.1109/tc.2023.3299030"},"language":"en","primary_location":{"id":"doi:10.1109/tc.2023.3299030","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2023.3299030","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","1557-9956","2326-3814"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Computers","raw_type":"journal-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/A5042838206","display_name":"Jinwoo Choi","orcid":"https://orcid.org/0000-0003-2363-501X"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Jinwoo Choi","raw_affiliation_strings":["Department of Computer Science, Yonsei University, Seoul, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026190390","display_name":"Yeonan Ha","orcid":"https://orcid.org/0000-0002-3418-5299"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Yeonan Ha","raw_affiliation_strings":["Department of Computer Science, Yonsei University, Seoul, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052472647","display_name":"Jounghoo Lee","orcid":"https://orcid.org/0000-0002-0463-7717"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]},{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jounghoo Lee","raw_affiliation_strings":["Department of Computer Science, Yonsei University, Seoul, South Korea","Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101475292","display_name":"Sangsu Lee","orcid":"https://orcid.org/0000-0003-1033-9625"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Sangsu Lee","raw_affiliation_strings":["Department of Computer Science, Yonsei University, Seoul, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100447633","display_name":"Jinho Lee","orcid":"https://orcid.org/0000-0003-4010-6611"},"institutions":[{"id":"https://openalex.org/I139264467","display_name":"Seoul National University","ror":"https://ror.org/04h9pn542","country_code":"KR","type":"education","lineage":["https://openalex.org/I139264467"]},{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Jinho Lee","raw_affiliation_strings":["Department of Computer Science, Yonsei University, Seoul, South Korea","Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]},{"raw_affiliation_string":"Department of Electrical and Computer Engineering, Seoul National University, Seoul, South Korea","institution_ids":["https://openalex.org/I139264467"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013182828","display_name":"Hanhwi Jang","orcid":"https://orcid.org/0000-0003-3722-4131"},"institutions":[{"id":"https://openalex.org/I57664883","display_name":"Ajou University","ror":"https://ror.org/03tzb2h73","country_code":"KR","type":"education","lineage":["https://openalex.org/I57664883"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hanhwi Jang","raw_affiliation_strings":["Department of Electrical and Computer Engineering, Ajou University, Suwon, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Electrical and Computer Engineering, Ajou University, Suwon, South Korea","institution_ids":["https://openalex.org/I57664883"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5088551237","display_name":"Youngsok Kim","orcid":"https://orcid.org/0000-0002-1015-9969"},"institutions":[{"id":"https://openalex.org/I193775966","display_name":"Yonsei University","ror":"https://ror.org/01wjejq96","country_code":"KR","type":"education","lineage":["https://openalex.org/I193775966"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Youngsok Kim","raw_affiliation_strings":["Department of Computer Science, Yonsei University, Seoul, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science, Yonsei University, Seoul, South Korea","institution_ids":["https://openalex.org/I193775966"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5042838206"],"corresponding_institution_ids":["https://openalex.org/I193775966"],"apc_list":null,"apc_paid":null,"fwci":1.2036,"has_fulltext":false,"cited_by_count":10,"citation_normalized_percentile":{"value":0.81249346,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":99},"biblio":{"volume":"72","issue":"12","first_page":"3383","last_page":"3398"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9997000098228455,"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":0.9997000098228455,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.9991000294685364,"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/T12808","display_name":"Ferroelectric and Negative Capacitance Devices","score":0.9940999746322632,"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/human-multitasking","display_name":"Human multitasking","score":0.7542960047721863},{"id":"https://openalex.org/keywords/dataflow","display_name":"Dataflow","score":0.6986303329467773},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.698189377784729},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.45080018043518066},{"id":"https://openalex.org/keywords/systolic-array","display_name":"Systolic array","score":0.4384400248527527},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.3260859251022339},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.12816229462623596}],"concepts":[{"id":"https://openalex.org/C107418235","wikidata":"https://www.wikidata.org/wiki/Q1520565","display_name":"Human multitasking","level":2,"score":0.7542960047721863},{"id":"https://openalex.org/C96324660","wikidata":"https://www.wikidata.org/wiki/Q205446","display_name":"Dataflow","level":2,"score":0.6986303329467773},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.698189377784729},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.45080018043518066},{"id":"https://openalex.org/C150741067","wikidata":"https://www.wikidata.org/wiki/Q2377218","display_name":"Systolic array","level":3,"score":0.4384400248527527},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3260859251022339},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.12816229462623596},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.0},{"id":"https://openalex.org/C14580979","wikidata":"https://www.wikidata.org/wiki/Q876049","display_name":"Very-large-scale integration","level":2,"score":0.0},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tc.2023.3299030","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2023.3299030","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","1557-9956","2326-3814"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Computers","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G6250211539","display_name":null,"funder_award_id":"2021R1F1A1062902","funder_id":"https://openalex.org/F4320322120","funder_display_name":"National Research Foundation of Korea"},{"id":"https://openalex.org/G747069884","display_name":null,"funder_award_id":"2022R1C1C1008131","funder_id":"https://openalex.org/F4320322120","funder_display_name":"National Research Foundation of Korea"},{"id":"https://openalex.org/G8446193549","display_name":null,"funder_award_id":"2022R1C1C1011307","funder_id":"https://openalex.org/F4320322120","funder_display_name":"National Research Foundation of Korea"}],"funders":[{"id":"https://openalex.org/F4320315121","display_name":"Samsung Advanced Institute of Technology","ror":null},{"id":"https://openalex.org/F4320322120","display_name":"National Research Foundation of Korea","ror":"https://ror.org/013aysd81"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":32,"referenced_works":["https://openalex.org/W2017369466","https://openalex.org/W2097117768","https://openalex.org/W2129381159","https://openalex.org/W2194775991","https://openalex.org/W2346205343","https://openalex.org/W2442974303","https://openalex.org/W2604319603","https://openalex.org/W2605350416","https://openalex.org/W2606722458","https://openalex.org/W2618530766","https://openalex.org/W2725159389","https://openalex.org/W2905515056","https://openalex.org/W2909419460","https://openalex.org/W2946332877","https://openalex.org/W2998732502","https://openalex.org/W3006586535","https://openalex.org/W3008591352","https://openalex.org/W3016939927","https://openalex.org/W3043406639","https://openalex.org/W3043571714","https://openalex.org/W3097528158","https://openalex.org/W3101026687","https://openalex.org/W3128330332","https://openalex.org/W3158233068","https://openalex.org/W3159535809","https://openalex.org/W3190761184","https://openalex.org/W4312518053","https://openalex.org/W4323038488","https://openalex.org/W4360831984","https://openalex.org/W4385245566","https://openalex.org/W6739901393","https://openalex.org/W6790607911"],"related_works":["https://openalex.org/W4388263628","https://openalex.org/W4242930893","https://openalex.org/W1988895983","https://openalex.org/W3140336604","https://openalex.org/W2488058330","https://openalex.org/W2998381397","https://openalex.org/W2940047422","https://openalex.org/W4237084280","https://openalex.org/W4242426637","https://openalex.org/W132134666"],"abstract_inverted_index":{"Neural":[0],"Processing":[1],"Units":[2],"(NPUs)":[3],"frequently":[4],"suffer":[5],"from":[6],"low":[7,37],"hardware":[8,39],"utilization":[9],"as":[10,145],"the":[11,20,36,42,89,132,138,150,159,163,173,189,196],"efficiency":[12],"of":[13,22,84,91,109,135,167],"their":[14],"systolic":[15,111],"arrays":[16],"heavily":[17],"depends":[18],"on":[19,158],"characteristics":[21],"a":[23,31,72,110,168],"deep":[24],"neural":[25],"network":[26],"(DNN).":[27],"Spatial":[28],"multitasking":[29],"is":[30,86],"promising":[32],"solution":[33],"to":[34,51,87,130,170,193],"overcome":[35],"NPU":[38,45,69,75,127,198],"utilization;":[40],"however,":[41],"state-of-the-art":[43,197],"spatial-multitasking":[44,74,133],"architecture":[46,76,199],"achieves":[47,142],"sub-optimal":[48,143],"performance":[49,128,134,140,144,165,190,201],"due":[50],"its":[52],"coarse-grained":[53],"systolic-array":[54,79,126],"distribution":[55],"and":[56,115,184,200],"incurs":[57],"significant":[58],"implementation":[59],"costs.":[60],"In":[61],"this":[62],"paper,":[63],"we":[64],"propose":[65,119],"<italic":[66,120],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[67,121],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">dataflow-mirroring":[68],"(DM-NPU)</i>":[70],",":[71,123],"novel":[73],"supporting":[77],"fine-grained":[78],"distribution.":[80],"The":[81],"key":[82],"idea":[83],"DM-NPU":[85,99,183],"reverse":[88],"dataflows":[90],"co-located":[92],"DNNs":[93,180],"in":[94],"horizontal":[95],"and/or":[96],"vertical":[97],"directions.":[98],"can":[100,186],"place":[101],"allocation":[102],"boundaries":[103],"between":[104],"any":[105],"adjacent":[106],"processing":[107],"elements":[108],"array,":[112],"both":[113],"horizontally":[114],"vertically.":[116],"We":[117],"then":[118],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">DM-Perf</i>":[122],"an":[124],"accurate":[125],"model,":[129],"maximize":[131],"DM-NPU.":[136],"Utilizing":[137],"existing":[139],"models":[141],"they":[146],"cannot":[147],"accurately":[148,171],"capture":[149,172],"resource":[151,174],"contention":[152],"caused":[153],"by":[154,191],"spatial":[155],"multitasking.":[156],"DM-Perf,":[157],"other":[160],"hand,":[161],"exploits":[162],"per-layer":[164],"profiles":[166],"DNN":[169],"contention.":[175],"Our":[176],"evaluation":[177],"using":[178],"MLPerf":[179],"shows":[181],"that":[182],"DM-Perf":[185],"greatly":[187],"improve":[188],"up":[192],"35.1%":[194],"over":[195],"model.":[202]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":7},{"year":2024,"cited_by_count":2}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}
