{"id":"https://openalex.org/W4405361299","doi":"https://doi.org/10.1007/s42514-024-00200-3","title":"Optimizing depthwise separable convolution on DCU","display_name":"Optimizing depthwise separable convolution on DCU","publication_year":2024,"publication_date":"2024-12-01","ids":{"openalex":"https://openalex.org/W4405361299","doi":"https://doi.org/10.1007/s42514-024-00200-3"},"language":"en","primary_location":{"id":"doi:10.1007/s42514-024-00200-3","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s42514-024-00200-3","pdf_url":"https://link.springer.com/content/pdf/10.1007/s42514-024-00200-3.pdf","source":{"id":"https://openalex.org/S4210190911","display_name":"CCF Transactions on High Performance Computing","issn_l":"2524-4922","issn":["2524-4922","2524-4930"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"CCF Transactions on High Performance Computing","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://link.springer.com/content/pdf/10.1007/s42514-024-00200-3.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100423698","display_name":"Zheng Liu","orcid":"https://orcid.org/0000-0002-7260-209X"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zheng Liu","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100750431","display_name":"Hao Meng","orcid":"https://orcid.org/0009-0000-3476-3135"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Meng Hao","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5085221201","display_name":"Weizhe Zhang","orcid":"https://orcid.org/0000-0003-4783-876X"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Weizhe Zhang","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen, 518055, China"],"raw_orcid":"https://orcid.org/0000-0003-4783-876X","affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]},{"raw_affiliation_string":"School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen, 518055, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5082302042","display_name":"Gangzhao Lu","orcid":null},"institutions":[{"id":"https://openalex.org/I4210130112","display_name":"China Academy of Information and Communications Technology","ror":"https://ror.org/038dte259","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210130112","https://openalex.org/I890469752"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Gangzhao Lu","raw_affiliation_strings":["China Nanhu Academy of Electronics and Information Technology, Jiaxing, 314001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China Nanhu Academy of Electronics and Information Technology, Jiaxing, 314001, China","institution_ids":["https://openalex.org/I4210130112"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Xueyang Tian","orcid":null},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xueyang Tian","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101988298","display_name":"Siyu Yang","orcid":"https://orcid.org/0000-0001-9088-7482"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Siyu Yang","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083685156","display_name":"Min Xie","orcid":"https://orcid.org/0000-0002-2967-3490"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Mingdong Xie","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5076673098","display_name":"Jie Dai","orcid":"https://orcid.org/0000-0001-9848-6894"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jie Dai","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Chenyu Yuan","orcid":null},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chenyu Yuan","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100733744","display_name":"Desheng Wang","orcid":"https://orcid.org/0000-0002-7502-7094"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Desheng Wang","raw_affiliation_strings":["School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen, 518055, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen, 518055, China","institution_ids":["https://openalex.org/I204983213"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5032239293","display_name":"Hongwei Yang","orcid":"https://orcid.org/0000-0002-8386-0131"},"institutions":[{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongwei Yang","raw_affiliation_strings":["Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Faculty of Computing, Harbin Institute of Technology, Harbin, 150001, China","institution_ids":["https://openalex.org/I204983213"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5085221201"],"corresponding_institution_ids":["https://openalex.org/I204983213"],"apc_list":{"value":2190,"currency":"EUR","value_usd":2790},"apc_paid":{"value":2190,"currency":"EUR","value_usd":2790},"fwci":1.4845,"has_fulltext":true,"cited_by_count":8,"citation_normalized_percentile":{"value":0.83386515,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":98,"max":99},"biblio":{"volume":"6","issue":"6","first_page":"646","last_page":"664"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13114","display_name":"Image Processing Techniques and Applications","score":0.9988999962806702,"subfield":{"id":"https://openalex.org/subfields/2214","display_name":"Media Technology"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T13114","display_name":"Image Processing Techniques and Applications","score":0.9988999962806702,"subfield":{"id":"https://openalex.org/subfields/2214","display_name":"Media Technology"},"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/T10500","display_name":"Sparse and Compressive Sensing Techniques","score":0.998199999332428,"subfield":{"id":"https://openalex.org/subfields/2206","display_name":"Computational Mechanics"},"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/T11105","display_name":"Advanced Image Processing Techniques","score":0.998199999332428,"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/separable-space","display_name":"Separable space","score":0.6946854591369629},{"id":"https://openalex.org/keywords/convolution","display_name":"Convolution (computer science)","score":0.624304473400116},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.45302364230155945},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.3633359968662262},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.27685147523880005},{"id":"https://openalex.org/keywords/mathematical-analysis","display_name":"Mathematical analysis","score":0.16140782833099365},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.06257167458534241}],"concepts":[{"id":"https://openalex.org/C70710897","wikidata":"https://www.wikidata.org/wiki/Q680081","display_name":"Separable space","level":2,"score":0.6946854591369629},{"id":"https://openalex.org/C45347329","wikidata":"https://www.wikidata.org/wiki/Q5166604","display_name":"Convolution (computer science)","level":3,"score":0.624304473400116},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.45302364230155945},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3633359968662262},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.27685147523880005},{"id":"https://openalex.org/C134306372","wikidata":"https://www.wikidata.org/wiki/Q7754","display_name":"Mathematical analysis","level":1,"score":0.16140782833099365},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.06257167458534241}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1007/s42514-024-00200-3","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s42514-024-00200-3","pdf_url":"https://link.springer.com/content/pdf/10.1007/s42514-024-00200-3.pdf","source":{"id":"https://openalex.org/S4210190911","display_name":"CCF Transactions on High Performance Computing","issn_l":"2524-4922","issn":["2524-4922","2524-4930"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"CCF Transactions on High Performance Computing","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1007/s42514-024-00200-3","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s42514-024-00200-3","pdf_url":"https://link.springer.com/content/pdf/10.1007/s42514-024-00200-3.pdf","source":{"id":"https://openalex.org/S4210190911","display_name":"CCF Transactions on High Performance Computing","issn_l":"2524-4922","issn":["2524-4922","2524-4930"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"CCF Transactions on High Performance Computing","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G6668131176","display_name":null,"funder_award_id":"2023YFB4503205","funder_id":"https://openalex.org/F4320335774","funder_display_name":"Key Technologies Research and Development Program"},{"id":"https://openalex.org/G8725492322","display_name":null,"funder_award_id":"62202123","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/F4320335774","display_name":"Key Technologies Research and Development Program","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4405361299.pdf","grobid_xml":"https://content.openalex.org/works/W4405361299.grobid-xml"},"referenced_works_count":41,"referenced_works":["https://openalex.org/W1972971542","https://openalex.org/W2072912074","https://openalex.org/W2097117768","https://openalex.org/W2117539524","https://openalex.org/W2122729175","https://openalex.org/W2153190325","https://openalex.org/W2163687928","https://openalex.org/W2194775991","https://openalex.org/W2531409750","https://openalex.org/W2605739168","https://openalex.org/W2622148586","https://openalex.org/W2883780447","https://openalex.org/W2887936511","https://openalex.org/W2913790721","https://openalex.org/W2946289095","https://openalex.org/W2953212265","https://openalex.org/W2963037989","https://openalex.org/W2963163009","https://openalex.org/W2963404959","https://openalex.org/W2963918968","https://openalex.org/W2965658867","https://openalex.org/W2969388332","https://openalex.org/W2970971581","https://openalex.org/W2978452617","https://openalex.org/W2983670846","https://openalex.org/W2998957070","https://openalex.org/W3100789280","https://openalex.org/W3136416617","https://openalex.org/W3166036439","https://openalex.org/W3169512402","https://openalex.org/W3172509117","https://openalex.org/W4225124368","https://openalex.org/W4308155352","https://openalex.org/W4381280256","https://openalex.org/W4385698913","https://openalex.org/W4386896826","https://openalex.org/W4399154859","https://openalex.org/W4404918643","https://openalex.org/W6604344240","https://openalex.org/W6698039833","https://openalex.org/W6702248584"],"related_works":["https://openalex.org/W2899084033","https://openalex.org/W4391375266","https://openalex.org/W2530952058","https://openalex.org/W2582836483","https://openalex.org/W4299366318","https://openalex.org/W3158431807","https://openalex.org/W4297580547","https://openalex.org/W2951583185","https://openalex.org/W2531409750","https://openalex.org/W4308155352"],"abstract_inverted_index":{"Abstract":[0],"The":[1],"integration":[2],"of":[3,18,25,45,60,205],"Large":[4],"Language":[5],"Models":[6],"(LLMs)":[7],"with":[8],"Convolutional":[9],"Neural":[10],"Networks":[11],"(CNNs)":[12],"is":[13,28,42],"significantly":[14],"advancing":[15],"the":[16,22,40,43,58,78,81,166,170,203,210],"development":[17],"large":[19,26],"models.":[20],"However,":[21],"computational":[23],"cost":[24],"models":[27],"high,":[29],"necessitating":[30],"optimization":[31],"for":[32],"greater":[33],"efficiency.":[34,64],"One":[35],"effective":[36],"way":[37],"to":[38,56,80,102,120,174,188,213],"optimize":[39],"CNN":[41],"use":[44],"depthwise":[46,93,144,184],"separable":[47],"convolution":[48,147,185],"(DSC),":[49],"which":[50],"decouples":[51],"spatial":[52],"and":[53,62,72,107,130,132,145,150,186],"channel":[54,139],"convolutions":[55],"reduce":[57],"number":[59],"parameters":[61],"enhance":[63,134],"In":[65],"this":[66],"study,":[67],"we":[68,95,114,133],"focus":[69],"on":[70,169],"porting":[71],"optimizing":[73],"DSC":[74],"kernel":[75,148,163],"functions":[76,149,164],"from":[77],"GPU":[79],"Deep":[82],"Computing":[83],"Unit":[84],"(DCU),":[85],"a":[86,97,138,175,189],"computing":[87],"accelerator":[88],"developed":[89],"in":[90,183,197,208],"China.":[91],"For":[92,111],"convolution,":[94,113],"implement":[96,143],"row":[98],"data":[99,105],"reuse":[100],"algorithm":[101],"minimize":[103],"redundant":[104],"loading":[106],"memory":[108],"access":[109],"overhead.":[110],"pointwise":[112,146,198],"extend":[115],"our":[116,161,206],"dynamic":[117],"tiling":[118],"strategy":[119],"improve":[121],"hardware":[122],"utilization":[123],"by":[124],"balancing":[125],"resource":[126],"allocation":[127],"among":[128],"blocks":[129],"threads,":[131],"arithmetic":[135],"intensity":[136],"through":[137],"distribution":[140],"algorithm.":[141],"We":[142],"integrate":[151],"them":[152],"into":[153],"PyTorch":[154],"as":[155],"extension":[156],"modules.":[157],"Experiments":[158],"demonstrate":[159],"that":[160],"optimized":[162],"outperform":[165],"MIOpen":[167],"library":[168],"DCU,":[171],"achieving":[172],"up":[173,187],"3.59":[176],"$$\\times$$":[177,191],"<mml:math":[178,192],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\">":[179,193],"<mml:mo>\u00d7</mml:mo>":[180,194],"</mml:math>":[181,195],"speedup":[182,196],"3.54":[190],"convolution.":[199],"These":[200],"results":[201],"highlight":[202],"effectiveness":[204],"approach":[207],"leveraging":[209],"DCU\u2019s":[211],"architecture":[212],"accelerate":[214],"deep":[215],"learning":[216],"operations.":[217]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":5}],"updated_date":"2026-07-15T18:14:33.161393","created_date":"2025-10-10T00:00:00"}