{"id":"https://openalex.org/W4386896826","doi":"https://doi.org/10.1145/3625004","title":"Advancing Direct Convolution Using Convolution Slicing Optimization and ISA Extensions","display_name":"Advancing Direct Convolution Using Convolution Slicing Optimization and ISA Extensions","publication_year":2023,"publication_date":"2023-09-20","ids":{"openalex":"https://openalex.org/W4386896826","doi":"https://doi.org/10.1145/3625004"},"language":"en","primary_location":{"id":"doi:10.1145/3625004","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3625004","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3625004","source":{"id":"https://openalex.org/S26056741","display_name":"ACM Transactions on Architecture and Code Optimization","issn_l":"1544-3566","issn":["1544-3566","1544-3973"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Architecture and Code Optimization","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"diamond","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3625004","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5010935709","display_name":"Victor Ferrari","orcid":"https://orcid.org/0000-0002-2051-2877"},"institutions":[{"id":"https://openalex.org/I181391015","display_name":"Universidade Estadual de Campinas (UNICAMP)","ror":"https://ror.org/04wffgt70","country_code":"BR","type":"education","lineage":["https://openalex.org/I181391015"]},{"id":"https://openalex.org/I4210148978","display_name":"Hospital de Cl\u00ednicas da Unicamp","ror":"https://ror.org/05g89bp20","country_code":"BR","type":"healthcare","lineage":["https://openalex.org/I4210148978"]}],"countries":["BR"],"is_corresponding":true,"raw_author_name":"Victor Ferrari","raw_affiliation_strings":["Institute of Computing-UNICAMP, Brazil","Institute of Computing - UNICAMP, Brazil"],"affiliations":[{"raw_affiliation_string":"Institute of Computing-UNICAMP, Brazil","institution_ids":["https://openalex.org/I4210148978"]},{"raw_affiliation_string":"Institute of Computing - UNICAMP, Brazil","institution_ids":["https://openalex.org/I181391015"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103064453","display_name":"Rafael Sousa","orcid":"https://orcid.org/0000-0002-2673-6601"},"institutions":[{"id":"https://openalex.org/I181391015","display_name":"Universidade Estadual de Campinas (UNICAMP)","ror":"https://ror.org/04wffgt70","country_code":"BR","type":"education","lineage":["https://openalex.org/I181391015"]},{"id":"https://openalex.org/I4210148978","display_name":"Hospital de Cl\u00ednicas da Unicamp","ror":"https://ror.org/05g89bp20","country_code":"BR","type":"healthcare","lineage":["https://openalex.org/I4210148978"]}],"countries":["BR"],"is_corresponding":false,"raw_author_name":"Rafael Sousa","raw_affiliation_strings":["Institute of Computing-UNICAMP, Brazil","Institute of Computing - UNICAMP, Brazil"],"affiliations":[{"raw_affiliation_string":"Institute of Computing-UNICAMP, Brazil","institution_ids":["https://openalex.org/I4210148978"]},{"raw_affiliation_string":"Institute of Computing - UNICAMP, Brazil","institution_ids":["https://openalex.org/I181391015"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101887660","display_name":"Marcio Pereira","orcid":"https://orcid.org/0000-0003-1140-4513"},"institutions":[{"id":"https://openalex.org/I181391015","display_name":"Universidade Estadual de Campinas (UNICAMP)","ror":"https://ror.org/04wffgt70","country_code":"BR","type":"education","lineage":["https://openalex.org/I181391015"]},{"id":"https://openalex.org/I4210148978","display_name":"Hospital de Cl\u00ednicas da Unicamp","ror":"https://ror.org/05g89bp20","country_code":"BR","type":"healthcare","lineage":["https://openalex.org/I4210148978"]}],"countries":["BR"],"is_corresponding":false,"raw_author_name":"Marcio Pereira","raw_affiliation_strings":["Institute of Computing-UNICAMP, Brazil","Institute of Computing - UNICAMP, Brazil"],"affiliations":[{"raw_affiliation_string":"Institute of Computing-UNICAMP, Brazil","institution_ids":["https://openalex.org/I4210148978"]},{"raw_affiliation_string":"Institute of Computing - UNICAMP, Brazil","institution_ids":["https://openalex.org/I181391015"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090480187","display_name":"Jo\u00e3o P. L. de Carvalho","orcid":"https://orcid.org/0000-0002-3476-184X"},"institutions":[{"id":"https://openalex.org/I154425047","display_name":"University of Alberta","ror":"https://ror.org/0160cpw27","country_code":"CA","type":"education","lineage":["https://openalex.org/I154425047"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Jo\u00e3o P. L. De Carvalho","raw_affiliation_strings":["University of Alberta, Canada"],"affiliations":[{"raw_affiliation_string":"University of Alberta, Canada","institution_ids":["https://openalex.org/I154425047"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048554669","display_name":"Jos\u00e9 Nelson Amaral","orcid":"https://orcid.org/0000-0002-9943-1809"},"institutions":[{"id":"https://openalex.org/I154425047","display_name":"University of Alberta","ror":"https://ror.org/0160cpw27","country_code":"CA","type":"education","lineage":["https://openalex.org/I154425047"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Jos\u00e9 Nelson Amaral","raw_affiliation_strings":["University of Alberta, Canada"],"affiliations":[{"raw_affiliation_string":"University of Alberta, Canada","institution_ids":["https://openalex.org/I154425047"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055326969","display_name":"Jos\u00e9 E. Moreira","orcid":"https://orcid.org/0000-0001-7029-6327"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jos\u00e9 Moreira","raw_affiliation_strings":["IBM Research, United States of America"],"affiliations":[{"raw_affiliation_string":"IBM Research, United States of America","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5067675731","display_name":"Guido Ara\u00fajo","orcid":"https://orcid.org/0000-0003-4869-5190"},"institutions":[{"id":"https://openalex.org/I181391015","display_name":"Universidade Estadual de Campinas (UNICAMP)","ror":"https://ror.org/04wffgt70","country_code":"BR","type":"education","lineage":["https://openalex.org/I181391015"]},{"id":"https://openalex.org/I4210148978","display_name":"Hospital de Cl\u00ednicas da Unicamp","ror":"https://ror.org/05g89bp20","country_code":"BR","type":"healthcare","lineage":["https://openalex.org/I4210148978"]}],"countries":["BR"],"is_corresponding":false,"raw_author_name":"Guido Araujo","raw_affiliation_strings":["Institute of Computing\u2013UNICAMP, Brazil","Institute of Computing - UNICAMP, Brazil"],"affiliations":[{"raw_affiliation_string":"Institute of Computing\u2013UNICAMP, Brazil","institution_ids":["https://openalex.org/I4210148978"]},{"raw_affiliation_string":"Institute of Computing - UNICAMP, Brazil","institution_ids":["https://openalex.org/I181391015"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5010935709"],"corresponding_institution_ids":["https://openalex.org/I181391015","https://openalex.org/I4210148978"],"apc_list":null,"apc_paid":null,"fwci":3.3275,"has_fulltext":true,"cited_by_count":11,"citation_normalized_percentile":{"value":0.93033604,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":99},"biblio":{"volume":"20","issue":"4","first_page":"1","last_page":"26"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9997000098228455,"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":0.9997000098228455,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9979000091552734,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.9830999970436096,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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":0.8375349044799805},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.7122251391410828},{"id":"https://openalex.org/keywords/kernel","display_name":"Kernel (algebra)","score":0.5809947848320007},{"id":"https://openalex.org/keywords/kernelization","display_name":"Kernelization","score":0.48613378405570984},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.485041081905365},{"id":"https://openalex.org/keywords/x86","display_name":"x86","score":0.4401785433292389},{"id":"https://openalex.org/keywords/convolution","display_name":"Convolution (computer science)","score":0.415219783782959},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3888806700706482},{"id":"https://openalex.org/keywords/computational-science","display_name":"Computational science","score":0.37843868136405945},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2370208501815796},{"id":"https://openalex.org/keywords/parameterized-complexity","display_name":"Parameterized complexity","score":0.22112047672271729},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.2019309401512146},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.11609584093093872},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.09603986144065857}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8375349044799805},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.7122251391410828},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.5809947848320007},{"id":"https://openalex.org/C207225210","wikidata":"https://www.wikidata.org/wiki/Q1759539","display_name":"Kernelization","level":3,"score":0.48613378405570984},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.485041081905365},{"id":"https://openalex.org/C170723468","wikidata":"https://www.wikidata.org/wiki/Q182933","display_name":"x86","level":3,"score":0.4401785433292389},{"id":"https://openalex.org/C45347329","wikidata":"https://www.wikidata.org/wiki/Q5166604","display_name":"Convolution (computer science)","level":3,"score":0.415219783782959},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3888806700706482},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.37843868136405945},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2370208501815796},{"id":"https://openalex.org/C165464430","wikidata":"https://www.wikidata.org/wiki/Q1570441","display_name":"Parameterized complexity","level":2,"score":0.22112047672271729},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.2019309401512146},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.11609584093093872},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.09603986144065857},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3625004","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3625004","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3625004","source":{"id":"https://openalex.org/S26056741","display_name":"ACM Transactions on Architecture and Code Optimization","issn_l":"1544-3566","issn":["1544-3566","1544-3973"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Architecture and Code Optimization","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1145/3625004","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3625004","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3625004","source":{"id":"https://openalex.org/S26056741","display_name":"ACM Transactions on Architecture and Code Optimization","issn_l":"1544-3566","issn":["1544-3566","1544-3973"],"is_oa":true,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Architecture and Code Optimization","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4386896826.pdf","grobid_xml":"https://content.openalex.org/works/W4386896826.grobid-xml"},"referenced_works_count":23,"referenced_works":["https://openalex.org/W1667652561","https://openalex.org/W2073061372","https://openalex.org/W2097117768","https://openalex.org/W2194775991","https://openalex.org/W2252007067","https://openalex.org/W2560674852","https://openalex.org/W2578027111","https://openalex.org/W2605739168","https://openalex.org/W2913668833","https://openalex.org/W2950094539","https://openalex.org/W2998957070","https://openalex.org/W3093756531","https://openalex.org/W3093772369","https://openalex.org/W3123054690","https://openalex.org/W3147409832","https://openalex.org/W3156745629","https://openalex.org/W4200342808","https://openalex.org/W4206841102","https://openalex.org/W4282959697","https://openalex.org/W4285060868","https://openalex.org/W4308596141","https://openalex.org/W4311119077","https://openalex.org/W4311543031"],"related_works":["https://openalex.org/W3020640542","https://openalex.org/W2292973071","https://openalex.org/W2476602023","https://openalex.org/W2381408042","https://openalex.org/W4245720151","https://openalex.org/W1606657688","https://openalex.org/W4301454895","https://openalex.org/W4318567365","https://openalex.org/W2909144890","https://openalex.org/W3157543420"],"abstract_inverted_index":{"Convolution":[0,56,75],"is":[1,24,173,195],"one":[2],"of":[3,123,130,145,177,218],"the":[4,27,71,121,124,128,175,219],"most":[5],"computationally":[6],"intensive":[7],"operations":[8],"that":[9,45,65,80,120],"must":[10],"be":[11,47],"performed":[12],"for":[13,103,168,179,184,192],"machine":[14,116],"learning":[15,117],"model":[16,143,171,193],"inference.":[17],"A":[18],"traditional":[19],"approach":[20],"to":[21,83],"computing":[22,211],"convolutions":[23,104,112,213],"known":[25],"as":[26],"Im2Col":[28,125,159],"+":[29,160],"BLAS":[30,161,166,208],"method.":[31],"This":[32,52],"article":[33],"proposes":[34],"SConv:":[35],"a":[36,85,136],"direct-convolution":[37,87],"algorithm":[38,53],"based":[39,98,163],"on":[40,67,99,110,141,147,152,164,197,202],"an":[41,158],"MLIR/LLVM":[42],"code-generation":[43,78],"toolchain":[44],"can":[46],"integrated":[48],"into":[49],"machine-learning":[50,170],"compilers.":[51],"introduces:":[54],"(a)":[55],"Slicing":[57,76],"Analysis":[58],"(CSA)\u2014a":[59],"convolution-specific":[60],"3D":[61],"cache-blocking":[62],"analysis":[63],"pass":[64,79],"focuses":[66],"tile":[68],"reuse":[69],"over":[70,157],"cache":[72],"hierarchy;":[73],"(b)":[74],"Optimization\u2014a":[77],"uses":[81],"CSA":[82],"generate":[84],"tiled":[86],"macro-kernel;":[88],"and":[89,127,150,182,200],"(c)":[90],"Vector-based":[91],"Packing\u2014an":[92],"architecture-specific":[93],"optimized":[94],"input-tensor":[95],"packing":[96,132,138],"solution":[97],"vector-register":[100],"shift":[101],"instructions":[102],"with":[105],"unitary":[106],"stride.":[107],"Experiments":[108],"conducted":[109],"393":[111],"from":[113],"full":[114,142],"ONNX-MLIR":[115],"models":[118],"indicate":[119],"elimination":[122],"transformation":[126],"use":[129],"fast":[131],"routines":[133],"result":[134],"in":[135,174,214],"total":[137,189],"time":[139],"reduction,":[140],"inference,":[144],"2.3\u00d7\u20134.0\u00d7":[146],"Intel":[148,180,198],"x86":[149,181,199],"3.3\u00d7\u20135.9\u00d7":[151],"IBM":[153,185,203],"POWER10.":[154,204],"The":[155,188],"speed-up":[156],"method":[162],"current":[165],"implementations":[167],"end-to-end":[169],"inference":[172,194],"range":[176],"11%\u201327%":[178],"11%\u201334%":[183],"POWER10":[186],"architectures.":[187],"convolution":[190],"speedup":[191],"13%\u201328%":[196],"23%\u201339%":[201],"SConv":[205],"also":[206],"outperforms":[207],"GEMM,":[209],"when":[210],"pointwise":[212],"more":[215],"than":[216],"82%":[217],"219":[220],"tested":[221],"instances.":[222]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":4},{"year":2023,"cited_by_count":2}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}
