{"id":"https://openalex.org/W4408521258","doi":"https://doi.org/10.1109/tc.2025.3551937","title":"Deep Learning Operators Performance Tuning for Changeable Sized Input Data on Tensor Accelerate Hardware","display_name":"Deep Learning Operators Performance Tuning for Changeable Sized Input Data on Tensor Accelerate Hardware","publication_year":2025,"publication_date":"2025-03-17","ids":{"openalex":"https://openalex.org/W4408521258","doi":"https://doi.org/10.1109/tc.2025.3551937"},"language":"en","primary_location":{"id":"doi:10.1109/tc.2025.3551937","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2025.3551937","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","0016-9340","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/A5092246483","display_name":"Pengyu Mu","orcid":"https://orcid.org/0000-0002-4847-8148"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Pengyu Mu","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-4847-8148","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yi Liu","orcid":"https://orcid.org/0000-0003-1829-2817"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yi Liu","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-1829-2817","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Rui Wang","orcid":"https://orcid.org/0000-0003-2741-6033"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Rui Wang","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-2741-6033","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101656256","display_name":"Guoxiang Liu","orcid":"https://orcid.org/0009-0001-6163-2074"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Guoxiang Liu","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0001-6163-2074","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hangcheng An","orcid":"https://orcid.org/0009-0000-8932-1297"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hangcheng An","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0000-8932-1297","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5106660574","display_name":"Qianhe Zhao","orcid":null},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qianhe Zhao","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0006-1379-0576","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5018705589","display_name":"Hailong Yang","orcid":"https://orcid.org/0000-0003-1101-7927"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hailong Yang","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0003-1101-7927","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024508481","display_name":"Chenhao Xie","orcid":"https://orcid.org/0000-0002-1399-0352"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chenhao Xie","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-1399-0352","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074183877","display_name":"Zhongzhi Luan","orcid":"https://orcid.org/0000-0002-7186-0556"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhongzhi Luan","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-7186-0556","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5061134505","display_name":"Chunye Gong","orcid":"https://orcid.org/0000-0003-0349-1100"},"institutions":[{"id":"https://openalex.org/I4210165734","display_name":"National Supercomputing Center of Tianjin","ror":"https://ror.org/05tngxm14","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210165734"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chunye Gong","raw_affiliation_strings":["National Supercomputer Center, Tianjin, China"],"raw_orcid":"https://orcid.org/0000-0003-0349-1100","affiliations":[{"raw_affiliation_string":"National Supercomputer Center, Tianjin, China","institution_ids":["https://openalex.org/I4210165734"]}]},{"author_position":"last","author":{"id":null,"display_name":"Depei Qian","orcid":"https://orcid.org/0000-0002-5382-1473"},"institutions":[{"id":"https://openalex.org/I82880672","display_name":"Beihang University","ror":"https://ror.org/00wk2mp56","country_code":"CN","type":"education","lineage":["https://openalex.org/I82880672"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Depei Qian","raw_affiliation_strings":["School of Computer Science and Engineering, Beihang University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-5382-1473","affiliations":[{"raw_affiliation_string":"School of Computer Science and Engineering, Beihang University, Beijing, China","institution_ids":["https://openalex.org/I82880672"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":4.188,"has_fulltext":false,"cited_by_count":3,"citation_normalized_percentile":{"value":0.93341458,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":98,"max":99},"biblio":{"volume":"74","issue":"6","first_page":"2101","last_page":"2113"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13650","display_name":"Computational Physics and Python Applications","score":0.9829000234603882,"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/T13650","display_name":"Computational Physics and Python Applications","score":0.9829000234603882,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9763000011444092,"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/T12303","display_name":"Tensor decomposition and applications","score":0.965499997138977,"subfield":{"id":"https://openalex.org/subfields/2605","display_name":"Computational Mathematics"},"field":{"id":"https://openalex.org/fields/26","display_name":"Mathematics"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7122288942337036},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.5578119158744812},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.475154846906662},{"id":"https://openalex.org/keywords/tensor","display_name":"Tensor (intrinsic definition)","score":0.4623177647590637},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.4071807861328125},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.35657793283462524},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3314647674560547},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.32836222648620605},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.0926329493522644}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7122288942337036},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.5578119158744812},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.475154846906662},{"id":"https://openalex.org/C155281189","wikidata":"https://www.wikidata.org/wiki/Q3518150","display_name":"Tensor (intrinsic definition)","level":2,"score":0.4623177647590637},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.4071807861328125},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.35657793283462524},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3314647674560547},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.32836222648620605},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0926329493522644},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tc.2025.3551937","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tc.2025.3551937","pdf_url":null,"source":{"id":"https://openalex.org/S157670870","display_name":"IEEE Transactions on Computers","issn_l":"0018-9340","issn":["0018-9340","0016-9340","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/G4039879177","display_name":null,"funder_award_id":"U23B2020","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G5610567887","display_name":null,"funder_award_id":"92373110","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G5774836301","display_name":"\u9ad8\u6027\u80fd\u3001\u9ad8\u53ef\u6269\u5c55\u548c\u9ad8\u53ef\u79fb\u690d\u7684Stencil\u4ee3\u7801\u751f\u6210\u548c\u4f18\u5316\u6846\u67b6\u7814\u7a76","funder_award_id":"62072018","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G5832729736","display_name":null,"funder_award_id":"62032023","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G6139273093","display_name":null,"funder_award_id":"62322201","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G8014835415","display_name":null,"funder_award_id":"U22A2028","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G928764723","display_name":null,"funder_award_id":"62341411","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G932256194","display_name":null,"funder_award_id":"2022ZD0161902","funder_id":"https://openalex.org/F4320335777","funder_display_name":"National Key Research and Development Program 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/F4320335777","display_name":"National Key Research and Development Program of China","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":28,"referenced_works":["https://openalex.org/W2295598076","https://openalex.org/W2896457183","https://openalex.org/W2961619211","https://openalex.org/W3012249773","https://openalex.org/W3122286897","https://openalex.org/W3123639931","https://openalex.org/W3161395920","https://openalex.org/W3177452048","https://openalex.org/W4281932226","https://openalex.org/W4296473245","https://openalex.org/W4308596141","https://openalex.org/W4312727772","https://openalex.org/W4318541538","https://openalex.org/W4327930463","https://openalex.org/W4376466880","https://openalex.org/W4381785750","https://openalex.org/W4387350504","https://openalex.org/W4387789669","https://openalex.org/W4394627365","https://openalex.org/W4394841768","https://openalex.org/W4394998528","https://openalex.org/W6637151318","https://openalex.org/W6751349269","https://openalex.org/W6761772364","https://openalex.org/W6779627212","https://openalex.org/W6779728309","https://openalex.org/W6839338829","https://openalex.org/W6893864439"],"related_works":["https://openalex.org/W4375867731","https://openalex.org/W2611989081","https://openalex.org/W2731899572","https://openalex.org/W4230611425","https://openalex.org/W4294635752","https://openalex.org/W4304166257","https://openalex.org/W4383066092","https://openalex.org/W3215138031","https://openalex.org/W2804383999","https://openalex.org/W2802049774"],"abstract_inverted_index":{"The":[0,77],"operator":[1,53,127,185,255],"library":[2,15,54],"is":[3,20,31,71,91,105,116],"the":[4,14,23,43,52,61,69,86,96,110,113,119,140,154,180,184,203,211,221,227,232,251,265,275],"fundamental":[5],"infrastructure":[6],"of":[7,35,88,99,112,149,158,169,179,183,196,267,281],"deep":[8,48,125,260],"learning":[9,49,126,261],"acceleration":[10],"hardware.":[11,137],"Automatically":[12],"generating":[13],"and":[16,28,38,198,206,237,273],"tuning":[17,171,222,276],"its":[18],"performance":[19,70,128,244,248,266],"promising":[21],"because":[22,60,85],"manual":[24],"development":[25],"by":[26,230,269,278],"well-trained":[27],"skillful":[29],"programmers":[30],"costly":[32],"in":[33,80],"terms":[34],"both":[36],"time":[37,277],"money.":[39],"Tensor":[40],"hardware":[41,63,103,234],"has":[42],"best":[44],"computing":[45,97,151,197,204,213],"efficiency":[46],"for":[47,131,160,243],"applications,":[50],"but":[51],"programs":[55],"are":[56],"hard":[57],"to":[58,73,101,173,201,209,215,250,258],"tune":[59],"tensor":[62,102,115,136,163,216],"primitives":[64],"have":[65],"many":[66],"limitations.":[67],"Otherwise,":[68],"difficult":[72],"be":[74],"fully":[75],"explored.":[76],"recent":[78],"advancement":[79],"LLM":[81],"exacerbates":[82],"this":[83],"problem":[84],"size":[87],"input":[89,114,133,141,162],"data":[90,134],"not":[92],"fixed.":[93],"Therefore,":[94],"mapping":[95],"tasks":[98],"operators":[100,159],"units":[104,152,214],"a":[106,124,147,167,175],"significant":[107],"challenge":[108],"when":[109],"shape":[111],"unknown":[117],"before":[118],"runtime.":[120],"We":[121,165],"propose":[122],"DSAT,":[123],"autotuning":[129],"technique":[130],"changeable-sized":[132,161],"on":[135,188,271],"To":[138,218],"match":[139],"tensor's":[142],"undetermined":[143],"shape,":[144],"we":[145,191,224],"choose":[146],"group":[148,168],"abstract":[150,212],"as":[153],"basic":[155],"building":[156],"blocks":[157],"shapes.":[164],"design":[166],"programming":[170],"rules":[172],"construct":[174,192],"large":[176],"exploration":[177],"space":[178,229],"variant":[181],"implementation":[182],"programs.":[186],"Based":[187],"these":[189],"rules,":[190],"an":[193,239,279],"intermediate":[194],"representation":[195],"memory":[199],"access":[200],"describe":[202],"process":[205],"use":[207],"it":[208,263],"map":[210],"primitives.":[217],"speed":[219],"up":[220],"process,":[223],"narrow":[225],"down":[226],"optimization":[228],"predicting":[231],"actual":[233],"resource":[235],"requirement":[236],"providing":[238],"optimized":[240],"cost":[241],"model":[242],"prediction.":[245],"DSAT":[246],"achieves":[247],"comparable":[249],"vendor's":[252],"manually":[253],"tuned":[254],"libraries.":[256],"Compared":[257],"state-of-the-art":[259],"compilers,":[262],"improves":[264],"inference":[268],"13%":[270],"average":[272],"decreases":[274],"order":[280],"magnitude.":[282]},"counts_by_year":[{"year":2026,"cited_by_count":3}],"updated_date":"2025-12-21T01:58:51.020947","created_date":"2025-03-18T00:00:00"}