{"id":"https://openalex.org/W4405055872","doi":"https://doi.org/10.1109/tpds.2024.3511543","title":"TOP: Task-Based Operator Parallelism for Asynchronous Deep Learning Inference on GPU","display_name":"TOP: Task-Based Operator Parallelism for Asynchronous Deep Learning Inference on GPU","publication_year":2024,"publication_date":"2024-12-05","ids":{"openalex":"https://openalex.org/W4405055872","doi":"https://doi.org/10.1109/tpds.2024.3511543"},"language":"en","primary_location":{"id":"doi:10.1109/tpds.2024.3511543","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2024.3511543","pdf_url":null,"source":{"id":"https://openalex.org/S97130795","display_name":"IEEE Transactions on Parallel and Distributed Systems","issn_l":"1045-9219","issn":["1045-9219","1558-2183","2161-9883"],"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 Parallel and Distributed Systems","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/A5065564945","display_name":"Changyao Lin","orcid":"https://orcid.org/0000-0001-6805-2649"},"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"]},{"id":"https://openalex.org/I4210161462","display_name":"Heilongjiang Institute of Technology","ror":"https://ror.org/05x0m9n95","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210161462"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Changyao Lin","raw_affiliation_strings":["School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Heilongjiang, China"],"affiliations":[{"raw_affiliation_string":"School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Heilongjiang, China","institution_ids":["https://openalex.org/I4210161462","https://openalex.org/I204983213"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024746141","display_name":"Zhenming Chen","orcid":"https://orcid.org/0000-0002-6461-0081"},"institutions":[{"id":"https://openalex.org/I874390226","display_name":"China Construction Bank","ror":"https://ror.org/00xkrw816","country_code":"CN","type":"other","lineage":["https://openalex.org/I874390226"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhenming Chen","raw_affiliation_strings":["China Construction Steel Structure Engineering Corp., LTD, Shenzhen, Guangdong, China"],"affiliations":[{"raw_affiliation_string":"China Construction Steel Structure Engineering Corp., LTD, Shenzhen, Guangdong, China","institution_ids":["https://openalex.org/I874390226"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014588008","display_name":"Ziyang Zhang","orcid":"https://orcid.org/0000-0003-2539-8257"},"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"]},{"id":"https://openalex.org/I4210161462","display_name":"Heilongjiang Institute of Technology","ror":"https://ror.org/05x0m9n95","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210161462"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ziyang Zhang","raw_affiliation_strings":["School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Heilongjiang, China"],"affiliations":[{"raw_affiliation_string":"School of Computer Science and Technology, Harbin Institute of Technology, Harbin, Heilongjiang, China","institution_ids":["https://openalex.org/I4210161462","https://openalex.org/I204983213"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100454005","display_name":"Jie Liu","orcid":"https://orcid.org/0000-0001-6209-6886"},"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 Liu","raw_affiliation_strings":["National Key Laboratory of Smart Farm Technologies and Systems, Harbin Institute of Technology, Harbin, Heilongjiang, China","International Research Institute for Artificial Intelligence, Harbin Institute of Technology, Shenzhen, Guangdong, China"],"affiliations":[{"raw_affiliation_string":"National Key Laboratory of Smart Farm Technologies and Systems, Harbin Institute of Technology, Harbin, Heilongjiang, China","institution_ids":["https://openalex.org/I204983213"]},{"raw_affiliation_string":"International Research Institute for Artificial Intelligence, Harbin Institute of Technology, Shenzhen, Guangdong, China","institution_ids":["https://openalex.org/I204983213"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5065564945"],"corresponding_institution_ids":["https://openalex.org/I204983213","https://openalex.org/I4210161462"],"apc_list":null,"apc_paid":null,"fwci":1.0245,"has_fulltext":false,"cited_by_count":3,"citation_normalized_percentile":{"value":0.81781645,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":96,"max":97},"biblio":{"volume":"36","issue":"2","first_page":"266","last_page":"281"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9269999861717224,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9269999861717224,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9222999811172485,"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/T10320","display_name":"Neural Networks and Applications","score":0.9204000234603882,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8758880496025085},{"id":"https://openalex.org/keywords/asynchronous-communication","display_name":"Asynchronous communication","score":0.7113842964172363},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.643711268901825},{"id":"https://openalex.org/keywords/parallelism","display_name":"Parallelism (grammar)","score":0.6064993739128113},{"id":"https://openalex.org/keywords/data-parallelism","display_name":"Data parallelism","score":0.6021349430084229},{"id":"https://openalex.org/keywords/task-parallelism","display_name":"Task parallelism","score":0.5835952758789062},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5356848239898682},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4991111755371094},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4801156520843506},{"id":"https://openalex.org/keywords/operator","display_name":"Operator (biology)","score":0.44862914085388184},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.40913158655166626}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8758880496025085},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.7113842964172363},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.643711268901825},{"id":"https://openalex.org/C2781172179","wikidata":"https://www.wikidata.org/wiki/Q853109","display_name":"Parallelism (grammar)","level":2,"score":0.6064993739128113},{"id":"https://openalex.org/C61483411","wikidata":"https://www.wikidata.org/wiki/Q3124522","display_name":"Data parallelism","level":3,"score":0.6021349430084229},{"id":"https://openalex.org/C42992933","wikidata":"https://www.wikidata.org/wiki/Q691169","display_name":"Task parallelism","level":3,"score":0.5835952758789062},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5356848239898682},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4991111755371094},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4801156520843506},{"id":"https://openalex.org/C17020691","wikidata":"https://www.wikidata.org/wiki/Q139677","display_name":"Operator (biology)","level":5,"score":0.44862914085388184},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.40913158655166626},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C158448853","wikidata":"https://www.wikidata.org/wiki/Q425218","display_name":"Repressor","level":4,"score":0.0},{"id":"https://openalex.org/C55493867","wikidata":"https://www.wikidata.org/wiki/Q7094","display_name":"Biochemistry","level":1,"score":0.0},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0},{"id":"https://openalex.org/C86339819","wikidata":"https://www.wikidata.org/wiki/Q407384","display_name":"Transcription factor","level":3,"score":0.0},{"id":"https://openalex.org/C104317684","wikidata":"https://www.wikidata.org/wiki/Q7187","display_name":"Gene","level":2,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tpds.2024.3511543","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2024.3511543","pdf_url":null,"source":{"id":"https://openalex.org/S97130795","display_name":"IEEE Transactions on Parallel and Distributed Systems","issn_l":"1045-9219","issn":["1045-9219","1558-2183","2161-9883"],"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 Parallel and Distributed Systems","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3537643952","display_name":null,"funder_award_id":"62350710797","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"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":31,"referenced_works":["https://openalex.org/W2034940213","https://openalex.org/W2183341477","https://openalex.org/W2194775991","https://openalex.org/W2416799949","https://openalex.org/W2804032941","https://openalex.org/W2962851801","https://openalex.org/W2981758446","https://openalex.org/W3011667710","https://openalex.org/W3049640275","https://openalex.org/W3212184265","https://openalex.org/W3217445637","https://openalex.org/W4308089797","https://openalex.org/W4312199008","https://openalex.org/W4320060064","https://openalex.org/W4372271902","https://openalex.org/W4385245566","https://openalex.org/W4395668837","https://openalex.org/W6637151318","https://openalex.org/W6684205842","https://openalex.org/W6687681856","https://openalex.org/W6695314431","https://openalex.org/W6738796088","https://openalex.org/W6755207826","https://openalex.org/W6762488536","https://openalex.org/W6766978945","https://openalex.org/W6775385080","https://openalex.org/W6775686901","https://openalex.org/W6785429063","https://openalex.org/W6790722028","https://openalex.org/W6801411073","https://openalex.org/W6809979497"],"related_works":["https://openalex.org/W2950520577","https://openalex.org/W1501159154","https://openalex.org/W2003935582","https://openalex.org/W74409296","https://openalex.org/W1554644772","https://openalex.org/W2494130044","https://openalex.org/W2468095077","https://openalex.org/W29548032","https://openalex.org/W2593878938","https://openalex.org/W305742777"],"abstract_inverted_index":{"Current":[0],"deep":[1],"learning":[2,110],"compilers":[3],"have":[4],"made":[5],"significant":[6],"strides":[7],"in":[8,156,188],"optimizing":[9],"computation":[10,51,130],"graphs":[11],"for":[12,22,38,65,78,191],"single-":[13,193],"and":[14,119,129,143,194],"multi-model":[15,195],"scenarios.":[16,198],"However,":[17],"they":[18],"lack":[19],"specific":[20],"optimizations":[21],"asynchronous":[23],"multi-task":[24],"inference":[25,36,100],"systems.":[26],"In":[27],"such":[28],"systems,":[29],"tasks":[30],"arrive":[31],"dynamically,":[32],"leading":[33],"to":[34,113,137,147],"diverse":[35],"progress":[37,101],"each":[39,103],"model.":[40,72,104],"This":[41],"renders":[42],"traditional":[43],"optimization":[44,154],"strategies":[45],"based":[46],"solely":[47],"on":[48,98,175],"the":[49,70,92,99,107,116,153,166,180,186,206],"original":[50],"graph":[52],"suboptimal":[53],"or":[54,216],"even":[55],"invalid.":[56],"Furthermore,":[57],"existing":[58],"operator":[59,189],"scheduling":[60,190],"methods":[61],"do":[62],"not":[63],"account":[64],"parallel":[66],"task":[67,95,117,149,196],"pipelines":[68,74],"involving":[69],"same":[71],"Task":[73],"present":[75],"additional":[76],"opportunities":[77],"optimization.":[79],"Therefore,":[80],"we":[81,158,178,202],"propose":[82],"Task-based":[83],"Operator":[84],"Parallelism":[85],"(TOP).":[86],"TOP":[87,135],"incorporates":[88],"an":[89,123],"understanding":[90],"of":[91,94,102,126,134,182,208],"impact":[93],"arrival":[96],"patterns":[97],"It":[105,184],"leverages":[106],"multi-agent":[108],"reinforcement":[109],"algorithm":[111],"MADDPG":[112],"cooperatively":[114],"optimize":[115],"launcher":[118],"model":[120,211],"scheduler,":[121],"generating":[122],"optimal":[124],"pair":[125],"dequeue":[127],"frequency":[128],"graph.":[131],"The":[132],"objective":[133],"is":[136],"enhance":[138,205],"resource":[139],"utilization,":[140],"increase":[141],"throughput,":[142],"allocate":[144],"resources":[145],"judiciously":[146],"prevent":[148],"backlog.":[150],"To":[151],"expedite":[152],"process":[155],"TOP,":[157,201],"introduce":[159],"a":[160,209],"novel":[161],"stage":[162],"partition":[163],"method":[164],"using":[165],"GNN-based":[167],"Policy":[168],"Gradient":[169],"(GPG)":[170],"algorithm.":[171],"Through":[172],"extensive":[173],"experiments":[174],"various":[176],"devices,":[177],"demonstrate":[179],"efficacy":[181],"TOP.":[183],"outperforms":[185],"state-of-the-art":[187],"both":[192],"processing":[197],"Benefiting":[199],"from":[200],"can":[203],"significantly":[204],"throughput":[207],"single":[210],"by":[212],"increasing":[213],"its":[214],"concurrency":[215],"batch":[217],"size,":[218],"thereby":[219],"achieving":[220],"self-acceleration.":[221]},"counts_by_year":[{"year":2025,"cited_by_count":3}],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2025-10-10T00:00:00"}