{"id":"https://openalex.org/W4400770592","doi":"https://doi.org/10.1109/tpds.2024.3430063","title":"InSS: An Intelligent Scheduling Orchestrator for Multi-GPU Inference With Spatio-Temporal Sharing","display_name":"InSS: An Intelligent Scheduling Orchestrator for Multi-GPU Inference With Spatio-Temporal Sharing","publication_year":2024,"publication_date":"2024-07-18","ids":{"openalex":"https://openalex.org/W4400770592","doi":"https://doi.org/10.1109/tpds.2024.3430063"},"language":"en","primary_location":{"id":"doi:10.1109/tpds.2024.3430063","doi":"https://doi.org/10.1109/tpds.2024.3430063","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2024.3430063","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,"listed_in":["ccf-a","cwts-core","dongbi-a","jufo-2","norway-2","poland-140","russia-white-list-1","vabb-shw"],"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"},"parent_works":[],"lineage":["https://openalex.org/W4400770592"],"child_works_count":0,"type":"article","is_traditional":true,"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/A5102132211","display_name":"Ziyi Han","orcid":"https://orcid.org/0009-0001-9456-9948","observed_orcids":["https://orcid.org/0009-0001-9456-9948"]},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ziyi Han","raw_affiliation_strings":["School of Cyber Science and Engineering, Wuhan University, Wuhan, China"],"raw_orcid":"https://orcid.org/0009-0001-9456-9948","affiliations":[{"raw_affiliation_string":"School of Cyber Science and Engineering, Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024396134","display_name":"Ruiting Zhou","orcid":"https://orcid.org/0000-0001-9681-6482","observed_orcids":["https://orcid.org/0000-0001-9681-6482","https://orcid.org/0000-0002-2136-5389"]},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ruiting Zhou","raw_affiliation_strings":["School of Cyber Science and Engineering, Wuhan University, Wuhan, China"],"raw_orcid":"https://orcid.org/0000-0001-9681-6482","affiliations":[{"raw_affiliation_string":"School of Cyber Science and Engineering, Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5012773300","display_name":"Chengzhong Xu","orcid":"https://orcid.org/0000-0001-9480-0356","observed_orcids":["https://orcid.org/0000-0001-9480-0356"]},"institutions":[{"id":"https://openalex.org/I204512498","display_name":"University of Macau","ror":"https://ror.org/01r4q9n85","country_code":"MO","type":"education","lineage":["https://openalex.org/I204512498"]}],"countries":["MO"],"is_corresponding":false,"raw_author_name":"Chengzhong Xu","raw_affiliation_strings":["Faculty of Science and Technology, University of Macau, Taipa, China"],"raw_orcid":"https://orcid.org/0000-0001-9480-0356","affiliations":[{"raw_affiliation_string":"Faculty of Science and Technology, University of Macau, Taipa, China","institution_ids":["https://openalex.org/I204512498"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014761841","display_name":"Yifan Zeng","orcid":"https://orcid.org/0000-0001-9954-5614","observed_orcids":["https://orcid.org/0000-0001-9954-5614"]},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yifan Zeng","raw_affiliation_strings":["School of Cyber Science and Engineering, Wuhan University, Wuhan, China"],"raw_orcid":"https://orcid.org/0000-0001-9954-5614","affiliations":[{"raw_affiliation_string":"School of Cyber Science and Engineering, Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5030661958","display_name":"Renli Zhang","orcid":"https://orcid.org/0000-0003-1727-0309","observed_orcids":["https://orcid.org/0000-0003-1727-0309"]},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Renli Zhang","raw_affiliation_strings":["School of Cyber Science and Engineering, Wuhan University, Wuhan, China"],"raw_orcid":"https://orcid.org/0000-0003-1727-0309","affiliations":[{"raw_affiliation_string":"School of Cyber Science and Engineering, Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":{"value":2495,"currency":"USD","value_usd":2495},"apc_paid":null,"fwci":7.4589,"has_fulltext":false,"cited_by_count":32,"citation_normalized_percentile":{"value":0.98205181,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":100},"biblio":{"volume":"35","issue":"10","first_page":"1735","last_page":"1748"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.6240000128746033,"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":0.6240000128746033,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.21789999306201935,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.09860000014305115,"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/gpu-resource-allocation","display_name":"GPU resource allocation","score":0.6589999794960022},{"id":"https://openalex.org/keywords/model-placement","display_name":"model placement","score":0.7960000038146973},{"id":"https://openalex.org/keywords/service-level-objectives","display_name":"service level objectives","score":0.734000027179718},{"id":"https://openalex.org/keywords/inference-throughput","display_name":"inference throughput","score":0.6880000233650208},{"id":"https://openalex.org/keywords/nvidia-a100","display_name":"NVIDIA A100","score":0.8230000138282776}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8826082944869995},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.684941291809082},{"id":"https://openalex.org/C2984822820","wikidata":"https://www.wikidata.org/wiki/Q1123036","display_name":"Processor scheduling","level":3,"score":0.6300029158592224},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.6009069681167603},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5873140692710876},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5502472519874573},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.38516145944595337},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.21105247735977173},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.1330455243587494},{"id":"https://openalex.org/C68387754","wikidata":"https://www.wikidata.org/wiki/Q7271585","display_name":"Schedule","level":2,"score":0.10186690092086792},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/tpds.2024.3430063","doi":"https://doi.org/10.1109/tpds.2024.3430063","is_oa":false,"landing_page_url":"https://doi.org/10.1109/tpds.2024.3430063","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,"listed_in":["ccf-a","cwts-core","dongbi-a","jufo-2","norway-2","poland-140","russia-white-list-1","vabb-shw"],"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":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G36496830","display_name":null,"funder_award_id":"62232004","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G3762208358","display_name":"\u5206\u5e03\u5f0f\u673a\u5668\u5b66\u4e60\u4efb\u52a1\u52a8\u6001\u5b9a\u4ef7\u4e0e\u90e8\u7f72\u5173\u952e\u6280\u672f\u7684\u7814\u7a76","funder_award_id":"62072344","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G8024709401","display_name":null,"funder_award_id":"U20A20177","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":55,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W2017977879","https://openalex.org/W2108598243","https://openalex.org/W2513730816","https://openalex.org/W2546571074","https://openalex.org/W2549139847","https://openalex.org/W2612445135","https://openalex.org/W2678047256","https://openalex.org/W2749928749","https://openalex.org/W2787938642","https://openalex.org/W2798291715","https://openalex.org/W2886953980","https://openalex.org/W2897417898","https://openalex.org/W2906773779","https://openalex.org/W2941327948","https://openalex.org/W2944542720","https://openalex.org/W2946948417","https://openalex.org/W2956461999","https://openalex.org/W2965757358","https://openalex.org/W2968986602","https://openalex.org/W2981037657","https://openalex.org/W2982157693","https://openalex.org/W2982356264","https://openalex.org/W2996600026","https://openalex.org/W3005664618","https://openalex.org/W3022200715","https://openalex.org/W3033527233","https://openalex.org/W3097411828","https://openalex.org/W3104481311","https://openalex.org/W3130571442","https://openalex.org/W3130689885","https://openalex.org/W3200486361","https://openalex.org/W3208777667","https://openalex.org/W3209166877","https://openalex.org/W3209811848","https://openalex.org/W3214476430","https://openalex.org/W4206943884","https://openalex.org/W4214690606","https://openalex.org/W4221141427","https://openalex.org/W4226203901","https://openalex.org/W4289828024","https://openalex.org/W4292240644","https://openalex.org/W4293025059","https://openalex.org/W4313229743","https://openalex.org/W4320005538","https://openalex.org/W4323897035","https://openalex.org/W4360831842","https://openalex.org/W4362653492","https://openalex.org/W4366564134","https://openalex.org/W4377235497","https://openalex.org/W4386436407","https://openalex.org/W4395112660","https://openalex.org/W6756009870","https://openalex.org/W6785197036","https://openalex.org/W6802095294"],"related_works":["https://openalex.org/W4402592509","https://openalex.org/W7037034321","https://openalex.org/W4403182945","https://openalex.org/W7117536227","https://openalex.org/W3198738574","https://openalex.org/W2906773779","https://openalex.org/W4313229743","https://openalex.org/W7113756933","https://openalex.org/W4391952397","https://openalex.org/W4413436944"],"abstract_inverted_index":{"As":[0],"the":[1,11,22,44,70,75,85,103,117,130,143,158,166,176],"applications":[2],"of":[3,13,25,77],"AI":[4],"proliferate,":[5],"it":[6,31],"is":[7,125],"critical":[8],"to":[9,61,68,101,127,162,165],"increase":[10],"throughput":[12,159],"online":[14],"DNN":[15,40],"inference":[16,52,95],"services.":[17],"Multi-process":[18],"service":[19],"(MPS)":[20],"improves":[21],"utilization":[23],"rate":[24],"GPU":[26,46,133,168],"resources":[27],"by":[28,141,160],"spatial-sharing,":[29],"but":[30],"also":[32],"brings":[33],"unique":[34],"challenges.":[35],"First,":[36],"interference":[37],"between":[38],"co-located":[39],"models":[41],"deployed":[42],"on":[43,150],"same":[45],"must":[47],"be":[48,62],"accurately":[49],"modeled.":[50],"Second,":[51],"tasks":[53],"arrive":[54],"dynamically":[55],"online,":[56],"and":[57,136],"each":[58],"task":[59,118],"needs":[60],"served":[63],"within":[64],"a":[65],"bounded":[66],"time":[67],"meet":[69],"service-level":[71],"objective":[72],"(SLO).":[73],"Third,":[74],"problem":[76],"fragments":[78],"has":[79],"become":[80],"more":[81],"serious.":[82],"To":[83],"address":[84],"above":[86],"three":[87],"challenges,":[88],"we":[89],"propose":[90],"anIntelligentScheduling":[91],"orchestrator":[92],"for":[93],"multi-GPU":[94],"servers":[96],"with":[97],"spatio-temporalSharing":[98],"(InSS),":[99],"aiming":[100],"maximize":[102],"system":[104],"throughput.InSSexploits":[105],"two":[106],"key":[107],"innovations:":[108],"i)":[109],"An":[110],"interference-aware":[111],"latency":[112,144],"analytical":[113,145],"model":[114,131],"which":[115],"estimates":[116],"latency.":[119],"ii)":[120],"A":[121],"two-stage":[122],"intelligent":[123],"scheduler":[124],"tailored":[126],"jointly":[128],"optimize":[129],"placement,":[132],"resource":[134],"allocation":[135],"adaptively":[137],"decides":[138],"batch":[139],"size":[140],"coupling":[142],"model.":[146],"Our":[147],"prototype":[148],"implementation":[149],"four":[151],"NVIDIA":[152],"A100":[153],"GPUs":[154],"shows":[155],"thatInSScan":[156],"improve":[157],"up":[161],"86%":[163],"compared":[164],"state-of-the-art":[167],"schedulers,":[169],"while":[170],"satisfying":[171],"SLOs.":[172],"We":[173],"further":[174],"show":[175],"scalability":[177],"ofInSSon":[178],"64":[179],"GPUs.":[180]},"counts_by_year":[{"year":2026,"cited_by_count":18},{"year":2025,"cited_by_count":10},{"year":2024,"cited_by_count":4}],"updated_date":"2026-11-10T11:39:43.021295","created_date":"2024-07-19T00:00:00"}