{"id":"https://openalex.org/W4402805367","doi":"https://doi.org/10.1145/3688351.3689156","title":"Optimizing GPU Sharing for Container-Based DNN Serving with Multi-Instance GPUs","display_name":"Optimizing GPU Sharing for Container-Based DNN Serving with Multi-Instance GPUs","publication_year":2024,"publication_date":"2024-09-16","ids":{"openalex":"https://openalex.org/W4402805367","doi":"https://doi.org/10.1145/3688351.3689156"},"language":"en","primary_location":{"id":"doi:10.1145/3688351.3689156","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3688351.3689156","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 17th ACM International Systems and Storage Conference on ZZZ","raw_type":"proceedings-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/A5113399773","display_name":"X. J. Wei","orcid":null},"institutions":[{"id":"https://openalex.org/I130701444","display_name":"Georgia Institute of Technology","ror":"https://ror.org/01zkghx44","country_code":"US","type":"education","lineage":["https://openalex.org/I130701444"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Xinpeng Wei","raw_affiliation_strings":["Georgia Institute of Technology"],"affiliations":[{"raw_affiliation_string":"Georgia Institute of Technology","institution_ids":["https://openalex.org/I130701444"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103043568","display_name":"Zhichao Li","orcid":"https://orcid.org/0000-0001-7388-7341"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhichao Li","raw_affiliation_strings":["ByteDance Inc"],"affiliations":[{"raw_affiliation_string":"ByteDance Inc","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5102759092","display_name":"Cheng Tan","orcid":"https://orcid.org/0000-0002-1420-5125"},"institutions":[{"id":"https://openalex.org/I87182695","display_name":"Universidad del Noreste","ror":"https://ror.org/02ahky613","country_code":"MX","type":"education","lineage":["https://openalex.org/I87182695"]}],"countries":["MX"],"is_corresponding":false,"raw_author_name":"Cheng Tan","raw_affiliation_strings":["Northeastern University"],"affiliations":[{"raw_affiliation_string":"Northeastern University","institution_ids":["https://openalex.org/I87182695"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5113399773"],"corresponding_institution_ids":["https://openalex.org/I130701444"],"apc_list":null,"apc_paid":null,"fwci":1.1123,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.78256077,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"68","last_page":"82"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.998199999332428,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.998199999332428,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"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/T10714","display_name":"Software-Defined Networks and 5G","score":0.9977999925613403,"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"}},{"id":"https://openalex.org/T10080","display_name":"Energy Efficient Wireless Sensor Networks","score":0.9968000054359436,"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":0.8465659618377686},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.6448346972465515},{"id":"https://openalex.org/keywords/container","display_name":"Container (type theory)","score":0.5877231359481812},{"id":"https://openalex.org/keywords/gpu-cluster","display_name":"GPU cluster","score":0.5121912360191345},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.4163636863231659},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.3701949417591095},{"id":"https://openalex.org/keywords/materials-science","display_name":"Materials science","score":0.06062114238739014}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8465659618377686},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.6448346972465515},{"id":"https://openalex.org/C2781018962","wikidata":"https://www.wikidata.org/wiki/Q5164884","display_name":"Container (type theory)","level":2,"score":0.5877231359481812},{"id":"https://openalex.org/C2781335571","wikidata":"https://www.wikidata.org/wiki/Q2633544","display_name":"GPU cluster","level":3,"score":0.5121912360191345},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.4163636863231659},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.3701949417591095},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.06062114238739014},{"id":"https://openalex.org/C159985019","wikidata":"https://www.wikidata.org/wiki/Q181790","display_name":"Composite material","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3688351.3689156","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3688351.3689156","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 17th ACM International Systems and Storage Conference on ZZZ","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":33,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W749350555","https://openalex.org/W1999244290","https://openalex.org/W2027403233","https://openalex.org/W2082649910","https://openalex.org/W2084026635","https://openalex.org/W2149935414","https://openalex.org/W2407965972","https://openalex.org/W2500898769","https://openalex.org/W2531894595","https://openalex.org/W2565600385","https://openalex.org/W2734941459","https://openalex.org/W2899071864","https://openalex.org/W2917975220","https://openalex.org/W2919594608","https://openalex.org/W2975712713","https://openalex.org/W2982356264","https://openalex.org/W3022298203","https://openalex.org/W3029416730","https://openalex.org/W3033527233","https://openalex.org/W3041498984","https://openalex.org/W3048441649","https://openalex.org/W3083558459","https://openalex.org/W3095922049","https://openalex.org/W3096484587","https://openalex.org/W3097053732","https://openalex.org/W3097411828","https://openalex.org/W3203600045","https://openalex.org/W3210776666","https://openalex.org/W4255654664","https://openalex.org/W4256272899","https://openalex.org/W4400770592","https://openalex.org/W6750647073"],"related_works":["https://openalex.org/W3213381848","https://openalex.org/W2012954338","https://openalex.org/W2392023973","https://openalex.org/W2939411666","https://openalex.org/W1973046741","https://openalex.org/W2056717482","https://openalex.org/W2017587301","https://openalex.org/W2082176405","https://openalex.org/W2096672917","https://openalex.org/W2005148983"],"abstract_inverted_index":{"The":[0],"trade-off":[1],"of":[2],"throughput":[3,91],"versus":[4],"latency":[5],"is":[6],"important":[7],"in":[8,72],"serving":[9,35,57],"deep":[10],"neural":[11],"networks":[12],"(DNNs)":[13],"on":[14,59,96],"GPUs.":[15],"A":[16],"hardware":[17],"feature---Multi-Instance":[18],"GPU":[19,31],"(MIG)---provides":[20],"an":[21],"additional":[22],"dimension":[23],"to":[24,48],"this":[25],"trade-off.":[26],"This":[27],"paper":[28],"studies":[29],"the":[30,112],"sharing":[32],"problem":[33],"for":[34],"DNNs":[36],"with":[37],"MIG.":[38],"We":[39,93],"propose":[40],"a":[41,50,65,77,83],"system":[42],"Jormungandr":[43,63,95,102],"that":[44,68,81,101],"allocates":[45],"MIG-enabled":[46],"GPUs":[47,110],"optimize":[49],"user-define":[51],"utility":[52],"function,":[53],"and":[54],"places":[55],"DNN":[56],"containers":[58],"heterogeneous":[60],"MIG":[61],"instances.":[62],"introduces":[64],"utility-first":[66],"search":[67],"yields":[69],"efficient":[70],"solutions":[71],"practice.":[73],"Additionally,":[74],"it":[75],"presents":[76],"robust":[78],"transition":[79],"protocol":[80],"guarantees":[82],"seamless":[84],"switch":[85],"between":[86],"two":[87],"GPU-cluster":[88],"configurations,":[89],"minimizing":[90],"fluctuations.":[92],"implement":[94],"Kubernetes.":[97],"Our":[98],"experiments":[99],"show":[100],"provides":[103],"near-optimal":[104],"solutions,":[105],"using":[106],"<":[107],"6%":[108],"more":[109],"than":[111],"optimal":[113],"solutions.":[114]},"counts_by_year":[{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":1}],"updated_date":"2025-12-21T01:58:51.020947","created_date":"2025-10-10T00:00:00"}