{"id":"https://openalex.org/W4407218580","doi":"https://doi.org/10.1145/3669940.3707251","title":"Dilu: Enabling GPU Resourcing-on-Demand for Serverless DL Serving via Introspective Elasticity","display_name":"Dilu: Enabling GPU Resourcing-on-Demand for Serverless DL Serving via Introspective Elasticity","publication_year":2025,"publication_date":"2025-02-06","ids":{"openalex":"https://openalex.org/W4407218580","doi":"https://doi.org/10.1145/3669940.3707251"},"language":"en","primary_location":{"id":"doi:10.1145/3669940.3707251","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3669940.3707251","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 1","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/A5071437610","display_name":"Cunchi Lv","orcid":"https://orcid.org/0009-0001-7089-0315"},"institutions":[{"id":"https://openalex.org/I4210101410","display_name":"International Centre for Theoretical Physics Asia-Pacific","ror":"https://ror.org/01z2px678","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210101410","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Cunchi Lv","raw_affiliation_strings":["ICT, CAS, Beijing, China and UCAS, Beijing, China"],"affiliations":[{"raw_affiliation_string":"ICT, CAS, Beijing, China and UCAS, Beijing, China","institution_ids":["https://openalex.org/I4210101410"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101653986","display_name":"Xiao Shi","orcid":"https://orcid.org/0000-0001-7105-8355"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao Shi","raw_affiliation_strings":["ICT, CAS, Beijing, China and Nanjing Institute of InforSuperBahn, Nanjing, China"],"affiliations":[{"raw_affiliation_string":"ICT, CAS, Beijing, China and Nanjing Institute of InforSuperBahn, Nanjing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000246945","display_name":"Zhengyu Lei","orcid":"https://orcid.org/0009-0004-2170-6560"},"institutions":[{"id":"https://openalex.org/I4210101410","display_name":"International Centre for Theoretical Physics Asia-Pacific","ror":"https://ror.org/01z2px678","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210101410","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhengyu Lei","raw_affiliation_strings":["ICT, CAS, Beijing, China and UCAS, Beijing, China"],"affiliations":[{"raw_affiliation_string":"ICT, CAS, Beijing, China and UCAS, Beijing, China","institution_ids":["https://openalex.org/I4210101410"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111358260","display_name":"Jinyue Huang","orcid":"https://orcid.org/0009-0004-9015-777X"},"institutions":[{"id":"https://openalex.org/I4210101410","display_name":"International Centre for Theoretical Physics Asia-Pacific","ror":"https://ror.org/01z2px678","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210101410","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jinyue Huang","raw_affiliation_strings":["ICT, CAS, Beijng, China and UCAS, Beijing, China"],"affiliations":[{"raw_affiliation_string":"ICT, CAS, Beijng, China and UCAS, Beijing, China","institution_ids":["https://openalex.org/I4210101410"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5077957152","display_name":"Wenting Tan","orcid":null},"institutions":[{"id":"https://openalex.org/I4210101410","display_name":"International Centre for Theoretical Physics Asia-Pacific","ror":"https://ror.org/01z2px678","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210101410","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wenting Tan","raw_affiliation_strings":["ICT, CAS, Beijing, China"],"affiliations":[{"raw_affiliation_string":"ICT, CAS, Beijing, China","institution_ids":["https://openalex.org/I4210101410"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113679400","display_name":"Xingfei Zheng","orcid":null},"institutions":[{"id":"https://openalex.org/I4210101410","display_name":"International Centre for Theoretical Physics Asia-Pacific","ror":"https://ror.org/01z2px678","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210101410","https://openalex.org/I4210165038"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaohui Zheng","raw_affiliation_strings":["ICT, CAS, Beijing, China"],"affiliations":[{"raw_affiliation_string":"ICT, CAS, Beijing, China","institution_ids":["https://openalex.org/I4210101410"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5110739105","display_name":"Xiaofang Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiaofang Zhao","raw_affiliation_strings":["ICT, CAS, Beijing, China and IICT, Suzhou, CAS, Suzhou, China"],"affiliations":[{"raw_affiliation_string":"ICT, CAS, Beijing, China and IICT, Suzhou, CAS, Suzhou, China","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5071437610"],"corresponding_institution_ids":["https://openalex.org/I4210101410"],"apc_list":null,"apc_paid":null,"fwci":5.955,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.95438059,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"311","last_page":"325"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10772","display_name":"Distributed systems and fault tolerance","score":0.9965000152587891,"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"}},"topics":[{"id":"https://openalex.org/T10772","display_name":"Distributed systems and fault tolerance","score":0.9965000152587891,"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/T10273","display_name":"IoT and Edge/Fog Computing","score":0.9943000078201294,"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.9939000010490417,"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/introspection","display_name":"Introspection","score":0.7733227014541626},{"id":"https://openalex.org/keywords/elasticity","display_name":"Elasticity (physics)","score":0.6716442704200745},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5351340174674988},{"id":"https://openalex.org/keywords/on-demand","display_name":"On demand","score":0.4695499539375305},{"id":"https://openalex.org/keywords/materials-science","display_name":"Materials science","score":0.27116087079048157},{"id":"https://openalex.org/keywords/composite-material","display_name":"Composite material","score":0.119730144739151},{"id":"https://openalex.org/keywords/multimedia","display_name":"Multimedia","score":0.11203256249427795},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.05498066544532776}],"concepts":[{"id":"https://openalex.org/C129671850","wikidata":"https://www.wikidata.org/wiki/Q210501","display_name":"Introspection","level":2,"score":0.7733227014541626},{"id":"https://openalex.org/C121854251","wikidata":"https://www.wikidata.org/wiki/Q62932","display_name":"Elasticity (physics)","level":2,"score":0.6716442704200745},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5351340174674988},{"id":"https://openalex.org/C2983523559","wikidata":"https://www.wikidata.org/wiki/Q410657","display_name":"On demand","level":2,"score":0.4695499539375305},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.27116087079048157},{"id":"https://openalex.org/C159985019","wikidata":"https://www.wikidata.org/wiki/Q181790","display_name":"Composite material","level":1,"score":0.119730144739151},{"id":"https://openalex.org/C49774154","wikidata":"https://www.wikidata.org/wiki/Q131765","display_name":"Multimedia","level":1,"score":0.11203256249427795},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.05498066544532776},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3669940.3707251","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3669940.3707251","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 1","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy","score":0.4099999964237213}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":33,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W1686810756","https://openalex.org/W2194775991","https://openalex.org/W2734941459","https://openalex.org/W2750779823","https://openalex.org/W2896457183","https://openalex.org/W2918828872","https://openalex.org/W2919594608","https://openalex.org/W2956461999","https://openalex.org/W2987607480","https://openalex.org/W3003011076","https://openalex.org/W3010543330","https://openalex.org/W3037377931","https://openalex.org/W3080451848","https://openalex.org/W3097411828","https://openalex.org/W3121702752","https://openalex.org/W3130689885","https://openalex.org/W3159219445","https://openalex.org/W3164865299","https://openalex.org/W3204075005","https://openalex.org/W3204216247","https://openalex.org/W3209503812","https://openalex.org/W4214690606","https://openalex.org/W4214764640","https://openalex.org/W4294904165","https://openalex.org/W4311080316","https://openalex.org/W4318541537","https://openalex.org/W4386436407","https://openalex.org/W4391045969","https://openalex.org/W4394944658","https://openalex.org/W4400770592","https://openalex.org/W6760979735","https://openalex.org/W6784871562"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W4387404794","https://openalex.org/W2380479582","https://openalex.org/W4387405288","https://openalex.org/W4387404506","https://openalex.org/W2084234399","https://openalex.org/W4387405632","https://openalex.org/W2500774003"],"abstract_inverted_index":{"Serverless":[0],"computing,":[1],"with":[2,21,109,120,138,172],"its":[3],"ease":[4],"of":[5,50,153,168],"management,":[6],"auto-scaling,":[7],"and":[8,52,82,104,185,190],"cost-effectiveness,":[9],"is":[10,34],"widely":[11],"adopted":[12],"by":[13,63],"deep":[14],"learning":[15],"(DL)":[16],"applications.":[17],"DL":[18,44,93,107,118,170],"workloads,":[19],"especially":[20],"large":[22],"language":[23],"models,":[24],"require":[25],"substantial":[26],"GPU":[27,38,55,89,136,154,174,177],"resources":[28],"to":[29,36,47,87,128,134,149,182,198],"ensure":[30],"QoS.":[31],"However,":[32],"it":[33,162],"prone":[35],"produce":[37],"fragments":[39],"(e.g.,":[40],"15%-94%)":[41],"in":[42,132,156],"serverless":[43,64,69,92,106],"systems":[45,70],"due":[46],"the":[48,60,129,151,166,199],"dynamicity":[49],"workloads":[51],"coarse-grained":[53],"static":[54],"allocation":[56],"mechanisms,":[57],"gradually":[58],"eroding":[59],"profits":[61],"offered":[62],"elasticity.":[65],"Different":[66],"from":[67],"classical":[68],"that":[71,161],"only":[72],"scale":[73],"horizontally,":[74],"we":[75,99],"present":[76],"introspective":[77],"elasticity":[78,152],"(IE),":[79],"a":[80,102],"fine-grained":[81],"adaptive":[83,145],"two-dimensional":[84],"co-scaling":[85,147],"mechanism":[86],"support":[88],"resourcing-on-demand":[90],"for":[91,117],"tasks.":[94],"Based":[95],"on":[96],"this":[97],"insight,":[98],"build":[100],"Dilu,":[101],"cross-layer":[103],"GPU-based":[105],"system":[108],"IE":[110],"support.":[111],"First,":[112],"Dilu":[113,126,142],"provides":[114],"multi-factor":[115],"profiling":[116],"tasks":[119],"efficient":[121],"pruning":[122],"search":[123],"methods.":[124],"Second,":[125],"adheres":[127],"resourcing-complementary":[130],"principles":[131],"scheduling":[133],"improve":[135],"utilization":[137],"QoS":[139,191],"guarantees.":[140],"Third,":[141],"adopts":[143],"an":[144],"2D":[146],"method":[148],"enhance":[150],"provisioning":[155],"real":[157],"time.":[158],"Evaluations":[159],"show":[160],"can":[163],"dynamically":[164],"adjust":[165],"resourcing":[167],"various":[169],"functions":[171],"low":[173],"fragmentation":[175],"(10%-46%":[176],"defragmentation),":[178],"high":[179],"throughput":[180,188],"(up":[181],"1.8\u00d7":[183],"inference":[184],"1.1\u00d7":[186],"training":[187],"increment)":[189],"guarantees":[192],"(11%-71%":[193],"violation":[194],"rate":[195],"reduction),":[196],"compared":[197],"SOTA":[200],"baselines.":[201]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":2}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}
