{"id":"https://openalex.org/W4411800481","doi":"https://doi.org/10.1109/ccgrid64434.2025.00040","title":"Draco: Dynamic Resource Allocation for Concurrent ML Applications","display_name":"Draco: Dynamic Resource Allocation for Concurrent ML Applications","publication_year":2025,"publication_date":"2025-05-19","ids":{"openalex":"https://openalex.org/W4411800481","doi":"https://doi.org/10.1109/ccgrid64434.2025.00040"},"language":"en","primary_location":{"id":"doi:10.1109/ccgrid64434.2025.00040","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ccgrid64434.2025.00040","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE 25th International Symposium on Cluster, Cloud and Internet Computing (CCGrid)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5118706986","display_name":"Theo Radig","orcid":null},"institutions":[{"id":"https://openalex.org/I143288331","display_name":"Hasso Plattner Institute","ror":"https://ror.org/058rn5r42","country_code":"DE","type":"facility","lineage":["https://openalex.org/I143288331","https://openalex.org/I176453806"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Theo Radig","raw_affiliation_strings":["Hasso-Plattner-Institute, Internet Technologies and Softwarization,Potsdam,Germany"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hasso-Plattner-Institute, Internet Technologies and Softwarization,Potsdam,Germany","institution_ids":["https://openalex.org/I143288331"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5019777818","display_name":"Holger Karl","orcid":"https://orcid.org/0000-0002-8343-6322"},"institutions":[{"id":"https://openalex.org/I143288331","display_name":"Hasso Plattner Institute","ror":"https://ror.org/058rn5r42","country_code":"DE","type":"facility","lineage":["https://openalex.org/I143288331","https://openalex.org/I176453806"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Holger Karl","raw_affiliation_strings":["Hasso-Plattner-Institute, Internet Technologies and Softwarization,Potsdam,Germany"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Hasso-Plattner-Institute, Internet Technologies and Softwarization,Potsdam,Germany","institution_ids":["https://openalex.org/I143288331"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I143288331"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.12559618,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"11","last_page":"20"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9994999766349792,"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":0.9994999766349792,"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.9952999949455261,"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.9952999949455261,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7008422017097473},{"id":"https://openalex.org/keywords/resource-allocation","display_name":"Resource allocation","score":0.6293368935585022},{"id":"https://openalex.org/keywords/resource-management","display_name":"Resource management (computing)","score":0.47834107279777527},{"id":"https://openalex.org/keywords/resource","display_name":"Resource (disambiguation)","score":0.4718969762325287},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.2569219470024109},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.13134145736694336}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7008422017097473},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.6293368935585022},{"id":"https://openalex.org/C2780609101","wikidata":"https://www.wikidata.org/wiki/Q17156588","display_name":"Resource management (computing)","level":2,"score":0.47834107279777527},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.4718969762325287},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.2569219470024109},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.13134145736694336}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ccgrid64434.2025.00040","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ccgrid64434.2025.00040","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE 25th International Symposium on Cluster, Cloud and Internet Computing (CCGrid)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.47999998927116394,"id":"https://metadata.un.org/sdg/8","display_name":"Decent work and economic growth"}],"awards":[{"id":"https://openalex.org/G7419989365","display_name":null,"funder_award_id":"01IS22092","funder_id":"https://openalex.org/F4320321114","funder_display_name":"Bundesministerium f\u00fcr Bildung und Forschung"}],"funders":[{"id":"https://openalex.org/F4320321114","display_name":"Bundesministerium f\u00fcr Bildung und Forschung","ror":"https://ror.org/04pz7b180"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":17,"referenced_works":["https://openalex.org/W2734941459","https://openalex.org/W2804032941","https://openalex.org/W2949650786","https://openalex.org/W2979041789","https://openalex.org/W3174529902","https://openalex.org/W4295312788","https://openalex.org/W4385623261","https://openalex.org/W4388855507","https://openalex.org/W4394944658","https://openalex.org/W4394969374","https://openalex.org/W4402855523","https://openalex.org/W4409248734","https://openalex.org/W6751349269","https://openalex.org/W6752943062","https://openalex.org/W6766978945","https://openalex.org/W6871042906","https://openalex.org/W6872772671"],"related_works":["https://openalex.org/W2923452570","https://openalex.org/W206598027","https://openalex.org/W2978610750","https://openalex.org/W2022931285","https://openalex.org/W1589966275","https://openalex.org/W2086872282","https://openalex.org/W2137789903","https://openalex.org/W2138781885","https://openalex.org/W4315630413","https://openalex.org/W2153007255"],"abstract_inverted_index":{"GPUs":[0],"are":[1,10,17,27,107,182],"increasingly":[2],"used":[3,11,67],"to":[4,12,43,84,115,173,176,184,204],"accelerate":[5],"ML":[6,127],"applications.":[7],"Often,":[8],"they":[9],"offload":[13],"inference":[14,25,37,70,101,118,132,142,174,209],"tasks,":[15],"which":[16],"usually":[18],"latency-critical.":[19],"But":[20],"the":[21,46,58,77,131,138,153,185,191,197],"arrival":[22],"patterns":[23],"of":[24,49,61,79,141,164,196],"requests":[26],"often":[28],"bursty":[29],"and":[30,81,89,144,169],"include":[31],"periods":[32,146],"without":[33,129],"any":[34],"load.":[35],"Furthermore,":[36,159],"tasks":[38,119,143,175],"may":[39],"not":[40],"be":[41],"able":[42],"fully":[44],"utilize":[45],"compute":[47,105],"resources":[48,106,181],"a":[50,62,113,121,162],"GPU,":[51],"even":[52],"with":[53,120,147],"larger":[54],"batch":[55,122,187,198],"sizes.":[56],"Consequently,":[57],"average":[59],"utilization":[60,88],"GPU":[63],"that":[64,94],"is":[65,72],"exclusively":[66],"for":[68],"an":[69,126],"service":[71],"low.":[73,211],"Industry":[74],"has":[75],"recognized":[76],"problem":[78],"underutilization":[80],"offers":[82],"solutions":[83,97,206],"co-locate":[85,116],"applications,":[86],"improving":[87],"cost-efficiency.":[90],"We":[91,110],"show,":[92],"however,":[93],"these":[95],"state-of-the-art":[96],"only":[98],"maintain":[99],"low":[100,148],"latency":[102,133,178,210],"when":[103],"their":[104],"significantly":[108],"overprovisioned.":[109],"propose":[111],"Draco,":[112],"system":[114],"latency-critical":[117],"job,":[123],"e.g.,":[124],"training":[125],"model,":[128],"violating":[130],"requirements.":[134,179],"Draco":[135,160,193],"autonomously":[136],"estimates":[137],"resource":[139],"requirements":[140],"detects":[145],"load":[149],"by":[150,200],"perodically":[151],"sampling":[152],"GPU's":[154],"performance":[155],"monitoring":[156],"unit":[157],"(PMU).":[158],"manages":[161],"pool":[163],"streaming":[165],"multiprocessor":[166],"(SM)":[167],"partitions":[168],"dynamically":[170],"assigns":[171],"them":[172],"meet":[177],"Leftover":[180],"granted":[183],"co-located":[186],"job.":[188],"Depending":[189],"on":[190],"workload,":[192],"increases":[194],"throughput":[195],"job":[199],"up":[201],"to$4":[202],"x$compared":[203],"industrial":[205],"while":[207],"keeping":[208]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}
