{"id":"https://openalex.org/W7158305544","doi":"https://doi.org/10.1145/3805621.3807643","title":"Bridging CPU and GPU Autoscaling for Cost-Efficient Inference Serving","display_name":"Bridging CPU and GPU Autoscaling for Cost-Efficient Inference Serving","publication_year":2026,"publication_date":"2026-04-27","ids":{"openalex":"https://openalex.org/W7158305544","doi":"https://doi.org/10.1145/3805621.3807643"},"language":null,"primary_location":{"id":"doi:10.1145/3805621.3807643","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805621.3807643","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Sixth European Workshop on Machine Learning and Systems","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3805621.3807643","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5091376011","display_name":"Mehran Salmani","orcid":"https://orcid.org/0009-0008-6362-0967"},"institutions":[{"id":"https://openalex.org/I119449181","display_name":"Technische Universit\u00e4t Ilmenau","ror":"https://ror.org/01weqhp73","country_code":"DE","type":"education","lineage":["https://openalex.org/I119449181"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Mehran Salmani","raw_affiliation_strings":["Computer Science and Automation, TU Ilmenau, Ilmenau, Germany"],"raw_orcid":"https://orcid.org/0009-0008-6362-0967","affiliations":[{"raw_affiliation_string":"Computer Science and Automation, TU Ilmenau, Ilmenau, Germany","institution_ids":["https://openalex.org/I119449181"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045974365","display_name":"Kamran Razavi","orcid":"https://orcid.org/0000-0002-3232-5657"},"institutions":[{"id":"https://openalex.org/I4210097355","display_name":"HeidelbergCement (Germany)","ror":"https://ror.org/01q1dy966","country_code":"DE","type":"company","lineage":["https://openalex.org/I4210097355"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Kamran Razavi","raw_affiliation_strings":["turbalance, Heidelberg, Germany"],"raw_orcid":"https://orcid.org/0000-0002-3232-5657","affiliations":[{"raw_affiliation_string":"turbalance, Heidelberg, Germany","institution_ids":["https://openalex.org/I4210097355"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5025996509","display_name":"Peter Amthor","orcid":"https://orcid.org/0000-0001-7711-4450"},"institutions":[{"id":"https://openalex.org/I119449181","display_name":"Technische Universit\u00e4t Ilmenau","ror":"https://ror.org/01weqhp73","country_code":"DE","type":"education","lineage":["https://openalex.org/I119449181"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Peter Amthor","raw_affiliation_strings":["Computer Science and Automation, TU Ilmenau, Ilmenau, Germany"],"raw_orcid":"https://orcid.org/0000-0001-7711-4450","affiliations":[{"raw_affiliation_string":"Computer Science and Automation, TU Ilmenau, Ilmenau, Germany","institution_ids":["https://openalex.org/I119449181"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5025082646","display_name":"Boris Koldehofe","orcid":"https://orcid.org/0000-0002-1588-2056"},"institutions":[{"id":"https://openalex.org/I161103922","display_name":"Philipps University of Marburg","ror":"https://ror.org/01rdrb571","country_code":"DE","type":"education","lineage":["https://openalex.org/I161103922"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Boris Koldehofe","raw_affiliation_strings":["Mathematics and Computer Science, Marburg University, Marburg, Germany"],"raw_orcid":"https://orcid.org/0000-0002-1588-2056","affiliations":[{"raw_affiliation_string":"Mathematics and Computer Science, Marburg University, Marburg, Germany","institution_ids":["https://openalex.org/I161103922"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.67550987,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"163","last_page":"170"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.5587000250816345,"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":0.5587000250816345,"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":0.18000000715255737,"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/T12127","display_name":"Software System Performance and Reliability","score":0.06430000066757202,"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/scalability","display_name":"Scalability","score":0.8151999711990356},{"id":"https://openalex.org/keywords/cloud-computing","display_name":"Cloud computing","score":0.7128000259399414},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6783999800682068},{"id":"https://openalex.org/keywords/bridging","display_name":"Bridging (networking)","score":0.6164000034332275},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.5835000276565552},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.5392000079154968},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.44350001215934753},{"id":"https://openalex.org/keywords/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.40049999952316284}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.866599977016449},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.8151999711990356},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.7128000259399414},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6783999800682068},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.6164000034332275},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.5835000276565552},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.5392000079154968},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.44350001215934753},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.42890000343322754},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.40049999952316284},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3781999945640564},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.37720000743865967},{"id":"https://openalex.org/C157742956","wikidata":"https://www.wikidata.org/wiki/Q3237776","display_name":"Frequency scaling","level":3,"score":0.3686999976634979},{"id":"https://openalex.org/C49154492","wikidata":"https://www.wikidata.org/wiki/Q5300","display_name":"Central processing unit","level":2,"score":0.3587999939918518},{"id":"https://openalex.org/C5119721","wikidata":"https://www.wikidata.org/wiki/Q220501","display_name":"Quality of service","level":2,"score":0.31850001215934753},{"id":"https://openalex.org/C2989134064","wikidata":"https://www.wikidata.org/wiki/Q288510","display_name":"Execution time","level":2,"score":0.31790000200271606},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.2996000051498413},{"id":"https://openalex.org/C48677424","wikidata":"https://www.wikidata.org/wiki/Q6888088","display_name":"Mode (computer interface)","level":2,"score":0.2766999900341034},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.263700008392334},{"id":"https://openalex.org/C2778160497","wikidata":"https://www.wikidata.org/wiki/Q869830","display_name":"Service-level agreement","level":3,"score":0.2619999945163727}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3805621.3807643","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805621.3807643","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Sixth European Workshop on Machine Learning and Systems","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3805621.3807643","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3805621.3807643","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Sixth European Workshop on Machine Learning and Systems","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":11,"referenced_works":["https://openalex.org/W2150871235","https://openalex.org/W3095488153","https://openalex.org/W3097411828","https://openalex.org/W3209166877","https://openalex.org/W4283727273","https://openalex.org/W4368353274","https://openalex.org/W4393904734","https://openalex.org/W4401408750","https://openalex.org/W4408844874","https://openalex.org/W4408846413","https://openalex.org/W4412477285"],"related_works":[],"abstract_inverted_index":{"Autoscaling":[0],"mechanisms":[1,21],"have":[2,22],"been":[3],"used":[4],"by":[5,35,44,116,200],"inference":[6,88,131],"serving":[7,89],"systems":[8],"to":[9,27,53,106,133,153,178,181,204],"maintain":[10],"service":[11,145],"quality":[12],"and":[13,47,50,69,99,108,129,170],"efficiently":[14],"use":[15],"cloud":[16],"resources":[17,94,125],"(CPUs/GPUs).":[18],"Horizontal":[19],"scaling-based":[20],"a":[23,101,118,162,182,191,210],"responsiveness":[24],"issue":[25],"due":[26],"cold-start":[28],"delays.":[29],"In-place":[30],"CPU":[31],"vertical":[32],"scaling":[33,110,152,186],"adopted":[34],"recent":[36],"works":[37],"avoids":[38],"cold-start;":[39],"however,":[40],"it":[41,173],"is":[42,51,74,82,146],"constrained":[43],"hardware":[45],"limits":[46],"model":[48],"parallelizability,":[49],"limited":[52],"CPUs,":[54],"which":[55],"are":[56],"efficient":[57],"for":[58,63,157],"low":[59],"workloads":[60,73,169],"but":[61],"cost-inefficient":[62],"large":[64],"workloads.":[65],"Thus,":[66],"combining":[67],"CPUs":[68,105],"GPUs":[70,107,111],"under":[71,166,190,209],"dynamic":[72,211],"worthwhile.":[75],"However,":[76],"enabling":[77],"responsive":[78],"switching":[79],"between":[80],"them":[81],"challenging.":[83],"We":[84,160],"propose":[85],"VCG,":[86],"an":[87],"system":[90],"that":[91,172],"utilizes":[92],"different":[93,168],"based":[95],"on":[96],"the":[97,113,134,142,198],"workload,":[98],"enables":[100],"smooth":[102],"transition":[103],"from":[104],"vertically":[109],"(adjusting":[112],"GPU":[114,123,188],"percentage)":[115],"forking":[117],"new":[119],"process":[120,136],"with":[121],"enough":[122],"compute":[124],"using":[126,184],"CUDA":[127],"MPS,":[128],"delivering":[130],"requests":[132],"child":[135],"through":[137],"inter-process":[138],"communication":[139],"pipes.":[140],"When":[141],"capacity":[143],"of":[144,164,187],"(nearly)":[147],"exhausted,":[148],"VCG":[149,165],"uses":[150],"horizontal":[151,185],"provide":[154],"scalability":[155],"headroom":[156],"future":[158],"steps.":[159],"evaluate":[161],"prototype":[163],"two":[167,205],"show":[171],"reduces":[174],"SLO":[175],"violations":[176],"up":[177],"8\u00d7":[179],"compared":[180,203],"baseline":[183],"instances":[189],"bursty":[192],"workload.":[193,213],"It":[194],"can":[195],"also":[196],"reduce":[197],"cost":[199],"around":[201],"60%":[202],"CPU-based":[206],"autoscaling":[207],"baselines":[208],"(non-bursty)":[212]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-04-30T00:00:00"}