{"id":"https://openalex.org/W4389628299","doi":"https://doi.org/10.1145/3626793","title":"Automated Backend Allocation for Multi-Model, On-Device AI Inference","display_name":"Automated Backend Allocation for Multi-Model, On-Device AI Inference","publication_year":2023,"publication_date":"2023-12-07","ids":{"openalex":"https://openalex.org/W4389628299","doi":"https://doi.org/10.1145/3626793"},"language":"en","primary_location":{"id":"doi:10.1145/3626793","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3626793","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3626793","source":{"id":"https://openalex.org/S4210193547","display_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","issn_l":"2476-1249","issn":["2476-1249"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"bronze","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3626793","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5042109946","display_name":"Venkatraman Iyer","orcid":null},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Venkatraman Iyer","raw_affiliation_strings":["Samsung Electronics, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0004-8707-3196","affiliations":[{"raw_affiliation_string":"Samsung Electronics, Seoul, South Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5010724789","display_name":"Sungho Lee","orcid":"https://orcid.org/0009-0005-3808-1592"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Sungho Lee","raw_affiliation_strings":["Samsung Electronics, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0005-3808-1592","affiliations":[{"raw_affiliation_string":"Samsung Electronics, Seoul, South Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5064421035","display_name":"Semun Lee","orcid":"https://orcid.org/0009-0001-3550-7958"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Semun Lee","raw_affiliation_strings":["Samsung Electronics, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0001-3550-7958","affiliations":[{"raw_affiliation_string":"Samsung Electronics, Seoul, South Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5003174140","display_name":"Juitem Joonwoo Kim","orcid":"https://orcid.org/0009-0008-5709-0875"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Juitem Joonwoo Kim","raw_affiliation_strings":["Samsung Electronics, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0008-5709-0875","affiliations":[{"raw_affiliation_string":"Samsung Electronics, Seoul, South Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102859295","display_name":"Hyunjun Kim","orcid":"https://orcid.org/0009-0001-3539-2679"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyunjun Kim","raw_affiliation_strings":["Samsung Electronics, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0001-3539-2679","affiliations":[{"raw_affiliation_string":"Samsung Electronics, Seoul, South Korea","institution_ids":["https://openalex.org/I2250650973"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5014882842","display_name":"Young-Jae Shin","orcid":"https://orcid.org/0009-0005-5749-2995"},"institutions":[{"id":"https://openalex.org/I2250650973","display_name":"Samsung (South Korea)","ror":"https://ror.org/04w3jy968","country_code":"KR","type":"company","lineage":["https://openalex.org/I2250650973"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Youngjae Shin","raw_affiliation_strings":["Samsung Electronics, Seoul, South Korea"],"raw_orcid":"https://orcid.org/0009-0005-5749-2995","affiliations":[{"raw_affiliation_string":"Samsung Electronics, Seoul, South Korea","institution_ids":["https://openalex.org/I2250650973"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5042109946"],"corresponding_institution_ids":["https://openalex.org/I2250650973"],"apc_list":null,"apc_paid":null,"fwci":0.683,"has_fulltext":true,"cited_by_count":6,"citation_normalized_percentile":{"value":0.72584955,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":95,"max":98},"biblio":{"volume":"7","issue":"3","first_page":"1","last_page":"33"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9984999895095825,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9984999895095825,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.9952999949455261,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9937000274658203,"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":0.8711705207824707},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6233739852905273},{"id":"https://openalex.org/keywords/pareto-principle","display_name":"Pareto principle","score":0.609748125076294},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.5190283060073853},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.4836955666542053},{"id":"https://openalex.org/keywords/computation-offloading","display_name":"Computation offloading","score":0.471241295337677},{"id":"https://openalex.org/keywords/delegate","display_name":"Delegate","score":0.46511390805244446},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.4224960505962372},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.33417046070098877},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.32862916588783264},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.32810232043266296},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.30352672934532166},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.23239180445671082},{"id":"https://openalex.org/keywords/cloud-computing","display_name":"Cloud computing","score":0.20959004759788513},{"id":"https://openalex.org/keywords/edge-computing","display_name":"Edge computing","score":0.10888153314590454},{"id":"https://openalex.org/keywords/mathematical-optimization","display_name":"Mathematical optimization","score":0.0902777910232544},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.08296653628349304}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8711705207824707},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6233739852905273},{"id":"https://openalex.org/C137635306","wikidata":"https://www.wikidata.org/wiki/Q182667","display_name":"Pareto principle","level":2,"score":0.609748125076294},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.5190283060073853},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4836955666542053},{"id":"https://openalex.org/C2781041963","wikidata":"https://www.wikidata.org/wiki/Q18348618","display_name":"Computation offloading","level":4,"score":0.471241295337677},{"id":"https://openalex.org/C143273055","wikidata":"https://www.wikidata.org/wiki/Q2382794","display_name":"Delegate","level":2,"score":0.46511390805244446},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.4224960505962372},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.33417046070098877},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.32862916588783264},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.32810232043266296},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.30352672934532166},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.23239180445671082},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.20959004759788513},{"id":"https://openalex.org/C2778456923","wikidata":"https://www.wikidata.org/wiki/Q5337692","display_name":"Edge computing","level":3,"score":0.10888153314590454},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.0902777910232544},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.08296653628349304},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C555944384","wikidata":"https://www.wikidata.org/wiki/Q249","display_name":"Wireless","level":2,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3626793","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3626793","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3626793","source":{"id":"https://openalex.org/S4210193547","display_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","issn_l":"2476-1249","issn":["2476-1249"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1145/3626793","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3626793","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3626793","source":{"id":"https://openalex.org/S4210193547","display_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","issn_l":"2476-1249","issn":["2476-1249"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","raw_type":"journal-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/9","score":0.44999998807907104,"display_name":"Industry, innovation and infrastructure"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4389628299.pdf","grobid_xml":"https://content.openalex.org/works/W4389628299.grobid-xml"},"referenced_works_count":54,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W1990912193","https://openalex.org/W2030782791","https://openalex.org/W2046868034","https://openalex.org/W2062368747","https://openalex.org/W2097643185","https://openalex.org/W2098907614","https://openalex.org/W2106334424","https://openalex.org/W2126105956","https://openalex.org/W2153190325","https://openalex.org/W2183341477","https://openalex.org/W2194775991","https://openalex.org/W2300242332","https://openalex.org/W2323693848","https://openalex.org/W2416799949","https://openalex.org/W2468875367","https://openalex.org/W2625157458","https://openalex.org/W2803550716","https://openalex.org/W2887327791","https://openalex.org/W2896225285","https://openalex.org/W2897268228","https://openalex.org/W2905692112","https://openalex.org/W2941630349","https://openalex.org/W2962725887","https://openalex.org/W2962773068","https://openalex.org/W2963122961","https://openalex.org/W2963918968","https://openalex.org/W2964217527","https://openalex.org/W2980856918","https://openalex.org/W3014810041","https://openalex.org/W3022536434","https://openalex.org/W3022539872","https://openalex.org/W3026458074","https://openalex.org/W3035424742","https://openalex.org/W3039050620","https://openalex.org/W3080624533","https://openalex.org/W3088076788","https://openalex.org/W3096858590","https://openalex.org/W3158233068","https://openalex.org/W3175471559","https://openalex.org/W3186289964","https://openalex.org/W3202465876","https://openalex.org/W3210626783","https://openalex.org/W4212986322","https://openalex.org/W4214668489","https://openalex.org/W4223975181","https://openalex.org/W4234552385","https://openalex.org/W4236099117","https://openalex.org/W4236786453","https://openalex.org/W4236853429","https://openalex.org/W4287750421","https://openalex.org/W4289305285","https://openalex.org/W4301104990","https://openalex.org/W4317928033"],"related_works":["https://openalex.org/W2031284701","https://openalex.org/W3141203889","https://openalex.org/W1969486489","https://openalex.org/W2418771911","https://openalex.org/W3140622617","https://openalex.org/W2994187804","https://openalex.org/W2078379271","https://openalex.org/W2043460294","https://openalex.org/W4210712718","https://openalex.org/W1905398909"],"abstract_inverted_index":{"On-Device":[0],"Artificial":[1],"Intelligence":[2],"(AI)":[3],"services":[4,25],"such":[5],"as":[6,136,138],"face":[7],"recognition,":[8],"object":[9],"tracking":[10],"and":[11,35,57,87,96,109,127,157,184,191],"voice":[12],"recognition":[13],"are":[14],"rapidly":[15],"scaling":[16],"up":[17],"deployments":[18],"on":[19,33,188,195],"embedded,":[20],"memory-constrained":[21],"hardware":[22],"devices.":[23],"These":[24],"typically":[26],"delegate":[27],"AI":[28,78],"inference":[29,107],"models":[30],"for":[31],"execution":[32],"CPU":[34,86],"GPU":[36,40,88,161,170],"computing":[37],"backends.":[38],"While":[39],"delegation":[41],"is":[42,94,117],"a":[43,70,83,99,118,149,159,164],"common":[44],"practice":[45],"to":[46,72,167],"achieve":[47],"high":[48],"speed":[49],"computation,":[50],"the":[51,173,180],"approach":[52],"suffers":[53],"from":[54],"degraded":[55],"throughput":[56,132],"completion":[58],"times":[59],"under":[60],"multi-model":[61,143],"scenarios,":[62],"i.e.":[63],"concurrently":[64],"executing":[65],"services.":[66],"This":[67],"paper":[68,116],"introduces":[69],"solution":[71],"sustain":[73],"performance":[74],"in":[75,131],"multi-model,":[76],"on-device":[77],"contexts":[79],"by":[80,98,186,192],"dynamically":[81],"allocating":[82],"combination":[84],"of":[85,101,114,179],"backends":[89],"per":[90,125],"model.":[91],"The":[92],"allocation":[93,120],"feedback-driven,":[95],"guided":[97],"knowledge":[100],"model-specific,":[102],"multi-objective":[103],"pareto":[104,151,174,189],"fronts":[105],"comprising":[106],"latency":[108],"memory":[110],"consumption.":[111],"Primary":[112],"contribution":[113],"this":[115],"backend":[119],"algorithm":[121,166],"that":[122,154],"runs":[123,155],"online":[124],"model,":[126],"achieves":[128],"25-100%":[129],"improvement":[130],"over":[133],"static":[134],"allocations":[135],"well":[137],"load-balancing":[139],"scheduler":[140],"solutions":[141],"targeting":[142],"scenarios.":[144],"Other":[145],"noteworthy":[146],"contributions":[147],"include":[148],"novel":[150],"front":[152,175],"estimator":[153,176],"on-device,":[156],"also":[158],"software-based":[160],"profiler":[162],"with":[163],"lightweight":[165],"detect":[168],"changing":[169],"workloads.":[171],"Specifically,":[172],"outperforms":[177],"state":[178],"art":[181],"algorithms":[182],"NSGA-II":[183],"SPEA2":[185],"94%":[187],"coverage,":[190],"almost":[193],"2x":[194],"computational":[196],"overhead.":[197]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":3}],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2025-10-10T00:00:00"}
