{"id":"https://openalex.org/W4402896975","doi":"https://doi.org/10.1109/iwqos61813.2024.10682915","title":"HarmonyBatch: Batching multi-SLO DNN Inference with Heterogeneous Serverless Functions","display_name":"HarmonyBatch: Batching multi-SLO DNN Inference with Heterogeneous Serverless Functions","publication_year":2024,"publication_date":"2024-06-19","ids":{"openalex":"https://openalex.org/W4402896975","doi":"https://doi.org/10.1109/iwqos61813.2024.10682915"},"language":"en","primary_location":{"id":"doi:10.1109/iwqos61813.2024.10682915","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iwqos61813.2024.10682915","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE/ACM 32nd International Symposium on Quality of Service (IWQoS)","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/A5112120011","display_name":"Jiabin Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I66867065","display_name":"East China Normal University","ror":"https://ror.org/02n96ep67","country_code":"CN","type":"education","lineage":["https://openalex.org/I66867065"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Jiabin Chen","raw_affiliation_strings":["East China Normal University,Shanghai Key Laboratory of Multidimensional Information Processing"],"affiliations":[{"raw_affiliation_string":"East China Normal University,Shanghai Key Laboratory of Multidimensional Information Processing","institution_ids":["https://openalex.org/I66867065"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029626585","display_name":"Fei Xu","orcid":"https://orcid.org/0000-0003-1590-5323"},"institutions":[{"id":"https://openalex.org/I66867065","display_name":"East China Normal University","ror":"https://ror.org/02n96ep67","country_code":"CN","type":"education","lineage":["https://openalex.org/I66867065"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fei Xu","raw_affiliation_strings":["East China Normal University,Shanghai Key Laboratory of Multidimensional Information Processing"],"affiliations":[{"raw_affiliation_string":"East China Normal University,Shanghai Key Laboratory of Multidimensional Information Processing","institution_ids":["https://openalex.org/I66867065"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5024880975","display_name":"Yikun Gu","orcid":"https://orcid.org/0000-0002-7645-3386"},"institutions":[{"id":"https://openalex.org/I66867065","display_name":"East China Normal University","ror":"https://ror.org/02n96ep67","country_code":"CN","type":"education","lineage":["https://openalex.org/I66867065"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yikun Gu","raw_affiliation_strings":["East China Normal University,Shanghai Key Laboratory of Multidimensional Information Processing"],"affiliations":[{"raw_affiliation_string":"East China Normal University,Shanghai Key Laboratory of Multidimensional Information Processing","institution_ids":["https://openalex.org/I66867065"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100379262","display_name":"Li Chen","orcid":"https://orcid.org/0000-0002-5842-838X"},"institutions":[{"id":"https://openalex.org/I79516672","display_name":"University of Louisiana at Lafayette","ror":"https://ror.org/01x8rc503","country_code":"US","type":"education","lineage":["https://openalex.org/I2799628689","https://openalex.org/I79516672"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Li Chen","raw_affiliation_strings":["University of Louisiana at Lafayette"],"affiliations":[{"raw_affiliation_string":"University of Louisiana at Lafayette","institution_ids":["https://openalex.org/I79516672"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101728901","display_name":"Fangming Liu","orcid":"https://orcid.org/0000-0001-9224-961X"},"institutions":[{"id":"https://openalex.org/I4210136793","display_name":"Peng Cheng Laboratory","ror":"https://ror.org/03qdqbt06","country_code":"CN","type":"facility","lineage":["https://openalex.org/I4210136793"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Fangming Liu","raw_affiliation_strings":["Peng Cheng Laboratory"],"affiliations":[{"raw_affiliation_string":"Peng Cheng Laboratory","institution_ids":["https://openalex.org/I4210136793"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5112113742","display_name":"Zhi Zhou","orcid":"https://orcid.org/0000-0002-4042-1104"},"institutions":[{"id":"https://openalex.org/I157773358","display_name":"Sun Yat-sen University","ror":"https://ror.org/0064kty71","country_code":"CN","type":"education","lineage":["https://openalex.org/I157773358"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhi Zhou","raw_affiliation_strings":["Sun Yat-sen University"],"affiliations":[{"raw_affiliation_string":"Sun Yat-sen University","institution_ids":["https://openalex.org/I157773358"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5112120011"],"corresponding_institution_ids":["https://openalex.org/I66867065"],"apc_list":null,"apc_paid":null,"fwci":4.0007,"has_fulltext":false,"cited_by_count":11,"citation_normalized_percentile":{"value":0.94539559,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"10"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.906499981880188,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.906499981880188,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10400","display_name":"Network Security and Intrusion Detection","score":0.9049999713897705,"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":0.8145681619644165},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.760402262210846},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.38155707716941833}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8145681619644165},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.760402262210846},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.38155707716941833}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iwqos61813.2024.10682915","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iwqos61813.2024.10682915","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2024 IEEE/ACM 32nd International Symposium on Quality of Service (IWQoS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320309392","display_name":"Louisiana Board of Regents","ror":"https://ror.org/00jv89z46"},{"id":"https://openalex.org/F4320321885","display_name":"Science and Technology Commission of Shanghai Municipality","ror":"https://ror.org/03kt66j61"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":30,"referenced_works":["https://openalex.org/W1686810756","https://openalex.org/W2333210517","https://openalex.org/W2934208298","https://openalex.org/W3047528232","https://openalex.org/W3105227206","https://openalex.org/W3130689885","https://openalex.org/W3204075005","https://openalex.org/W3215002152","https://openalex.org/W4214690606","https://openalex.org/W4214764640","https://openalex.org/W4221167396","https://openalex.org/W4250589301","https://openalex.org/W4292779060","https://openalex.org/W4294904165","https://openalex.org/W4313229743","https://openalex.org/W4318541537","https://openalex.org/W4385437022","https://openalex.org/W4386436407","https://openalex.org/W4388040908","https://openalex.org/W4388041408","https://openalex.org/W4388041447","https://openalex.org/W4391272922","https://openalex.org/W4402896975","https://openalex.org/W6637373629","https://openalex.org/W6755207826","https://openalex.org/W6765484274","https://openalex.org/W6775201933","https://openalex.org/W6778883912","https://openalex.org/W6810265253","https://openalex.org/W6861569072"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052","https://openalex.org/W2382290278","https://openalex.org/W4395014643"],"abstract_inverted_index":{"Deep":[0],"Neural":[1],"Network":[2],"(DNN)":[3],"inference":[4,24,46,79,108,131,147,186,210,230,242],"on":[5,21,39,148,168,219],"serverless":[6,22,44,89,134,240],"functions":[7,105],"is":[8,56,71],"gaining":[9],"prominence":[10],"due":[11],"to":[12,124,180,193,239,251,254],"its":[13],"potential":[14,93],"for":[15,75,106,128,200],"substantial":[16],"budget":[17,196],"savings.":[18],"Existing":[19],"works":[20],"DNN":[23,45,78,86,107,130,146,185,229,241],"solely":[25],"optimize":[26],"batching":[27,64,76],"requests":[28,80,187],"from":[29],"one":[30],"application":[31,189,202],"with":[32,81,132,227],"a":[33,62,118,170,174,215],"single":[34],"Service":[35],"Level":[36],"Objective":[37],"(SLO)":[38],"CPU":[40,102,150],"functions.":[41,135],"However,":[42],"production":[43],"traces":[47],"indicate":[48],"that":[49,233],"the":[50,92,157,183,195,246,255],"request":[51,163],"arrival":[52,164],"rate":[53,165],"of":[54,98,145,197,209,217],"applications":[55],"surprisingly":[57],"low,":[58],"which":[59],"inevitably":[60],"causes":[61],"long":[63],"time":[65],"and":[66,95,103,142,151,162],"SLO":[67],"violations.":[68],"Hence,":[69],"there":[70],"an":[72,139],"urgent":[73],"need":[74],"multiple":[77],"diverse":[82,206],"SLOs":[83,208],"(i.e.,":[84,101],"multi-SLO":[85,129,184],"inference)":[87],"in":[88,178],"platforms.":[90],"Moreover,":[91],"performance":[94,127,141,207,238],"cost":[96,143,248],"benefits":[97],"deploying":[99],"heterogeneous":[100,133],"GPU)":[104],"have":[109,213],"received":[110],"scant":[111],"attention.In":[112],"this":[113],"paper,":[114],"we":[115,137,172],"present":[116],"HarmonyBatch,":[117],"cost-efficient":[119],"resource":[120],"provisioning":[121,199],"framework":[122],"designed":[123],"achieve":[125],"predictable":[126,237],"Specifically,":[136],"construct":[138],"analytical":[140],"model":[144],"both":[149],"GPU":[152,158],"functions,":[153],"by":[154,249],"explicitly":[155],"considering":[156],"time-slicing":[159],"scheduling":[160],"mechanism":[161],"distribution.":[166],"Based":[167],"such":[169],"model,":[171],"devise":[173],"two-stage":[175],"merging":[176],"strategy":[177],"HarmonyBatch":[179,218,234],"judiciously":[181],"batch":[182],"into":[188],"groups.":[190],"It":[191],"aims":[192],"minimize":[194],"function":[198],"each":[201],"group":[203],"while":[204,244],"guaranteeing":[205],"applications.":[211],"We":[212],"implemented":[214],"prototype":[216,225],"Alibaba":[220],"Cloud":[221],"Function":[222],"Compute.":[223],"Extensive":[224],"experiments":[226],"representative":[228],"workloads":[231,243],"demonstrate":[232],"can":[235],"provide":[236],"reducing":[245],"monetary":[247],"up":[250],"82.9%":[252],"compared":[253],"state-of-the-art":[256],"methods.":[257]},"counts_by_year":[{"year":2025,"cited_by_count":3},{"year":2024,"cited_by_count":8}],"updated_date":"2025-12-27T23:08:20.325037","created_date":"2025-10-10T00:00:00"}