{"id":"https://openalex.org/W4285122249","doi":"https://doi.org/10.1109/access.2022.3184692","title":"Scale-Train: A Scalable DNN Training Framework for a Heterogeneous GPU Cloud","display_name":"Scale-Train: A Scalable DNN Training Framework for a Heterogeneous GPU Cloud","publication_year":2022,"publication_date":"2022-01-01","ids":{"openalex":"https://openalex.org/W4285122249","doi":"https://doi.org/10.1109/access.2022.3184692"},"language":"en","primary_location":{"id":"doi:10.1109/access.2022.3184692","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2022.3184692","pdf_url":"https://ieeexplore.ieee.org/ielx7/6287639/6514899/09801853.pdf","source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://ieeexplore.ieee.org/ielx7/6287639/6514899/09801853.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5067270140","display_name":"Kyeonglok Kim","orcid":"https://orcid.org/0000-0002-9811-9899"},"institutions":[{"id":"https://openalex.org/I848706","display_name":"Sungkyunkwan University","ror":"https://ror.org/04q78tk20","country_code":"KR","type":"education","lineage":["https://openalex.org/I848706"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Kyeonglok Kim","raw_affiliation_strings":["Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0002-9811-9899","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea","institution_ids":["https://openalex.org/I848706"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101878270","display_name":"Hyeonsu Lee","orcid":"https://orcid.org/0000-0003-3811-2449"},"institutions":[{"id":"https://openalex.org/I848706","display_name":"Sungkyunkwan University","ror":"https://ror.org/04q78tk20","country_code":"KR","type":"education","lineage":["https://openalex.org/I848706"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Hyeonsu Lee","raw_affiliation_strings":["Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea","institution_ids":["https://openalex.org/I848706"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5022036931","display_name":"Seungmin Oh","orcid":"https://orcid.org/0000-0002-2139-8746"},"institutions":[{"id":"https://openalex.org/I848706","display_name":"Sungkyunkwan University","ror":"https://ror.org/04q78tk20","country_code":"KR","type":"education","lineage":["https://openalex.org/I848706"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Seungmin Oh","raw_affiliation_strings":["Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea","institution_ids":["https://openalex.org/I848706"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5062665275","display_name":"Euiseong Seo","orcid":"https://orcid.org/0000-0003-2103-8019"},"institutions":[{"id":"https://openalex.org/I848706","display_name":"Sungkyunkwan University","ror":"https://ror.org/04q78tk20","country_code":"KR","type":"education","lineage":["https://openalex.org/I848706"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Euiseong Seo","raw_affiliation_strings":["Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea"],"raw_orcid":"https://orcid.org/0000-0003-2103-8019","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Sungkyunkwan University, Suwon-si, Republic of Korea","institution_ids":["https://openalex.org/I848706"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I848706"],"apc_list":{"value":1850,"currency":"USD","value_usd":1850},"apc_paid":{"value":1850,"currency":"USD","value_usd":1850},"fwci":0.243,"has_fulltext":true,"cited_by_count":4,"citation_normalized_percentile":{"value":0.45590923,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":97},"biblio":{"volume":"10","issue":null,"first_page":"68468","last_page":"68481"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9998999834060669,"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":0.9998999834060669,"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/T12702","display_name":"Brain Tumor Detection and Classification","score":0.9950000047683716,"subfield":{"id":"https://openalex.org/subfields/2808","display_name":"Neurology"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}},{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9933000206947327,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8651986122131348},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.8225235939025879},{"id":"https://openalex.org/keywords/cloud-computing","display_name":"Cloud computing","score":0.7658251523971558},{"id":"https://openalex.org/keywords/gpu-cluster","display_name":"GPU cluster","score":0.6573882699012756},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.6107056140899658},{"id":"https://openalex.org/keywords/graphics-processing-unit","display_name":"Graphics processing unit","score":0.6013869047164917},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.5863942503929138},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.5824731588363647},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.48180025815963745},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.4724229574203491},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.4400704503059387},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3844909369945526},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.261687308549881}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8651986122131348},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.8225235939025879},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.7658251523971558},{"id":"https://openalex.org/C2781335571","wikidata":"https://www.wikidata.org/wiki/Q2633544","display_name":"GPU cluster","level":3,"score":0.6573882699012756},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.6107056140899658},{"id":"https://openalex.org/C2779851693","wikidata":"https://www.wikidata.org/wiki/Q183484","display_name":"Graphics processing unit","level":2,"score":0.6013869047164917},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.5863942503929138},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.5824731588363647},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.48180025815963745},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.4724229574203491},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.4400704503059387},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3844909369945526},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.261687308549881},{"id":"https://openalex.org/C555944384","wikidata":"https://www.wikidata.org/wiki/Q249","display_name":"Wireless","level":2,"score":0.0},{"id":"https://openalex.org/C153294291","wikidata":"https://www.wikidata.org/wiki/Q25261","display_name":"Meteorology","level":1,"score":0.0},{"id":"https://openalex.org/C121332964","wikidata":"https://www.wikidata.org/wiki/Q413","display_name":"Physics","level":0,"score":0.0},{"id":"https://openalex.org/C62520636","wikidata":"https://www.wikidata.org/wiki/Q944","display_name":"Quantum mechanics","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/access.2022.3184692","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2022.3184692","pdf_url":"https://ieeexplore.ieee.org/ielx7/6287639/6514899/09801853.pdf","source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},{"id":"pmh:oai:doaj.org/article:7326cd25cd6640a9bd6db97fe8a5b224","is_oa":true,"landing_page_url":"https://doaj.org/article/7326cd25cd6640a9bd6db97fe8a5b224","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"IEEE Access, Vol 10, Pp 68468-68481 (2022)","raw_type":"article"}],"best_oa_location":{"id":"doi:10.1109/access.2022.3184692","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2022.3184692","pdf_url":"https://ieeexplore.ieee.org/ielx7/6287639/6514899/09801853.pdf","source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4285122249.pdf","grobid_xml":"https://content.openalex.org/works/W4285122249.grobid-xml"},"referenced_works_count":41,"referenced_works":["https://openalex.org/W2083842231","https://openalex.org/W2108598243","https://openalex.org/W2186615578","https://openalex.org/W2194775991","https://openalex.org/W2271840356","https://openalex.org/W2339765813","https://openalex.org/W2618530766","https://openalex.org/W2622263826","https://openalex.org/W2755443794","https://openalex.org/W2792327481","https://openalex.org/W2798515322","https://openalex.org/W2936343019","https://openalex.org/W2963149401","https://openalex.org/W2963964896","https://openalex.org/W2964350391","https://openalex.org/W3010045285","https://openalex.org/W3011751313","https://openalex.org/W3022298203","https://openalex.org/W3043004181","https://openalex.org/W3134991928","https://openalex.org/W4288079579","https://openalex.org/W4288115451","https://openalex.org/W4301239768","https://openalex.org/W6637373629","https://openalex.org/W6679393576","https://openalex.org/W6684859321","https://openalex.org/W6686509673","https://openalex.org/W6694517276","https://openalex.org/W6703420464","https://openalex.org/W6739622702","https://openalex.org/W6742061381","https://openalex.org/W6745458143","https://openalex.org/W6748645090","https://openalex.org/W6749102460","https://openalex.org/W6756009870","https://openalex.org/W6758283263","https://openalex.org/W6759814162","https://openalex.org/W6766978945","https://openalex.org/W6767421740","https://openalex.org/W6784425352","https://openalex.org/W6784836957"],"related_works":["https://openalex.org/W2119534391","https://openalex.org/W2078379271","https://openalex.org/W2023770367","https://openalex.org/W2503137108","https://openalex.org/W2043460294","https://openalex.org/W1966576946","https://openalex.org/W2020279179","https://openalex.org/W2056717482","https://openalex.org/W2030707850","https://openalex.org/W2393490604"],"abstract_inverted_index":{"In":[0,101,149],"order":[1],"to":[2,41,57,78,89,105,191,205,208],"cope":[3],"with":[4,130],"the":[5,17,103,108,112,115,132,145,161,164,167,178,181,199,209,220,245,248,253],"growing":[6],"scale":[7,66,71],"of":[8,19,32,114,137,163,180,194,202,219,240],"deep":[9],"neural":[10],"network":[11],"(DNN)":[12],"models":[13],"and":[14,135,252],"training":[15,25,47,92,109,147,214],"data,":[16],"use":[18],"cloud":[20,37,168],"computing":[21],"for":[22,226],"distributed":[23,45],"DNN":[24,46],"is":[26,76],"becoming":[27],"increasingly":[28],"popular.":[29],"The":[30,184],"amount":[31],"available":[33],"resources":[34,98],"in":[35,166,229],"a":[36,49,64,90,126,153,171,192],"continuously":[38],"changes":[39],"according":[40],"users’":[42],"demands.":[43],"Although":[44],"has":[48],"long":[50],"execution":[51],"time":[52],"ranging":[53],"from":[54],"several":[55,58],"hours":[56],"days,":[59],"existing":[60,200,210],"frameworks":[61],"cannot":[62],"provide":[63],"dynamic":[65,127,133,234],"function":[67],"or":[68],"have":[69],"high":[70],"in/out":[72],"overhead.":[73],"Therefore,":[74],"it":[75,119],"difficult":[77],"achieve":[79],"higher":[80],"performance":[81,162,172,222],"by":[82,189],"adding":[83],"graphics":[84],"processing":[85],"unit":[86],"(GPU)":[87],"nodes":[88],"running":[91],"cluster,":[93],"even":[94,159,237],"when":[95,118,160,238],"surplus":[96,182],"GPU":[97],"become":[99],"available.":[100],"addition,":[102,150],"inability":[104],"dynamically":[106],"reconfigure":[107],"cluster":[110,116,201],"prohibits":[111],"reform":[113],"topology":[117],"was":[120,215,257],"sub-optimally":[121],"created.":[122],"This":[123],"paper":[124],"proposes":[125],"scaling":[128,197],"technique":[129,156],"which":[131,232],"addition":[134,179],"removal":[136],"new":[138],"workers":[139,204],"can":[140,174],"be":[141,175],"performed":[142],"without":[143],"suspending":[144],"ongoing":[146],"job.":[148],"we":[151],"propose":[152],"heterogeneity-aware":[154],"straggler-proof":[155],"so":[157],"that,":[158],"GPUs":[165,239],"are":[169],"uneven,":[170],"benefit":[173],"guaranteed":[176],"through":[177],"resources.":[183],"proposed":[185],"scheme":[186],"improved":[187],"throughput":[188],"up":[190],"factor":[193],"17.52":[195],"during":[196],"out":[198],"five":[203],"ten":[206],"compared":[207],"checkpoint-based":[211],"scheme.":[212],"Furthermore,":[213],"continued":[216],"at":[217],"95.52%":[218],"maximum":[221],"while":[223],"being":[224],"stopped":[225],"841":[227],"seconds":[228],"Elastic":[230],"Horovod,":[231],"supports":[233],"scaling.":[235],"Finally,":[236],"different":[241],"performances":[242],"were":[243],"mixed,":[244],"error":[246],"between":[247],"determined":[249],"batch":[250,255],"size":[251,256],"optimal":[254],"3.37%":[258],"on":[259],"average.":[260]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":1}],"updated_date":"2026-07-22T07:51:19.307946","created_date":"2025-10-10T00:00:00"}