{"id":"https://openalex.org/W4414830116","doi":"https://doi.org/10.1145/3767295.3803568","title":"MegaScale-Data: Scaling DataLoader for Multisource Large Foundation Model Training","display_name":"MegaScale-Data: Scaling DataLoader for Multisource Large Foundation Model Training","publication_year":2026,"publication_date":"2026-04-24","ids":{"openalex":"https://openalex.org/W4414830116","doi":"https://doi.org/10.1145/3767295.3803568"},"language":"en","primary_location":{"id":"doi:10.1145/3767295.3803568","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803568","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 21st European Conference on Computer Systems","raw_type":"proceedings-article"},"type":"article","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3767295.3803568","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5087869611","display_name":"Juntao Zhao","orcid":"https://orcid.org/0000-0003-3376-0607"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]},{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":true,"raw_author_name":"Juntao Zhao","raw_affiliation_strings":["The University of Hong Kong, Hong Kong, Hong Kong"],"raw_orcid":"https://orcid.org/0000-0003-3376-0607","affiliations":[{"raw_affiliation_string":"The University of Hong Kong, Hong Kong, Hong Kong","institution_ids":["https://openalex.org/I177725633","https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5067546310","display_name":"Qi Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qi Lu","raw_affiliation_strings":["ByteDance Inc., Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0003-6926-5787","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Shanghai, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100690218","display_name":"Wei Jia","orcid":"https://orcid.org/0000-0002-9423-0143"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wei Jia","raw_affiliation_strings":["ByteDance Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0009-0001-8596-5877","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5093967737","display_name":"Borui Wan","orcid":"https://orcid.org/0009-0008-5902-1611"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]},{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Borui Wan","raw_affiliation_strings":["The University of Hong Kong, Hong Kong, Hong Kong"],"raw_orcid":"https://orcid.org/0009-0008-5902-1611","affiliations":[{"raw_affiliation_string":"The University of Hong Kong, Hong Kong, Hong Kong","institution_ids":["https://openalex.org/I177725633","https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090363951","display_name":"Lei Zuo","orcid":"https://orcid.org/0000-0002-7478-3648"},"institutions":[{"id":"https://openalex.org/I58610484","display_name":"Seattle University","ror":"https://ror.org/02jqc0m91","country_code":"US","type":"education","lineage":["https://openalex.org/I58610484"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Lei Zuo","raw_affiliation_strings":["ByteDance Inc., Seattle, USA"],"raw_orcid":"https://orcid.org/0000-0002-7423-1675","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Seattle, USA","institution_ids":["https://openalex.org/I58610484"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017329016","display_name":"Junda Feng","orcid":"https://orcid.org/0000-0003-3664-6615"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Junda Feng","raw_affiliation_strings":["ByteDance Inc., Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0003-3664-6615","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Shanghai, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5091837703","display_name":"Jianyu Jiang","orcid":"https://orcid.org/0000-0002-8684-8509"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jianyu Jiang","raw_affiliation_strings":["ByteDance Inc., Shenzhen, China"],"raw_orcid":"https://orcid.org/0000-0002-8684-8509","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Shenzhen, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062504314","display_name":"Yangrui Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yangrui Chen","raw_affiliation_strings":["ByteDance Inc., Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0007-6682-1783","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Shanghai, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071951396","display_name":"Shuaishuai Cao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shuaishuai Cao","raw_affiliation_strings":["ByteDance Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0009-0000-6870-8898","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102779490","display_name":"Jialing He","orcid":"https://orcid.org/0000-0001-8871-0012"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jialing He","raw_affiliation_strings":["ByteDance Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0009-0008-7858-8728","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Kaihua Jiang","orcid":"https://orcid.org/0009-0006-4598-3807"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kaihua Jiang","raw_affiliation_strings":["ByteDance Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0009-0006-4598-3807","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114585030","display_name":"Yuanzhe Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuanzhe Hu","raw_affiliation_strings":["ByteDance Inc., Beijing, China"],"raw_orcid":"https://orcid.org/0009-0006-6943-6090","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Beijing, China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114130887","display_name":"Shibiao Nong","orcid":null},"institutions":[{"id":"https://openalex.org/I58610484","display_name":"Seattle University","ror":"https://ror.org/02jqc0m91","country_code":"US","type":"education","lineage":["https://openalex.org/I58610484"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shibiao Nong","raw_affiliation_strings":["ByteDance Inc., Seattle, USA"],"raw_orcid":"https://orcid.org/0009-0002-0653-767X","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Seattle, USA","institution_ids":["https://openalex.org/I58610484"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013856812","display_name":"Yanghua Peng","orcid":"https://orcid.org/0000-0003-3989-4358"},"institutions":[{"id":"https://openalex.org/I58610484","display_name":"Seattle University","ror":"https://ror.org/02jqc0m91","country_code":"US","type":"education","lineage":["https://openalex.org/I58610484"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yanghua Peng","raw_affiliation_strings":["ByteDance Inc., Seattle, USA"],"raw_orcid":"https://orcid.org/0000-0003-3989-4358","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Seattle, USA","institution_ids":["https://openalex.org/I58610484"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5084047386","display_name":"Haibin Lin","orcid":"https://orcid.org/0000-0003-4879-5335"},"institutions":[{"id":"https://openalex.org/I58610484","display_name":"Seattle University","ror":"https://ror.org/02jqc0m91","country_code":"US","type":"education","lineage":["https://openalex.org/I58610484"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Haibin Lin","raw_affiliation_strings":["ByteDance Inc., Seattle, USA"],"raw_orcid":"https://orcid.org/0000-0003-4879-5335","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Seattle, USA","institution_ids":["https://openalex.org/I58610484"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5051257081","display_name":"Chuan Wu","orcid":null},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]},{"id":"https://openalex.org/I889458895","display_name":"University of Hong Kong","ror":"https://ror.org/02zhqgq86","country_code":"HK","type":"education","lineage":["https://openalex.org/I889458895"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Chuan Wu","raw_affiliation_strings":["The University of Hong Kong, Hong Kong, Hong Kong"],"raw_orcid":"https://orcid.org/0000-0002-3144-4398","affiliations":[{"raw_affiliation_string":"The University of Hong Kong, Hong Kong, Hong Kong","institution_ids":["https://openalex.org/I177725633","https://openalex.org/I889458895"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":16,"corresponding_author_ids":["https://openalex.org/A5087869611"],"corresponding_institution_ids":["https://openalex.org/I177725633","https://openalex.org/I889458895"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.00517036,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"18"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T14470","display_name":"Advanced Data Processing Techniques","score":0.8644000291824341,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T14470","display_name":"Advanced Data Processing Techniques","score":0.8644000291824341,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.8192999958992004,"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/T13734","display_name":"Advanced Computational Techniques and Applications","score":0.7365999817848206,"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/bottleneck","display_name":"Bottleneck","score":0.5432999730110168},{"id":"https://openalex.org/keywords/workflow","display_name":"Workflow","score":0.4950000047683716},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.435699999332428},{"id":"https://openalex.org/keywords/loader","display_name":"Loader","score":0.4334000051021576},{"id":"https://openalex.org/keywords/fault-tolerance","display_name":"Fault tolerance","score":0.4300000071525574},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.375},{"id":"https://openalex.org/keywords/workload","display_name":"Workload","score":0.3192000091075897},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.31450000405311584}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7653999924659729},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.5432999730110168},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.4950000047683716},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.46369999647140503},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.435699999332428},{"id":"https://openalex.org/C2779041774","wikidata":"https://www.wikidata.org/wiki/Q650550","display_name":"Loader","level":2,"score":0.4334000051021576},{"id":"https://openalex.org/C63540848","wikidata":"https://www.wikidata.org/wiki/Q3140932","display_name":"Fault tolerance","level":2,"score":0.4300000071525574},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.375900000333786},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.375},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.3192000091075897},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.31450000405311584},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3125999867916107},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.3109999895095825},{"id":"https://openalex.org/C47487241","wikidata":"https://www.wikidata.org/wiki/Q5227230","display_name":"Data access","level":2,"score":0.30640000104904175},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.2962000072002411},{"id":"https://openalex.org/C147494362","wikidata":"https://www.wikidata.org/wiki/Q2078905","display_name":"Troubleshooting","level":2,"score":0.28450000286102295},{"id":"https://openalex.org/C34736171","wikidata":"https://www.wikidata.org/wiki/Q918333","display_name":"Preprocessor","level":2,"score":0.2842999994754791},{"id":"https://openalex.org/C70061542","wikidata":"https://www.wikidata.org/wiki/Q989016","display_name":"Distributed database","level":2,"score":0.2797999978065491},{"id":"https://openalex.org/C62611344","wikidata":"https://www.wikidata.org/wiki/Q1062658","display_name":"Node (physics)","level":2,"score":0.27239999175071716},{"id":"https://openalex.org/C10551718","wikidata":"https://www.wikidata.org/wiki/Q5227332","display_name":"Data pre-processing","level":2,"score":0.2712000012397766},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2700999975204468},{"id":"https://openalex.org/C45340560","wikidata":"https://www.wikidata.org/wiki/Q215382","display_name":"Disjoint sets","level":2,"score":0.2685999870300293},{"id":"https://openalex.org/C138827492","wikidata":"https://www.wikidata.org/wiki/Q6661985","display_name":"Data processing","level":2,"score":0.2685999870300293},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.25870001316070557},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.25760000944137573},{"id":"https://openalex.org/C75684735","wikidata":"https://www.wikidata.org/wiki/Q858810","display_name":"Big data","level":2,"score":0.25619998574256897},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.2547999918460846}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1145/3767295.3803568","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803568","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 21st European Conference on Computer Systems","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2504.09844","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2504.09844","pdf_url":"https://arxiv.org/pdf/2504.09844","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:arXiv.org:2504.09844","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2504.09844","pdf_url":"https://arxiv.org/pdf/2504.09844","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2504.09844","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2504.09844","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.1145/3767295.3803568","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803568","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 21st European Conference on Computer Systems","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Modern":[0],"frameworks":[1],"for":[2,27],"training":[3,22,29,67],"large":[4],"foundation":[5],"models":[6],"(LFMs)":[7],"employ":[8],"dataloaders":[9],"in":[10,103],"a":[11,18],"data-parallel":[12,56],"manner,":[13],"with":[14],"each":[15],"loader":[16],"processing":[17],"disjoint":[19],"subset":[20],"of":[21,47],"data.":[23],"When":[24],"preparing":[25],"data":[26,72,93],"LFM":[28],"that":[30,79],"originates":[31],"from":[32],"multiple,":[33],"distinct":[34],"sources,":[35],"two":[36],"fundamental":[37],"challenges":[38],"arise.":[39],"First,":[40],"due":[41],"to":[42,59],"the":[43,48,51,66],"quadratic":[44],"computational":[45],"complexity":[46],"attention":[49],"operator,":[50],"non-uniform":[52],"sample":[53],"distribution":[54],"over":[55],"ranks":[57],"leads":[58],"significant":[60],"workload":[61],"imbalance":[62],"among":[63],"dataloaders,":[64],"degrading":[65],"efficiency.":[68],"Second,":[69],"supporting":[70],"diverse":[71],"sources":[73],"requires":[74],"per-dataset":[75],"file":[76],"access":[77],"states":[78],"are":[80],"redundantly":[81],"replicated":[82],"across":[83],"parallel":[84],"loaders,":[85],"consuming":[86],"excessive":[87],"memory.":[88],"This":[89],"also":[90],"hinders":[91],"dynamic":[92],"mixing":[94],"(e.g.,":[95],"curriculum":[96],"learning)":[97],"and":[98],"causes":[99],"redundant":[100],"access/memory":[101],"overhead":[102],"hybrid":[104],"parallelism.":[105]},"counts_by_year":[],"updated_date":"2026-05-11T08:15:01.531666","created_date":"2025-10-05T00:00:00"}