{"id":"https://openalex.org/W4415330770","doi":"https://doi.org/10.1145/3767295.3769369","title":"Zeppelin: Balancing Variable-length Workloads in Data Parallel Large Model Training","display_name":"Zeppelin: Balancing Variable-length Workloads in Data Parallel Large Model Training","publication_year":2026,"publication_date":"2026-04-24","ids":{"openalex":"https://openalex.org/W4415330770","doi":"https://doi.org/10.1145/3767295.3769369"},"language":"en","primary_location":{"id":"doi:10.1145/3767295.3769369","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3769369","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/10.1145/3767295.3769369","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5016345260","display_name":"Chang Chen","orcid":"https://orcid.org/0000-0003-3425-1118"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chang Chen","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0008-6416-7172","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5015040114","display_name":"Tiancheng Chen","orcid":"https://orcid.org/0009-0002-8071-2552"},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Tiancheng Chen","raw_affiliation_strings":["ETH Zurich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0009-0002-8071-2552","affiliations":[{"raw_affiliation_string":"ETH Zurich, Zurich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102584876","display_name":"Jiangfei Duan","orcid":null},"institutions":[{"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":"Jiangfei Duan","raw_affiliation_strings":["CUHK, Hong Kong, Hong Kong"],"raw_orcid":"https://orcid.org/0000-0002-6327-2033","affiliations":[{"raw_affiliation_string":"CUHK, Hong Kong, Hong Kong","institution_ids":["https://openalex.org/I889458895"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043475210","display_name":"Qianchao Zhu","orcid":"https://orcid.org/0009-0001-5021-2912"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qianchao Zhu","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0009-0001-5021-2912","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103187751","display_name":"Zerui Wang","orcid":"https://orcid.org/0000-0003-1784-1079"},"institutions":[{"id":"https://openalex.org/I4391012619","display_name":"Shanghai Artificial Intelligence Laboratory","ror":"https://ror.org/03wkvpx79","country_code":null,"type":"facility","lineage":["https://openalex.org/I4391012619"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zerui Wang","raw_affiliation_strings":["Shanghai AI Laboratory, Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0003-8081-8469","affiliations":[{"raw_affiliation_string":"Shanghai AI Laboratory, Shanghai, China","institution_ids":["https://openalex.org/I4391012619"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102016069","display_name":"Qinghao Hu","orcid":"https://orcid.org/0009-0007-5300-5787"},"institutions":[{"id":"https://openalex.org/I4210110987","display_name":"IIT@MIT","ror":"https://ror.org/01wp8zh54","country_code":"US","type":"facility","lineage":["https://openalex.org/I30771326","https://openalex.org/I4210110987"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qinghao Hu","raw_affiliation_strings":["MIT, Cambridge, USA"],"raw_orcid":"https://orcid.org/0000-0003-1034-7502","affiliations":[{"raw_affiliation_string":"MIT, Cambridge, USA","institution_ids":["https://openalex.org/I4210110987"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100614090","display_name":"Peng Sun","orcid":"https://orcid.org/0000-0002-1227-5533"},"institutions":[{"id":"https://openalex.org/I4391012619","display_name":"Shanghai Artificial Intelligence Laboratory","ror":"https://ror.org/03wkvpx79","country_code":null,"type":"facility","lineage":["https://openalex.org/I4391012619"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Peng Sun","raw_affiliation_strings":["Shanghai AI Laboratory, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0001-8456-0491","affiliations":[{"raw_affiliation_string":"Shanghai AI Laboratory, Shanghai, China","institution_ids":["https://openalex.org/I4391012619"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114246744","display_name":"Xiuhong Li","orcid":"https://orcid.org/0000-0002-4896-121X"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiuhong Li","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0002-4896-121X","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5014524927","display_name":"Chao Yang","orcid":"https://orcid.org/0000-0002-5204-8060"},"institutions":[{"id":"https://openalex.org/I20231570","display_name":"Peking University","ror":"https://ror.org/02v51f717","country_code":"CN","type":"education","lineage":["https://openalex.org/I20231570"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chao Yang","raw_affiliation_strings":["Peking University, Beijing, China"],"raw_orcid":"https://orcid.org/0000-0001-7426-6248","affiliations":[{"raw_affiliation_string":"Peking University, Beijing, China","institution_ids":["https://openalex.org/I20231570"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5102875426","display_name":"Torsten Hoefler","orcid":null},"institutions":[{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Torsten Hoefler","raw_affiliation_strings":["ETH Zurich, Zurich, Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-1333-9797","affiliations":[{"raw_affiliation_string":"ETH Zurich, Zurich, Switzerland","institution_ids":["https://openalex.org/I35440088"]}]}],"institutions":[],"countries_distinct_count":4,"institutions_distinct_count":5,"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":0.01231604,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1879","last_page":"1893"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12535","display_name":"Machine Learning and Data Classification","score":0.9641000032424927,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.9641000032424927,"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/T14339","display_name":"Image Processing and 3D Reconstruction","score":0.9280999898910522,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.6583999991416931},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.6137999892234802},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5669000148773193},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.5365999937057495},{"id":"https://openalex.org/keywords/layer","display_name":"Layer (electronics)","score":0.5185999870300293},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.4902999997138977},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.4717999994754791},{"id":"https://openalex.org/keywords/routing","display_name":"Routing (electronic design automation)","score":0.41749998927116394}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8055999875068665},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.6583999991416931},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.6137999892234802},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5669000148773193},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5428000092506409},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5365999937057495},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.5185999870300293},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5127000212669373},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.4902999997138977},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.4717999994754791},{"id":"https://openalex.org/C74172769","wikidata":"https://www.wikidata.org/wiki/Q1446839","display_name":"Routing (electronic design automation)","level":2,"score":0.41749998927116394},{"id":"https://openalex.org/C129844170","wikidata":"https://www.wikidata.org/wiki/Q41299","display_name":"Quadratic equation","level":2,"score":0.35109999775886536},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.3375000059604645},{"id":"https://openalex.org/C138959212","wikidata":"https://www.wikidata.org/wiki/Q1806783","display_name":"Load balancing (electrical power)","level":3,"score":0.3352999985218048},{"id":"https://openalex.org/C106515295","wikidata":"https://www.wikidata.org/wiki/Q26806595","display_name":"Parallel processing","level":2,"score":0.33379998803138733},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.329800009727478},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.32089999318122864},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.3102000057697296},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.2856000065803528},{"id":"https://openalex.org/C37829124","wikidata":"https://www.wikidata.org/wiki/Q763575","display_name":"Parallel communication","level":3,"score":0.27480000257492065},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.27309998869895935},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.257099986076355}],"mesh":[],"locations_count":6,"locations":[{"id":"doi:10.1145/3767295.3769369","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3769369","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:www.research-collection.ethz.ch:20.500.11850/800395","is_oa":true,"landing_page_url":"http://hdl.handle.net/20.500.11850/800395","pdf_url":null,"source":{"id":"https://openalex.org/S4306402302","display_name":"Repository for Publications and Research Data (ETH Zurich)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I35440088","host_organization_name":"ETH Zurich","host_organization_lineage":["https://openalex.org/I35440088"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"EUROSYS '26: Proceedings of the 21st European Conference on Computer Systems","raw_type":"info:eu-repo/semantics/conferenceObject"},{"id":"pmh:oai:arXiv.org:2509.21841","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.21841","pdf_url":"https://arxiv.org/pdf/2509.21841","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:doi:10.3929/ethz-c-000792576","is_oa":true,"landing_page_url":"http://hdl.handle.net/20.500.11850/792576","pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","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","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Working Paper"},{"id":"doi:10.48550/arxiv.2509.21841","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2509.21841","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.3929/ethz-c-000800395","is_oa":true,"landing_page_url":"https://doi.org/10.3929/ethz-c-000800395","pdf_url":null,"source":{"id":"https://openalex.org/S7407051236","display_name":"ETH Z\u00fcrich Research Collection","issn_l":null,"issn":[],"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","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Other"}],"best_oa_location":{"id":"doi:10.1145/3767295.3769369","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3769369","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Training":[0],"large":[1],"language":[2],"models":[3],"(LLMs)":[4],"with":[5,45],"increasingly":[6],"long":[7],"and":[8,41,75,79],"varying":[9,58],"sequence":[10,46],"lengths":[11,65],"introduces":[12],"severe":[13],"load":[14],"imbalance":[15],"challenges":[16],"in":[17,49,66],"large-scale":[18],"data-parallel":[19],"training.":[20],"Recent":[21],"frameworks":[22],"attempt":[23],"to":[24],"mitigate":[25],"these":[26],"issues":[27],"through":[28],"data":[29],"reorganization":[30],"or":[31],"hybrid":[32],"parallel":[33,77],"strategies.":[34],"However,":[35],"they":[36],"often":[37],"overlook":[38],"how":[39],"computational":[40],"communication":[42],"costs":[43],"scale":[44],"length,":[47],"resulting":[48],"suboptimal":[50],"performance.":[51],"We":[52],"identify":[53],"three":[54],"critical":[55],"challenges:":[56],"(1)":[57],"computation-to-communication":[59],"ratios":[60],"across":[61],"sequences":[62],"of":[63],"different":[64],"distributed":[67],"attention,":[68],"(2)":[69],"mismatch":[70],"between":[71],"static":[72],"NIC-GPU":[73],"affinity":[74],"dynamic":[76],"workloads,":[78],"(3)":[80],"distinct":[81],"optimal":[82],"partitioning":[83],"strategies":[84],"required":[85],"for":[86],"quadratic":[87],"attention":[88],"versus":[89],"linear":[90],"components.":[91]},"counts_by_year":[],"updated_date":"2026-07-03T08:13:44.112507","created_date":"2025-10-19T00:00:00"}