{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T07:53:16Z","timestamp":1782287596910,"version":"3.54.5"},"reference-count":61,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"8","license":[{"start":{"date-parts":[[2024,8,1]],"date-time":"2024-08-01T00:00:00Z","timestamp":1722470400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2024,8,1]],"date-time":"2024-08-01T00:00:00Z","timestamp":1722470400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2024,8,1]],"date-time":"2024-08-01T00:00:00Z","timestamp":1722470400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62372055"],"award-info":[{"award-number":["62372055"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Sci-Tech Innovation 2030 Agenda of China","award":["2023ZD0120502"],"award-info":[{"award-number":["2023ZD0120502"]}]},{"DOI":"10.13039\/501100012226","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"publisher","award":["PASC DaCeMI project"],"award-info":[{"award-number":["PASC DaCeMI project"]}],"id":[{"id":"10.13039\/501100012226","id-type":"DOI","asserted-by":"publisher"}]},{"name":"European High-Performance Computing Joint Undertaking","award":["101034126"],"award-info":[{"award-number":["101034126"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Parallel Distrib. Syst."],"published-print":{"date-parts":[[2024,8]]},"DOI":"10.1109\/tpds.2024.3397800","type":"journal-article","created":{"date-parts":[[2024,5,7]],"date-time":"2024-05-07T17:39:50Z","timestamp":1715103590000},"page":"1331-1344","source":"Crossref","is-referenced-by-count":9,"title":["AutoDDL: Automatic Distributed Deep Learning With Near-Optimal Bandwidth Cost"],"prefix":"10.1109","volume":"35","author":[{"ORCID":"https:\/\/orcid.org\/0009-0006-4922-5165","authenticated-orcid":false,"given":"Jinfan","family":"Chen","sequence":"first","affiliation":[{"name":"Department of Computer Science, ETH Zurich, Zürich, Switzerland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0022-7865","authenticated-orcid":false,"given":"Shigang","family":"Li","sequence":"additional","affiliation":[{"name":"School of Computer Science, Beijing University of Posts and Telecommunications, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-6835-2467","authenticated-orcid":false,"given":"Ran","family":"Guo","sequence":"additional","affiliation":[{"name":"OneFlow, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0700-2645","authenticated-orcid":false,"given":"Jinhui","family":"Yuan","sequence":"additional","affiliation":[{"name":"OneFlow, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1333-9797","authenticated-orcid":false,"given":"Torsten","family":"Hoefler","sequence":"additional","affiliation":[{"name":"Department of Computer Science, ETH Zurich, Zürich, Switzerland"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","article-title":"Scaling laws for neural language models","author":"Kaplan","year":"2020"},{"key":"ref2","article-title":"GPT-4 technical report","year":"2023"},{"key":"ref3","first-page":"1877","article-title":"Language models are few-shot learners","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Brown"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/SC41404.2022.00016"},{"key":"ref5","article-title":"2.5-dimensional distributed model training","author":"Wang","year":"2021"},{"key":"ref6","article-title":"Maximizing parallelism in distributed training for huge neural networks","author":"Bian","year":"2021"},{"key":"ref7","article-title":"Colossal-AI: A unified deep learning system for large-scale parallel training","author":"Bian","year":"2021"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.5555\/2685048.2685095"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/3320060"},{"key":"ref10","article-title":"Horovod: Fast and easy distributed deep learning in TensorFlow","author":"Sergeev","year":"2018"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3332466.3374528"},{"key":"ref12","article-title":"Efficient large-scale language model training on GPU clusters","volume-title":"Proc. Int. Conf. High Perform. Comput. Netw. Storage Anal.","author":"Narayanan"},{"key":"ref13","article-title":"One weird trick for parallelizing convolutional neural networks","author":"Krizhevsky","year":"2014"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1145\/3437801.3441593"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1145\/3341301.3359646"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.1811.06965"},{"key":"ref17","article-title":"PipeMare: Asynchronous pipeline parallel DNN training","author":"Yang","year":"2019"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/3458817.3476145"},{"key":"ref19","first-page":"1","article-title":"GShard: Scaling giant models with conditional computation and automatic sharding","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lepikhin"},{"key":"ref20","first-page":"559","article-title":"Alpa: Automating inter- and intra-operator parallelism for distributed deep learning","volume-title":"Proc. 16th USENIX Symp. Operating Syst. Des. Implementation","author":"Zheng"},{"key":"ref21","first-page":"1","article-title":"Beyond data and model parallelism for deep neural networks","volume-title":"Proc. 2nd Conf. Mach. Learn. Syst.","author":"Jia"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3302424.3303953"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1145\/3295500.3356181"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1145\/3539781.3539784"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpdc.2004.03.021"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2013.80"},{"key":"ref27","article-title":"OneFlow: Redesign the distributed deep learning framework from scratch","author":"Yuan","year":"2021"},{"key":"ref28","article-title":"Googles neural machine translation system: Bridging the gap between human and machine translation","author":"Wu","year":"2016"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ipdps54959.2023.00031"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPS.2019.00031"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/3295500.3356207"},{"key":"ref32","volume-title":"A Cellular Computer to Implement the Kalman Filter Algorithm","author":"Cannon","year":"1969"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1147\/rd.395.0575"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1016\/0304-3975(90)90188-N"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-23397-5_10"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1145\/3545008.3545087"},{"key":"ref37","article-title":"PipeFisher: Efficient training of large language models using pipelining and fisher information matrices","volume-title":"Proc. Mach. Learn. Syst.","volume":"5","author":"Osawa"},{"key":"ref38","article-title":"TeraPipe: Token-level pipeline parallelism for training large-scale language models","author":"Li","year":"2021"},{"key":"ref39","first-page":"7937","article-title":"Memory-efficient pipeline-parallel DNN training","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Narayanan"},{"key":"ref40","doi-asserted-by":"publisher","DOI":"10.1109\/SC41405.2020.00049"},{"key":"ref41","first-page":"2279","article-title":"Exploring hidden dimensions in parallelizing convolutional neural networks","volume-title":"Proc. 35th Int. Conf. Mach. Learn.","author":"Jia"},{"key":"ref42","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1023\/A:1020281327116","article-title":"An introduction to MCMC for machine learning","volume":"50","author":"Andrieu","year":"2004","journal-title":"Mach. Learn."},{"key":"ref43","first-page":"267","article-title":"Unity: Accelerating DNN training through joint optimization of algebraic transformations and parallelization","volume-title":"Proc. USENIX Symp. Operating Syst. Des. Implementation","author":"Unger"},{"key":"ref44","first-page":"1025","article-title":"Pase: Parallelization strategies for efficient DNN training","volume-title":"Proc. IEEE Int. Parallel Distrib. Process. Symp.","author":"Elango"},{"key":"ref45","article-title":"Colossal-auto: Unified automation of parallelization and activation checkpoint for large-scale models","author":"Liu","year":"2023"},{"key":"ref46","article-title":"Training deep nets with sublinear memory cost","author":"Chen","year":"2016"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/ICDCS.2019.00220"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2019\/473"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1145\/3452296.3472904"},{"key":"ref50","first-page":"1707","article-title":"QSGD: Communication-efficient SGD via gradient quantization and encoding","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Alistarh"},{"key":"ref51","article-title":"signSGD: Compressed optimisation for non-convex problems","author":"Bernstein","year":"2018"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.5555\/3294771.3294915"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1016\/j.jpdc.2008.09.002"},{"key":"ref54","doi-asserted-by":"publisher","DOI":"10.1177\/1094342005051521"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1093\/biomet\/57.1.97"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/SMC42975.2020.9283201"},{"key":"ref57","first-page":"1106","article-title":"ImageNet classification with deep convolutional neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Krizhevsky"},{"key":"ref58","article-title":"Very deep convolutional networks for large-scale image recognition","author":"Simonyan","year":"2015"},{"key":"ref59","article-title":"Deep learning recommendation model for personalization and recommendation systems","author":"Naumov","year":"2019"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1145\/2807591.2807644"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"}],"container-title":["IEEE Transactions on Parallel and Distributed Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/71\/10557478\/10521778.pdf?arnumber=10521778","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,16]],"date-time":"2025-06-16T19:01:24Z","timestamp":1750100484000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10521778\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,8]]},"references-count":61,"journal-issue":{"issue":"8"},"URL":"https:\/\/doi.org\/10.1109\/tpds.2024.3397800","relation":{},"ISSN":["1045-9219","1558-2183","2161-9883"],"issn-type":[{"value":"1045-9219","type":"print"},{"value":"1558-2183","type":"electronic"},{"value":"2161-9883","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,8]]}}}