{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,2]],"date-time":"2026-04-02T09:41:40Z","timestamp":1775122900461,"version":"3.50.1"},"reference-count":36,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"11","license":[{"start":{"date-parts":[[2019,11,1]],"date-time":"2019-11-01T00:00:00Z","timestamp":1572566400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2019,11,1]],"date-time":"2019-11-01T00:00:00Z","timestamp":1572566400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2019,11,1]],"date-time":"2019-11-01T00:00:00Z","timestamp":1572566400000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100008982","name":"National Science Foundation","doi-asserted-by":"publisher","award":["OAC-1540931"],"award-info":[{"award-number":["OAC-1540931"]}],"id":[{"id":"10.13039\/501100008982","id-type":"DOI","asserted-by":"publisher"}]},{"name":"U.S. DOE Office of Science, Office of Advanced Scientific Computing Research, Applied Mathematics program","award":["DE-SC0010200"],"award-info":[{"award-number":["DE-SC0010200"]}]},{"DOI":"10.13039\/100000185","name":"DARPA","doi-asserted-by":"crossref","award":["HR0011-12- 2-0016"],"award-info":[{"award-number":["HR0011-12- 2-0016"]}],"id":[{"id":"10.13039\/100000185","id-type":"DOI","asserted-by":"crossref"}]},{"name":"ASPIRE Lab industrial sponsors and affiliates Intel"},{"DOI":"10.13039\/100006785","name":"Google","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100006785","id-type":"DOI","asserted-by":"publisher"}]},{"name":"HP"},{"DOI":"10.13039\/501100003816","name":"Huawei Technologies","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100003816","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100004118","name":"LG Display","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100004118","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100004356","name":"Nokia","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100004356","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100007065","name":"Nvidia","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100007065","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100004682","name":"Oracle","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100004682","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100004358","name":"Samsung","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100004358","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Parallel Distrib. Syst."],"published-print":{"date-parts":[[2019,11,1]]},"DOI":"10.1109\/tpds.2019.2913833","type":"journal-article","created":{"date-parts":[[2019,4,30]],"date-time":"2019-04-30T20:08:58Z","timestamp":1556654938000},"page":"2449-2462","source":"Crossref","is-referenced-by-count":52,"title":["Fast Deep Neural Network Training on Distributed Systems and Cloud TPUs"],"prefix":"10.1109","volume":"30","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-2816-4384","authenticated-orcid":false,"given":"Yang","family":"You","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5921-0035","authenticated-orcid":false,"given":"Zhao","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Cho-Jui","family":"Hsieh","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"James","family":"Demmel","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kurt","family":"Keutzer","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-24685-5_1"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1177\/1094342005051521"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2009.5206848"},{"key":"ref30","doi-asserted-by":"publisher","DOI":"10.1016\/0167-8191(96)00024-5"},{"key":"ref36","first-page":"1310","article-title":"On the difficulty of training recurrent neural networks","author":"pascanu","year":"2013","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1006\/jpdc.1994.1091"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-30218-6_13"},{"key":"ref10","article-title":"How to scale distributed deep learning?","author":"jin","year":"2016"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/ALLERTON.2016.7852343"},{"key":"ref12","first-page":"1223","article-title":"Large scale distributed deep networks","author":"dean","year":"2012","journal-title":"Proc Advances Neural Inf Process Syst"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2014.6853593"},{"key":"ref14","article-title":"Scale out for large minibatch SGD: Residual network training on imagenet-1k with improved accuracy and reduced time to train","author":"codreanu","year":"2017"},{"key":"ref15","article-title":"Extremely large minibatch SGD: Training resnet-50 on imagenet in 15 minutes","author":"akiba","year":"2017"},{"key":"ref16","article-title":"Imagenet training in 24 minutes","author":"you","year":"2017"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1016\/0364-0213(90)90002-E"},{"key":"ref18","volume":"1","author":"goodfellow","year":"2016","journal-title":"Deep Learning"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735"},{"key":"ref28","article-title":"Scaling distributed machine learning with system and algorithm co-design","author":"li","year":"2017"},{"key":"ref4","article-title":"Firecaffe: Near-linear acceleration of deep neural network training on compute clusters","volume":"abs 1511 175","author":"iandola","year":"2015","journal-title":"CoRR"},{"key":"ref27","article-title":"One weird trick for parallelizing convolutional neural networks","author":"krizhevsky","year":"2014"},{"key":"ref3","article-title":"Deep speech 2: End-to-end speech recognition in english and mandarin","author":"amodei","year":"2015"},{"key":"ref6","article-title":"Scaling SGD batch size to 32K for imagenet training","author":"you","year":"2017"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1145\/2647868.2654889"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.284"},{"key":"ref8","first-page":"693","article-title":"Hogwild: A lock-free approach to parallelizing stochastic gradient descent","author":"recht","year":"2011","journal-title":"Proc Advances Neural Inf Process Syst"},{"key":"ref7","article-title":"Distributed deep learning using synchronous stochastic gradient descent","author":"das","year":"2016"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1145\/3225058.3225069"},{"key":"ref9","first-page":"685","article-title":"Deep learning with elastic averaging SGD","author":"zhang","year":"2015","journal-title":"Proc Advances Neural Inf Process Syst"},{"key":"ref1","article-title":"Accurate, large minibatch sgd: Training imagenet in 1 hour","author":"goyal","year":"2017"},{"key":"ref20","article-title":"Google's neural machine translation system: Bridging the gap between human and machine translation","author":"wu","year":"2016"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1145\/3079856.3080246"},{"key":"ref21","article-title":"Neural machine translation (seq2seq) tutorial","author":"luong","year":"2017"},{"key":"ref24","first-page":"685","article-title":"Deep learning with elastic averaging SGD","author":"zhang","year":"2015","journal-title":"Proc Advances Neural Inf Process Syst"},{"key":"ref23","article-title":"Revisiting distributed synchronous SGD","author":"chen","year":"2016"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref25","article-title":"On large-batch training for deep learning: Generalization gap and sharp minima","author":"keskar","year":"2016"}],"container-title":["IEEE Transactions on Parallel and Distributed Systems"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/71\/8863608\/08703162.pdf?arnumber=8703162","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,7,13]],"date-time":"2022-07-13T20:43:49Z","timestamp":1657745029000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/8703162\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2019,11,1]]},"references-count":36,"journal-issue":{"issue":"11"},"URL":"https:\/\/doi.org\/10.1109\/tpds.2019.2913833","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":[[2019,11,1]]}}}