{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,10]],"date-time":"2026-03-10T20:50:50Z","timestamp":1773175850504,"version":"3.50.1"},"reference-count":49,"publisher":"IEEE","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2018,9]]},"DOI":"10.1109\/iiswc.2018.8573521","type":"proceedings-article","created":{"date-parts":[[2018,12,13]],"date-time":"2018-12-13T19:48:21Z","timestamp":1544730501000},"page":"122-133","source":"Crossref","is-referenced-by-count":30,"title":["Profiling DNN Workloads on a Volta-based DGX-1 System"],"prefix":"10.1109","author":[{"given":"Saiful A.","family":"Mojumder","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Marcia S","family":"Louis","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yifan","family":"Sun","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Amir Kavyan","family":"Ziabari","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jose L.","family":"Abellan","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"John","family":"Kim","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"David","family":"Kaeli","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ajay","family":"Joshi","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-44781-0_45"},{"key":"ref38","article-title":"An analysis of the delayed gradients problem in asynchronous sgd","author":"srinivasan","year":"2018"},{"key":"ref33","doi-asserted-by":"crossref","first-page":"117","DOI":"10.1016\/j.jpdc.2008.09.002","article-title":"Bandwidth optimal all-reduce algorithms for clusters of workstations","volume":"69","author":"patarasuk","year":"2009","journal-title":"Journal of Parallel and Distributed Computing"},{"key":"ref32","article-title":"Gpu asynchronous stochastic gradient descent to speed up neural network training","author":"paine","year":"2013","journal-title":"arXiv preprint arXiv 1312 6186"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2004.01.013"},{"key":"ref30","article-title":"Cublas library","volume":"15","author":"nvidia","year":"2008","journal-title":"NVIDIA Corporation Santa Clara California"},{"key":"ref37","article-title":"Don't decay the learning rate, increase the batch size","author":"smith","year":"2017","journal-title":"arXiv preprint arXiv 1711 03890"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1109\/CCBD.2016.029"},{"key":"ref35","article-title":"Performance modeling and evaluation of distributed deep learning frameworks on gpus","author":"shi","year":"2017","journal-title":"arXiv preprint arXiv 1711 03890"},{"key":"ref34","doi-asserted-by":"publisher","DOI":"10.1016\/j.neunet.2014.09.003"},{"key":"ref28","year":"2018","journal-title":"Nvidia cuDNN"},{"key":"ref27","year":"0","journal-title":"Nvidia dgx-1 with tesla v100 system architecture"},{"key":"ref29","year":"2018","journal-title":"Profiler User's Guide"},{"key":"ref2","article-title":"Tensorflow: Large-scale machine learning on heterogeneous distributed systems","author":"abadi","year":"2016","journal-title":"arXiv preprint arXiv 1603 04467"},{"key":"ref1","year":"2018","journal-title":"NVIDIA Collective Communications Library (NCCL)"},{"key":"ref20","article-title":"One weird trick for parallelizing convolutional neural networks","author":"krizhevsky","year":"2014","journal-title":"arXiv preprint arXiv 1404 5997"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/BDCloud-SocialCom-SustainCom.2016.76"},{"key":"ref21","first-page":"1097","article-title":"Imagenet classification with deep convolutional neural networks","author":"krizhevsky","year":"2012","journal-title":"Advances in neural information processing systems"},{"key":"ref24","doi-asserted-by":"crossref","DOI":"10.21437\/Interspeech.2014-211","article-title":"Distributed learning of multilingual dnn feature extractors using gpus","author":"miao","year":"2014"},{"key":"ref23","first-page":"583","article-title":"Scaling distributed machine learning with the parameter server","volume":"14","author":"li","year":"2014","journal-title":"OSDI"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.178"},{"key":"ref25","author":"nervanasystems","year":"2017","journal-title":"Deep learning frameworks"},{"key":"ref10","first-page":"1223","article-title":"Large scale distributed deep networks","author":"dean","year":"2012","journal-title":"Advances in neural information processing systems"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1145\/3184407.3184423"},{"key":"ref40","first-page":"3","article-title":"Evaluating on-node gpu interconnects for deep learning workloads","author":"tallent","year":"2017","journal-title":"International Workshop on Performance Modeling Benchmarking and Simulation of High Performance Computer Systems"},{"key":"ref12","article-title":"Software frameworks for deep learning at scale","author":"fox","year":"2016","journal-title":"Internal Indiana University Technical Report"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/IPDPSW.2017.36"},{"key":"ref14","first-page":"249","article-title":"Understanding the difficulty of training deep feedforward neural networks","author":"glorot","year":"2010","journal-title":"Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics"},{"key":"ref15","article-title":"Accurate, large minibatch sgd: training imagenet in 1 hour","author":"goyal","year":"2017","journal-title":"arXiv preprint arXiv 1706 02677"},{"key":"ref16","article-title":"Medical image processing in the age of deep learning","author":"hagerty","year":"2017"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1145\/2647868.2654889"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1145\/2834892.2834893"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/ISPASS.2017.7975270"},{"key":"ref4","article-title":"Comparative study of deep learning software frameworks","author":"bahrampour","year":"2015","journal-title":"arXiv preprint arXiv 1511 05271"},{"key":"ref3","article-title":"Optimized broadcast for deep learning workloads on dense-gpu infiniband clusters: Mpi or nccl?","author":"awan","year":"2017","journal-title":"arXiv preprint arXiv 1707 07816"},{"key":"ref6","article-title":"Mxnet: A flexible and efficient machine learning library for heterogeneous distributed systems","author":"chen","year":"2015","journal-title":"arXiv preprint arXiv 1512 00327"},{"key":"ref5","article-title":"Deep learning with gpus","author":"brown","year":"2015","journal-title":"Larry Brown Ph D Johns Hopkins University"},{"key":"ref8","article-title":"Torch7: A matlab-like environment for machine learning","author":"collobert","year":"2011","journal-title":"BigLearn NIPS Workshop"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2014.2325029"},{"key":"ref49","article-title":"Tbd: Benchmarking and analyzing deep neural network training","author":"zhu","year":"2018","journal-title":"arXiv preprint arXiv 1803 06905"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1145\/2901318.2901323"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2013.6638950"},{"key":"ref45","article-title":"Poseidon: A system architecture for efficient gpu-based deep learning on multiple machines","author":"zhang","year":"2015","journal-title":"arXiv preprint arXiv 1512 00327"},{"key":"ref48","article-title":"Asynchronous stochastic gradient descent with delay compensation","author":"zheng","year":"2016","journal-title":"arXiv preprint arXiv 1609 09861"},{"key":"ref47","first-page":"2379","article-title":"Fast asynchronous parallel stochastic gradient descent: A lock-free approach with convergence guarantee","author":"zhao","year":"2016","journal-title":"AAAI"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1007\/s12083-017-0574-4"},{"key":"ref41","article-title":"Multi-gpu training of convnets","author":"yadan","year":"2013","journal-title":"arXiv preprint arXiv 1312 5853"},{"key":"ref44","article-title":"An introduction to computational networks and the computational network toolkit","author":"yu","year":"2014","journal-title":"Microsoft Technical Report MSR-TR-2014–112"},{"key":"ref43","article-title":"100-epoch imagenet training with alexnet in 24 minutes","author":"you","year":"2017","journal-title":"ArXiv e-prints"}],"event":{"name":"2018 IEEE International Symposium on Workload Characterization (IISWC)","location":"Raleigh, NC","start":{"date-parts":[[2018,9,30]]},"end":{"date-parts":[[2018,10,2]]}},"container-title":["2018 IEEE International Symposium on Workload Characterization (IISWC)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/8554060\/8573472\/08573521.pdf?arnumber=8573521","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,1,27]],"date-time":"2022-01-27T08:25:48Z","timestamp":1643271948000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/8573521\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2018,9]]},"references-count":49,"URL":"https:\/\/doi.org\/10.1109\/iiswc.2018.8573521","relation":{},"subject":[],"published":{"date-parts":[[2018,9]]}}}