Channel and filter parallelism for large-scale CNN training

@inproceedings{DBLP:conf/sc/DrydenMMBSE19,
   author    = {Nikoli Dryden and
                Naoya Maruyama and
                Tim Moon and
                Tom Benson and
                Marc Snir and
                Brian Van Essen},
   editor    = {Michela Taufer and
                Pavan Balaji and
                Antonio J. Pe{\~{n}}a},
   title     = {Channel and filter parallelism for large-scale {CNN}
   training},
   booktitle = {Proceedings of the International Conference for High
   Performance Computing,
                Networking, Storage and Analysis, {SC} 2019, Denver,
                Colorado, USA,
                November 17-19, 2019},
   pages     = {10:1--10:20},
   publisher = {{ACM}},
   year      = {2019},
   url       = {https://doi.org/10.1145/3295500.3356207},
   doi       = {10.1145/3295500.3356207},
   timestamp = {Sat, 09 Nov 2019 12:06:02 +0100},
   biburl    = {https://dblp.org/rec/conf/sc/DrydenMMBSE19.bib},
   bibsource = {dblp computer science bibliography, https://dblp.org}
 }