TensorFlow at Scale: Performance and productivity analysis of distributed training with Horovod, MLSL, and Cray PE ML

Thorsten Kurth, Mikhail Smorkalov, Peter Mendygral, Srinivas Sridharan 0002, Amrita Mathuriya. TensorFlow at Scale: Performance and productivity analysis of distributed training with Horovod, MLSL, and Cray PE ML. Concurrency - Practice and Experience, 31(16), 2019. [doi]

@article{KurthSMSM19,
  title = {TensorFlow at Scale: Performance and productivity analysis of distributed training with Horovod, MLSL, and Cray PE ML},
  author = {Thorsten Kurth and Mikhail Smorkalov and Peter Mendygral and Srinivas Sridharan 0002 and Amrita Mathuriya},
  year = {2019},
  doi = {10.1002/cpe.4989},
  url = {https://doi.org/10.1002/cpe.4989},
  researchr = {https://researchr.org/publication/KurthSMSM19},
  cites = {0},
  citedby = {0},
  journal = {Concurrency - Practice and Experience},
  volume = {31},
  number = {16},
}