Vinay Amatya, Abhinav Vishnu, Charles Siegel, Jeff Daily. What does fault tolerant deep learning need from MPI?. In Antonio J. Peña, Pavan Balaji, William Gropp, Rajeev Thakur, editors, Proceedings of the 24th European MPI Users' Group Meeting, EuroMPI/USA 2017, Chicago, IL, USA, September 25-28, 2017. ACM, 2017. [doi]