Login / Signup
Optimized large-message broadcast for deep learning workloads: MPI, MPI+NCCL, or NCCL2?
Ammar Ahmad Awan
Karthik Vadambacheri Manian
Ching-Hsiang Chu
Hari Subramoni
Dhabaleswar K. Panda
Published in:
Parallel Comput. (2019)
Keyphrases
</>
deep learning
message passing
general purpose
unsupervised learning
unsupervised feature learning
deep architectures
email
mental models
weakly supervised
data sets
machine learning
data mining
pattern recognition
restricted boltzmann machine