Anatomy of high-performance deep learning convolutions on SIMD architectures.
Evangelos GeorganasSasikanth AvanchaKunal BanerjeeDhiraj D. KalamkarGreg HenryHans PabstAlexander HeineckePublished in: SC (2018)
Keyphrases
- deep learning
- array processor
- parallel architectures
- unsupervised learning
- single instruction multiple data
- unsupervised feature learning
- machine learning
- massively parallel
- mental models
- parallel algorithm
- weakly supervised
- restricted boltzmann machine
- deep architectures
- medical images
- parallel computers
- three dimensional
- decision making
- reinforcement learning
- data sets