Experimental Demonstration and Tolerancing of a Large-Scale Neural Network (165 000 Synapses) Using Phase-Change Memory as the Synaptic Weight Element
Lists
Only you see this
Abstract
Using two phase-change memory devices per synapse, a three-layer perceptron network with 164 885 synapses is trained on a subset (5000 examples) of the MNIST database of handwritten digits using a backpropagation variant suitable for nonvolatile memory (NVM) + selector crossbar arrays, obtaining a training (generalization) accuracy of 82.2% (82.9%). Using a neural network simulator matched to the experimental demonstrator, extensive tolerancing is performed with respect to NVM variability, yield, and the stochasticity, linearity, and asymmetry of the NVM-conductance response. We show that a bidirectional NVM with a symmetric, linear conductance response of high dynamic range is capable of delivering the same high classification accuracies on this problem as a conventional, software-based implementation of this same network.
BibTeX
@article{burr2015experimental,
author = {Geoffrey W. Burr and R. M. Shelby and Severin Sidler and Carmelo di Nolfo and Junwoo Jang and Irem Boybat and Rohit S. Shenoy and Pritish Narayanan and Kumar Virwani and Emanuele U. Giacometti and B. N. Kurdi and Hyunsang Hwang},
title = {{Experimental Demonstration and Tolerancing of a Large-Scale Neural Network (165 000 Synapses) Using Phase-Change Memory as the Synaptic Weight Element}},
journal = {IEEE Transactions on Electron Devices},
year = {2015},
doi = {10.1109/ted.2015.2439635},
}Loading related content...