Experimental Demonstration and Tolerancing of a Large-Scale Neural Network (165 000 Synapses) Using Phase-Change Memory as the Synaptic Weight Element
SCIE
SCOPUS
- Title
- Experimental Demonstration and Tolerancing of a Large-Scale Neural Network (165 000 Synapses) Using Phase-Change Memory as the Synaptic Weight Element
- Authors
- Burr, GW; Shelby, RM; Sidler, S; di Nolfo, C; Jang, J; Boybat, I; Shenoy, RS; Narayanan, P; Virwani, K; Giacometti, EU; Kurdi, BN; Hwang, H
- Date Issued
- 2015-11
- Publisher
- IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
- 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.
- URI
- https://oasis.postech.ac.kr/handle/2014.oak/35495
- DOI
- 10.1109/TED.2015.2439635
- ISSN
- 0018-9383
- Article Type
- Article
- Citation
- IEEE TRANSACTIONS ON ELECTRON DEVICES, vol. 62, no. 11, page. 3498 - 3507, 2015-11
- Files in This Item:
- There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.