Experimental Demonstration and Tolerancing of a Large-Scale Neural Network (165 000 Synapses) Using Phase-Change Memory as the Synaptic Weight Element
Experimental Demonstration and Tolerancing of a Large-Scale Neural Network (165 000 Synapses) Using Phase-Change Memory as the Synaptic Weight Element
复制标题
DOI：
10.1109/ted.2015.2439635
复制
发表时间：
2015-11-01
影响因子：
3.1
通讯作者：
Hwang, Hyunsang
中科院分区：
文献类型：
--
作者：
Burr, Geoffrey W.;Shelby, Robert M.;Hwang, Hyunsang
使用每个突触使用两个相变的内存设备，在使用适合非挥发性记忆（NVM） + selector crossbar的MNIST数据库的子集（5000个示例）上训练了一个具有164 885突触的三层感知器网络（164 885突触）。阵列获得82.2％（82.9％）的培训（概括）精度。使用与实验演示器相匹配的神经网络模拟器，相对于NVM的可变性，产率以及NVM-导电响应的随机性，线性和不对称性，进行了广泛的公差。我们表明，具有对称的，高动态范围的对称线性电导响应的双向NVM能够在此问题上提供相同的高分类精度，就像对同一网络的常规，基于软件的实现。
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.