Neuromorphic computing using non-volatile memory

• Published in: Advances in physics: X Vol. 2; no. 1; pp. 89 - 124
• Main Authors: Burr, Geoffrey W., Shelby, Robert M., Sebastian, Abu, Kim, Sangbum, Kim, Seyoung, Sidler, Severin, Virwani, Kumar, Ishii, Masatoshi, Narayanan, Pritish, Fumarola, Alessandro, Sanches, Lucas L., Boybat, Irem, Le Gallo, Manuel, Moon, Kibong, Woo, Jiyoo, Hwang, Hyunsang, Leblebici, Yusuf
• Format: Journal Article
• Language: English
• Published: Abingdon Taylor & Francis 01.01.2017; Taylor & Francis Ltd; Taylor & Francis Group
• Subjects: Algorithms; Arrays; Artificial neural networks; Back propagation; Computation; Linearity; Machine learning; Mathematical analysis; Matrix algebra; Matrix methods; Memory devices; Multiplication; Neural networks; Neuromorphic computing; non-volatile memory; NVM-based neurons; NVM-based synapses; Power management; Random access memory; Resistance; spike-timing-dependent-plasticity; spiking neural networks; vector-matrix multiplication
• DOI: 10.1080/23746149.2016.1259585
• ISSN: 2374-6149

Dense crossbar arrays of non-volatile memory (NVM) devices represent one possible path for implementing massively-parallel and highly energy-efficient neuromorphic computing systems. We first review recent advances in the application of NVM devices to three computing paradigms: spiking neural networks (SNNs), deep neural networks (DNNs), and 'Memcomputing'. In SNNs, NVM synaptic connections are updated by a local learning rule such as spike-timing-dependent-plasticity, a computational approach directly inspired by biology. For DNNs, NVM arrays can represent matrices of synaptic weights, implementing the matrix-vector multiplication needed for algorithms such as backpropagation in an analog yet massively-parallel fashion. This approach could provide significant improvements in power and speed compared to GPU-based DNN training, for applications of commercial significance. We then survey recent research in which different types of NVM devices - including phase change memory, conductive-bridging RAM, filamentary and non-filamentary RRAM, and other NVMs - have been proposed, either as a synapse or as a neuron, for use within a neuromorphic computing application. The relevant virtues and limitations of these devices are assessed, in terms of properties such as conductance dynamic range, (non)linearity and (a)symmetry of conductance response, retention, endurance, required switching power, and device variability.