000229322 001__ 229322
000229322 005__ 20180317092424.0
000229322 0247_ $$2doi$$a10.1109/TCSII.2017.2697457
000229322 02470 $$2ISI$$a000414427900018
000229322 037__ $$aARTICLE
000229322 245__ $$aNeuromorphic Architecture with 1M Memristive Synapses for Detection of Weakly Correlated Inputs
000229322 269__ $$a2017
000229322 260__ $$aPiscataway$$bIeee-Inst Electrical Electronics Engineers Inc$$c2017
000229322 300__ $$a5
000229322 336__ $$aJournal Articles
000229322 520__ $$aNeuromorphic computing takes inspiration from the brain to build highly parallel, energy- and area-efficient architectures. Recently, hardware realizations of neurons and synapses using memristive devices were proposed and applied for the task of correlation detection. However, for weakly correlated signals, this task becomes challenging because of the variability and the asymmetric conductance response of the memristive devices. In this paper, we propose a high-density memristive system realized using nanodevices based on phase-change technology. We present a noise-robust phase-change implementation of a neuron, and a synaptic learning rule that is capable of capturing patterns of weakly correlated inputs. We experimentally demonstrate operation with a correlation coefficient as low as 0.2 using a record number of 1M phase-change synapses.
000229322 6531_ $$aNeuromorphic
000229322 6531_ $$aphase-change
000229322 6531_ $$amemristors
000229322 6531_ $$aSTDP
000229322 6531_ $$aneuron
000229322 6531_ $$acorrelation detection
000229322 700__ $$aWozniak, Stanislaw
000229322 700__ $$aPantazi, Angeliki
000229322 700__ $$aSidler, Severin
000229322 700__ $$aPapandreou, Nikolaos
000229322 700__ $$0240162$$aLeblebici, Yusuf$$g112194
000229322 700__ $$aEleftheriou, Evangelos
000229322 773__ $$tIEEE Transactions on Circuits and Systems II: Express Briefs
000229322 909CO $$ooai:infoscience.tind.io:229322$$particle$$pSTI
000229322 909C0 $$0252051$$pLSM$$xU10325
000229322 917Z8 $$x230839
000229322 937__ $$aEPFL-ARTICLE-229322
000229322 973__ $$aEPFL$$rREVIEWED$$sPUBLISHED
000229322 980__ $$aARTICLE