AUTHOR=Kang Heebum , Kim Nayeon , Jeon Seonuk , Kim Hyun Wook , Hong Eunryeong , Kim Seyoung , Woo Jiyong 

TITLE=Analysis of electro-chemical RAM synaptic array for energy-efficient weight update

JOURNAL=Frontiers in Nanotechnology

VOLUME=4

YEAR=2022

URL=https://www.frontiersin.org/journals/nanotechnology/articles/10.3389/fnano.2022.1034357

DOI=10.3389/fnano.2022.1034357

ISSN=2673-3013

ABSTRACT=<p>While electro-chemical RAM (ECRAM)-based cross-point synaptic arrays are considered to be promising candidates for energy-efficient neural network computational hardware, array-level analyses to achieve energy-efficient update operations have not yet been performed. In this work, we fabricated CuO<sub>x</sub>/HfO<sub>x</sub>/WO<sub>x</sub> ECRAM arrays and demonstrated linear and symmetrical weight update capabilities in both fully parallel and sequential update operations. Based on the experimental measurements, we showed that the source-drain leakage current (I<sub>SD</sub>) through the unselected ECRAM cells and resultant energy consumption—which had been neglected thus far—contributed a large portion to the total update energy. We showed that both device engineering to reduce I<sub>SD</sub> and the selection of an update scheme—for example, column-by-column—that avoided I<sub>SD</sub> intervention <italic>via</italic> unselected cells were key to enable energy-efficient neuromorphic computing.</p>