Synaptic elements based on memory devices play an important role in boosting neuromorphic system performance. Here, we show two types of fab-friendly HfO2 material-based resistive memories categorized by configuration and an operating principle for a suitable analog synaptic device aimed at inference and training of neural networks. Since the inference task is mainly related to the number of states from a recognition accuracy perspective, we first demonstrate multilevel cell (MLC) properties of compact two-terminal resistive random-access memory (RRAM). The resistance state can be finely subdivided into an MLC by precisely controlling the evolution of conductive filament constructed by the local movement of oxygen vacancies. Specifically, we investigate how the thickness of the HfO2-switching layer is related to an MLC, which is understood by performing physics-based modeling in MATLAB from a microscopic view. Meanwhile, synaptic devices driven by an interfacial switching mechanism instead of local filamentary dynamics are preferred for training accelerated neuromorphic systems, where the analogous transition of each state ensures high accuracy. Thus, we introduce three-terminal electrochemical random-access memory that facilitates mobile ions across the entire HfO2 switching area uniformly, resulting in highly controllable and gradually tuned current proportional to the amount of migrated ions.
An analogous change in lateral channel current from source to drain in three-terminal synaptic devices actuated by mobile ions vertically provided from a gate can enhance neuromorphic computing performances. We demonstrate a gradually tuned channel current in a fully complementary metal–oxide–semiconductor compatible HfOx/WOx stack with Cu ions. By examining each layer in the three-terminal device, such as the channel, electrolyte, and mobile ion supplier, we identify which device structure can modulate the channel current effectively using mobile ions. Our findings reveal that the gate-tunable channel current response can be solely achieved when the Cu ions are not locally formed but migrate throughout the HfOx electrolyte. The linear dependence of the analog current operation on the channel width further proves the area-switching mechanism. The importance of ion movement can be indirectly verified from the uncontrollable channel currents using either Ag ions with faster mobility than Cu ions or a local path is created because of the thinned HfOx electrolyte.
To enhance the computing efficiency in a neuromorphic architecture, it is important to develop suitable memory devices that can emulate the role of biological synapses. More specifically, not only are multiple conductance states needed to be achieved in the memory but each state is also analogously adjusted by consecutive identical pulses. Recently, electrochemical random-access memory (ECRAM) has been dedicatedly designed to realize the desired synaptic characteristics. Electric-field-driven ion motion through various electrolytes enables the conductance of the ECRAM to be analogously modulated, resulting in a linear and symmetric response. Therefore, the aim of this study is to review recent advances in ECRAM technology from the material and device engineering perspectives. Since controllable mobile ions play an important role in achieving synaptic behavior, the prospect and challenges of ECRAM devices classified according to mobile ion species are discussed.
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 CuOx/HfOx/WOx 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 (ISD) 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 ISD and the selection of an update scheme—for example, column-by-column—that avoided ISD intervention via unselected cells were key to enable energy-efficient neuromorphic computing.
We demonstrate the synaptic characteristics of analogously modulated channel currents in Cu-ion-actuated electrochemical RAM (ECRAM) based on an HfOx electrolyte and a WOx channel. Uncontrolled synaptic response is found as a function of the gate pulse when a Cu-rich gate electrode delivers mobile ions, presumably due to many ions injected from the infinite ion reservoir. As a result, we propose a CuOx oxide electrode to limit ion sources, which is indirectly validated by a physical examination of the degree of chemical bonding between Cu and oxygen, thereby boosting gate controllability over the channel. In addition, the HfOx electrolyte needs to be designed to facilitate the adequate migration of Cu ions, considering thickness and film quality. Using material stack engineering, the channel current of optimized CuOx/HfOx/WOx ECRAM can be steadily tuned via repeated identical gate pulses. The channel current and its change are proportional to the device area and the amount of migrated ions relevant to the gate pulse conditions, respectively. The homogeneous flow of ions across the entire area can, thus, be used to explain the obtained analog switching. The gate-controllable synaptic behavior of the ECRAM accelerates deep neural network training based on backpropagation algorithms. An improved pattern recognition accuracy of ∼88% for handwritten digits is achieved by linearly tuned multiple current states with more than 100 pulses and asymmetric gate voltage conditions in a three-layer neural network validated in simulation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.