Reservoir computing is a highly efficient network for processing temporal signals due to its low training cost compared to standard recurrent neural networks, and generating rich reservoir states is critical in the hardware implementation. In this work, we report a parallel dynamic memristor-based reservoir computing system by applying a controllable mask process, in which the critical parameters, including state richness, feedback strength and input scaling, can be tuned by changing the mask length and the range of input signal. Our system achieves a low word error rate of 0.4% in the spoken-digit recognition and low normalized root mean square error of 0.046 in the time-series prediction of the Hénon map, which outperforms most existing hardware-based reservoir computing systems and also software-based one in the Hénon map prediction task. Our work could pave the road towards high-efficiency memristor-based reservoir computing systems to handle more complex temporal tasks in the future.
Exploring new type of synapse–like electronic devices with fusion of computing and memory is a promising strategy to fundamentally approach to intelligent machines. Herein, organic thin film memristors (OTFMs) are achieved, functioning as electrically programmable and erasable analog memory with continuous and nonvolatile device states. The memristive characteristics of OTFMs stem from the asymmetric electrode configuration and the cumulative charge trapping/detrapping in a polymer electret layer, which enables the state–dependent current modulation analogous to the synaptic weight change in biological synapses. OTFMs are demonstrated to successfully emulate the essential synaptic functions, including the reversible potentiation and depression, and the short‐term plasticity such as the paired‐pulse facilitation and the long‐term plasticity such as the spike–timing dependent plasticity.
Implementing power-e cient reservoir computing hardware systems is of great interest to the eld of neuromorphic computing. More and more studies attempt to use analog devices or components, such as memristors and spintronic oscillators, to partially replace fully digital systems to boost the power e ciency. However, a reservoir computing system operating real-time in fully analog fashion has not been demonstrated yet. In this work, a fully analog reservoir computing system was implemented using two types of memristors, where dynamic memristors were used to construct the reservoir while nonvolatile memristor arrays were used as the readout layer. The key features, such as threshold and window, extracted from the dynamic memristor-based physical nodes were found to have a signi cant impact on the system performance. By adjusting the features to the appropriate range, our system can e ciently process spatiotemporal signals in real time with extremely low power consumption, more than three orders of magnitudes lower than digital counterparts. Both temporal task of arrhythm detection and spatiotemporal task of dynamic gesture recognition were demonstrated, where high detection accuracy of 96.6% and recognition accuracy 97.9% were achieved respectively. Our work demonstrates that such memristor-based fully analog reservoir computing system could be attractive for spatial and temporal edge computing with extremely low power and hardware cost.
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.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.