What is it about?
Memristors are emerging structures that can act as synapses in brain-inspired computing systems (in other words, hardware neural networks). However, they possess intrinsic stochasticity, which is a major obstacle to their successful utilization. This paper focuses on the improvement of a promising organic memristor via the incorporation of metal nanoparticles into its polymer layer. The achieved improvements are explained in detail. The obtained memristors are introduced to a neural network simulation for efficient prediction of heart diseases.
Featured Image
Photo by Robina Weermeijer on Unsplash
Why is it important?
Traditional computing systems are limited for artificial intelligence tasks due to the separation of their computing and storage parts. It leads to a long transfer of information from one part of the system to another, which leads to an increase of time and energy consumption. This problem is solved by brain-inspired neuromorphic computing systems. Such systems based on transistors have already demonstrated high efficiency and fast calculations. However, transistors are not optimal components to mimic synapses. New structures, such as memristors, are of high interest for the construction of brain-inspired systems.
Perspectives
The improved organic memristors that are demonstrated in the manuscript may be used in future biocompatible computing systems, wearable electronics, prosthetics, etc.
Anna Matsukatova
Read the Original
This page is a summary of: Scalable nanocomposite parylene-based memristors: Multifilamentary resistive switching and neuromorphic applications, Nano Research, October 2022, Tsinghua University Press,
DOI: 10.1007/s12274-022-5027-6.
You can read the full text:
Contributors
The following have contributed to this page
