A ZnO Nanowire Based Memristive Device and Its Application in Artificial Synapse

With the advent of the Big Data era, there is a growing demand for artificial intelligence, the Internet of Things, and machine learning. However, traditional computers based on von Neumann architecture face a significant challenge in the processing of vast amounts of data due to the separation of memory and the central processing unit. The hardware-based neuromorphic computing, which is inspired by the human brain, has the potential to significantly reduce energy consumption and realize a wide range of applications. Although a great achievement has been obtained, the uniformity in the electrical characteristics is still a challenge. Therefore, we designed and fabricated a memristor using ZnO nanowires (NWs) to achieve consistent modulation of conductive filaments. This memristor can emulate various synaptic functions, including paired-pulse facilitation, and short-term and long-term plasticity. As a proof of concept, the convolutional neural network simulation for MNIST recognition with 98.87% accuracy was achieved by the ZnO NWs-based memristors.