Emerging Memristive Artificial Neurons for Energy-efficient Neuromorphic Electronic Systems

Abstract

Artificial neurons are fundamental building blocks for implementing neuromorphic electronic systems capable of highly efficient computing processing and cognitive functions. Among the various device platforms, emerging memristive switching devices have been widely utilized and have been suggested as artificial neurons. In this chapter, we introduce and describe emerging memristive artificial neurons in terms of their neuronal functionalities, switching mechanisms, and computing applications. We explain the basic concept of memristive neurons, focusing on structural and functional similarities, present several representative types of memristive neurons and fundamental switching mechanisms for neuronal dynamics, and provide an overview of several demonstrations of artificial neural networks based on the fabricated memristive neurons. This chapter focuses on short guidelines for memristive neurons and their neuromorphic computing applications.