Spiking neurons with spatiotemporal dynamics and gain modulation for monolithically integrated memristive neural networks

Abstract

As a key building block of biological cortex, neurons are powerful information processing units and can achieve highly complex nonlinear computations even in individual cells. Hardware implementation of artificial neurons with similar capability is of great significance for the construction of intelligent, neuromorphic systems. Here, we demonstrate an artificial neuron based on NbO x volatile memristor that not only realizes traditional all-or-nothing, threshold-driven spiking and spatiotemporal integration, but also enables dynamic logic including XOR function that is not linearly separable and multiplicative gain modulation among different dendritic inputs, therefore surpassing neuronal functions described by a simple point neuron model. A monolithically integrated 4 × 4 fully memristive neural network consisting of volatile NbO x memristor based neurons and nonvolatile TaO x memristor based synapses in a single crossbar array is experimentally demonstrated, showing capability in pattern recognition through online learning using a simplified δ-rule and coincidence detection, which paves the way for bio-inspired intelligent systems.

Ferroelectric Synaptic Transistor Network for Associative Memory

“…Secondly, despite of a high memory density using 2D semiconductors, the technique of large‐area 2D semiconductors with uniform electronic properties is still unmatured, which limits the yield and lateral size of arrays. These issues we believe will be conquered as the development of artificial neurons [ 60 ] and engineering of 2D semiconductors.…”

Multi-bifurcation cascaded bursting oscillations and mechanism in a novel 3D non-autonomous circuit system with parametric and external excitation

“…Letting a = 10, b = 2, c = 1, β 1 = β 2 = 3, f 1 = 0.003 and f 2 = 0.001, then w 2 = g(t), w 1 can be replaced by 4g(t) 3 − 3g(t) based on Table 2. Next, the corresponding equilibrium points E 0 and E ± are substituted into (11), and the following characteristic equation can be obtained as,…”

“…For example, researchers in [10] used several types of BOs to drive electrodynamical systems. Authors in [11] demonstrated a specific type of artificial neuron with threshold-driven spiking and built a neural network for pattern recognition. Furthermore, it can also be applied to the study of stability and control design of pulsed power supply systems according to similar oscillation characteristics [12].…”

A Single Transistor‐Based Threshold Switch for a Bio‐Inspired Reconfigurable Threshold Logic

“…[12][13][14][15] In order to circumvent the limitations of complementary metal-oxide-semiconductor (CMOS) circuit-based threshold logic, a few types of single devices that can provide threshold operation were suggested using a magnetic tunnel junction (MTJ) and metal-insulator transistor (MIT) material. [16,17] However, only three Boolean logic functions were accomplished. In addition, these devices require additional materials that may not be easily implemented for conventional CMOS fabrication.…”