Hybrid architecture based on two-dimensional memristor crossbar array and CMOS integrated circuit for edge computing

Abstract The fabrication of integrated circuits (ICs) employing two-dimensional (2D) materials is a major goal semiconductor industry for the next decade, as it may allow extension Moore’s law, aids in in-memory computing and enables advanced devices beyond conventional complementary metal-oxide-semiconductor (CMOS) technology. However, most circuital demonstrations so far utilizing 2D employ methods such mechanical exfoliation that are not up-scalable wafer-level fabrication, their application could achieve only simple functionalities logic gates. Here, we present crossbar array memristors using multilayer hexagonal boron nitride (h-BN) dielectric, exhibit analog bipolar resistive switching >96% devices, which ideal implementation multi-state memory element neural networks, edge machine learning applications. Instead this memristive to solve logical problem, here go step combination h-BN with CMOS circuitry implement extreme (ELM) algorithm. circuit used design encoder unit, based decoder functionality. proposed hybrid architecture demonstrated complex audio, image, other non-linear classification tasks on real-time datasets.