Non-linear Reconfigurable Threshold Logic Gates Based on Nanostructured Metallic films

Abstract The dominion of Boolean logic circuits (BLCs) in current electronic design and synthesis tools is increasingly being called into question by challenges posed by the extreme miniaturization and energy footprint typical of AI applications. Compared to BLCs, Threshold Logic Gates (TLGs), by programming the weights, allow for more than two input states, which can enhance complexity in simplifying circuit design. Memristors organized in crossbar arrays architectures have been proposed to implement the programmable weights of threshold logic elements. An alternative TLG design, called Receptron, has been recently proposed: it is based on nonlinear weights thus widening the spectrum of Boolean computable functions while simplifying training thanks to a random search protocol. Here we present a theoretical and an experimental characterization of the Receptron model to determine the connection between the structure of the weights and function computability, identifying sub-linearity as an enabling feature. These results allowed the fabrication of an improved version of a Receptron device, enhancing its sub-linearity and random search efficiency thanks to an ad hoc circuit. This nonlinear threshold logic gate device can be considered for the integration with other conventional logic components for higher levels of computational complexity. Corresponding authors: [email protected]; [email protected]