A Native SPICE Implementation of Memristor Models for Simulation of Neuromorphic Analog Signal Processing Circuits

Since the memristor emerged as a programmable analog storage device, it has stimulated research on the design of analog/mixed-signal circuits with the memristor as the enabler of in-memory computation. Due to the difficulty in evaluating the circuit-level nonidealities of both memristors and CMOS devices, SPICE-accuracy simulation tools are necessary for perfecting the art of neuromorphic analog/mixed-signal circuit design. This article is dedicated to a native SPICE implementation of the memristor device models published in the open literature and develops case studies of applying such a circuit simulation with MOSFET models to study how device-level imperfections can make adversarial effects on the analog circuits that implement neuromorphic analog signal processing. Methods on memristor stamping in the framework of modified nodal analysis formulation are presented, and implementation results are reported. Furthermore, functional simulations on neuromorphic signal processing circuits including memristors and CMOS devices are carried out to validate the effectiveness of the native SPICE implementation of memristor models from the perspectives of simulation accuracy, efficiency, and convergence for large-scale simulation tasks.

2D materials–based homogeneous transistor-memory architecture for neuromorphic hardware

In neuromorphic hardware, peripheral circuits and memories based on heterogeneous devices are generally physically separated. Thus exploring homogeneous devices for these components is an important issue for improving module integration and resistance matching. Inspired by ferroelectric proximity effect on two-dimensional materials, we present a tungsten diselenide-on-LiNbO3 cascaded architecture as a basic device that functions as a nonlinear transistor, assisting the design of operational amplifiers for analog signal processing (ASP). It also functions as a nonvolatile memory cell, achieving memory operating (MO). Based on this homogeneous architecture, an analog signal processing-memory operating integrated system for binary classification and the design of ternary content-addressable memory were investigated for potential use in neuromorphic hardware.

Analog signal processing through space-time digital metasurfaces

In the quest to realize analog signal processing using subwavelength metasurfaces, in this paper, we present the first demonstration of programmable time-modulated metasurface processors based on the key properties of spatial Fourier transformation. Exploiting space-time coding strategy enables local, independent, and real-time engineering of not only amplitude but also phase profile of the contributing reflective digital meta-atoms at both central and harmonic frequencies. Several illustrative examples are demonstrated to show that the proposed multifunctional calculus metasurface is capable of implementing a large class of useful mathematical operators, including 1st- and 2nd-order spatial differentiation, 1st-order spatial integration, and integro-differential equation solving accompanied by frequency conversions. Unlike the recent proposals based on the Green’s function (GF) method, the designed time-modulated signal processor effectively operates for input signals containing wide spatial frequency bandwidths with an acceptable gain level. Proof-of-principle simulations are also reported to demonstrate the successful realization of image processing functions like edge detection. This time-varying wave-based computing system can set the direction for future developments of programmable metasurfaces with highly promising applications in ultrafast equation solving, real-time and continuous signal processing, and imaging.