Fast and robust memristive cell based on 2D materials for energy-efficient in-memory computing
Year: 2026
Type: software
Abstract: Simulation and training code for binary neural networks (BNNs) deployed on 1T1R memristor arrays, including two-stage training with a 0/1-weight memristor crossbar classifier, robustness evaluation under noise, quantization and sparsity, and circuit-level 1T1R array simulation. This is the version of the code used in the associated Nature Communications paper.
Authors: Zhao, Ruoyu, Hong, Yuwan
Institutions: City University of Hong Kong, Tsinghua University
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7214987932