Controlled Majority-Inverter Graph Logic With Highly Nonlinear, Self-Rectifying Memristor
Run Ni, Ling Yang, Xiaodi Huang, Sheng‐Guang Ren, Tianqing Wan, Yi Li, Xiangshui Miao
Abstract
In this article, for the first time, self-rectifying memristors are exploited for logic-in-memory computation. We report a Pt/TaO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><i>x</i></sub> /Ta memristor with salient self-rectifying bipolar features (10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sup> ON-/ OFF-ratio, 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">5</sup> rectification ratio, 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">5</sup> nonlinearity, and ~1 pA leakage current), which could support a large passive crossbar array up to 160 Mb with the premise of 10% read margin. Moreover, we propose and experimentally validate a controlled majority-inverter graph logic method based on the self-rectifying switching behaviors, with advantages in computation complexity. Our work is a step forward toward in-memory computing in high-density or even 3-D memristor architectures.