基于ZnO忆阻器的神经突触仿生电子器件

本文采用ZnO忆阻器模拟了生物神经突触的记忆和学习功能。ZnO突触器件表现出典型的随时间指数衰减的突触后兴奋电流(EPSC),以及EPSC的双脉冲增强行为。在此基础上,实现了学习-遗忘-再学习的经验式学习行为,以及四种不同种类的电脉冲时刻依赖可塑性学习规则。ZnO突触器件实现了超低能耗操作,单次突触行为能耗最低为1.6pJ,表明其可以用来构筑未来的人工神经网络硬件系统,最终开发出与人脑结构类似的认知型计算机以及类人机器人。     

Control of Switching Modes and Conductance Quantization in Oxygen Engineered HfOx based Memristive Devices

Hafnium oxide (HfO_x)‐based memristive devices have tremendous potential as nonvolatile resistive random access memory (RRAM) and in neuromorphic electronics. Despite its seemingly simple two‐terminal structure, a myriad of RRAM devices reported in the rapidly growing literature exhibit rather complex resistive switching behaviors. Using Pt/HfO_x/TiN‐based metal–insulator–metal structures as model systems, it is shown that a well‐controlled oxygen stoichiometry governs the filament formation and the occurrence of multiple switching modes. The oxygen vacancy concentration is found to be the key factor in manipulating the balance between electric field and Joule heating during formation, rupture (reset), and reformation (set) of the conductive filaments in the dielectric. In addition, the engineering of oxygen vacancies stabilizes atomic size filament constrictions exhibiting integer and half‐integer conductance quantization at room temperature during set and reset. Identifying the materials conditions of different switching modes and conductance quantization contributes to a unified switching model correlating structural and functional properties of RRAM materials. The possibility to engineer the oxygen stoichiometry in HfO_x will allow creating quantum point contacts with multiple conductance quanta as a first step toward multilevel memristive quantum devices.     

可记忆电阻器及其应用展望综述

可记忆电阻器,本文简称为忆阻器,能够不需要功率源而储存信息,是一种纳米尺寸的器件,它于1971年被Leon Chua作为一种"丢失的电路元件"提出,2008年被惠普实验室研究人员Williams R S等证明的确存在,成为电路理论里电阻器、电容器和电感器以外的第四个基本电路元件,极有可能引领电子学的一次重要变革。本文回顾了忆阻器的概念和数学定义,重点介绍了惠普实验室的Pt/TiO2/Pt三明治结构的忆阻器薄膜器件模型和忆阻器元件某些值得关注的特性,如滞回曲线特性,最后探讨了可记忆元件在纳米级器件领域的应用前景和研究方向。     

A programmable ternary CPU using hybrid CMOS/memristor circuits

Abstract

The carry propagation of arithmetic operations is one of the major shortcomings of common binary number encodings as the two’s complement. Signed-digit arithmetic allows the addition of two numbers without carry propagation and in asymptotically constant time in dependence of the word length, while at the same time requiring a digit representation with more than two states. With the advent of memristors, it has become possible to store multiple states within a single memory cell. This paper proposes an implementation of a general purpose CPU using signed-digit arithmetic by exploiting memristors in order to implement multi-value registers. The proposed model of the CPU is evaluated by the execution of various image processing algorithms. It is shown that a break-even point exists at which signed-digit algorithms outperform conventional binary arithmetic operations. Furthermore, simulation results prove that the memristor device lends itself to store signed-digit data efficiently.     

A Habituation Sensory Nervous System with Memristors

The sensory nervous system (SNS) builds up the association between external stimuli and the response of organisms. In this system, habituation is a fundamental characteristic that filters out irrelevantly repetitive information and makes the SNS adapt to the external environment. To emulate this critical process in electronic devices, a Li_xSiO_y-based memristor (TiN/Li_xSiO_y/Pt) is developed where the temporal response under repetitive stimulation is similar to that of habituation. By connecting this synaptic device to a leaky integrate-and-fire neuron based on a Ag/SiO₂:Ag/Au memristor, a fully memristive SNS with habituation is experimentally demonstrated. Finally, a habituation spiking neural network based on the SNS is built and its application in obstacle avoidance for robot navigation is successfully presented. The results provide that a direct emulation of the biologically inspired learning process by memristors could be a sound choice for neuromorphic hardware implementation.     

SPICE modeling of nonlinear memristive behavior

The recent discovery of the ‘modern’ memristor has drawn great attention of both academia and industry. Given their favorable performance merits, memristors are expected to play a fundamental role in electronic industry. Modeling of memristive devices is essential for circuit design, and a number of Simulation Program with Integrated Circuit Emphasis (SPICE) models have already been introduced. The common problem in most models is that there is no threshold consideration; hence, only a few address the nonlinear nature of the device. This paper aims to present a SPICE implementation of a threshold‐type switching model of a voltage‐controlled memristive device that attributes the switching effect to a tunneling distance modulation. Threshold‐type switching is closer to the actual behavior of most experimentally realizable memristive systems, and our modeling approach addresses the issue of programming thresholds. Both the netlist and the simple schematic are provided, thus making it easy to comprehend and ready to be used. Compared with other modeling solutions, it involves significantly low‐complexity operation under an unlimited set of frequencies, and its simulation results are in good qualitative and quantitative agreement with the theoretical formulation. The proposed model is used to simulate an antiserial memristive switch, proving that it can be efficiently introduced in complex memristive circuits. Copyright © 2013 John Wiley & Sons, Ltd.