Survey on Memrister Models S. Singh 1 , P.W.C. Prasad 1 , Abeer Alsadoon 1 , A. Beg 2 , L. Pham 1 , A. Elchouemi 3 1 School of Computing and Mathematics, Charles Sturt University, Sydney, Australia 2 College of Information Technology, United Arab Emirates University, Al Ain, U.A.E. 3 Hewlett Packard Enterprise Abstract— While hardware in a computer have developed greatly, users still has faced problems with its speed, and memory in terms of its performance. The recent developments in memristors made it possible to reduce the problems, as memristive models have been be designed to suit the requirements of time. However, different characteristics are expected from memristors depending upon its applications. The paper aims to compare three major models of memristors focusing on their advantages and limitations. It identifies the most suitable model of memristor that satisfies the memristive device conditions. Out of the three models, Voltage threshold adaptive memristor model (VTEAM) fits into the requirements and it has sufficient accuracy and computational efficiency. Keywords— Memristors, VTEAM, Threshold Adaptive Memristor Model, Boundary Condition Model I. INTRODUCTION A. Background While using a computer has been prevalent, it still has issues with speed, time and memory. Memristors are electrical devices designed to address the problems. Memristors, as fourth passive element, were not known until 1971 when a professor Leon Chua reasoned the existence of one more passive element as one missing link between four circuit variables i.e. magnetic flux and charge. He named that element as memristor essentially a resistor with memory. This is because memristor is discovered after resistors, capacitors and inductors as a fundamental electronic passive element of circuit design. These elements show their distinct characteristics mainly at nanoscale. Memristor is a concatenation of memory and resistor characterized by the relationship between time integrals of current and voltage across two terminal device [1]. This function is similar to variable resistance called memristance. B. Application With recent developments in memristors, it is clear that it has vast applications in future as a potential building block of hardware. Memristors can be used in several applications like neuromorphic systems, logic, memory and analog circuits. In memristive circuits, memristor behaviour must be mathematically proven, sufficiently accurate and computationally efficient. Many research on memristor applications in various fields using alternative materials is seen in recent literatures. Memristors can be used as non-volatile memory, low-power and remote sensing applications, cross bar latches used as transistors or augmenters, analog computation and circuit applications and mimic neuromorphic circuits and biological systems, but as to how far this is reality depends on future developments [2]. The memristor circuit modelling appeared in literatures are few to mention such as SPICE macro-modelling using linear and non-linear drift ion models [3], Programmable fine resolution resistor that uses memristor [4], Flexible memristors used as non-volatile memory(inexpensive and low power device) [5], digital logic implementation of crossbar architecture using memristor [6] and many more. C. Purpose of Research Even though there has been so much advancement in the domain, there are still many problems with memristors. For example slow switching speed and leakage mechanism is not clear [7]. For memristor as linear circuit element, there is no mathematical difference between resistor and memristor and non-linear memristor have same unit as resistor unlike capacitors and inductors, this behaviour is similar to non- linear resistors having memory affects as dynamic generalization but not “fourth fundamental circuit element” [8]. It has to be noted that physically realised memristor must satisfy the mathematical requirements of memristive system [9]. Memristor models have been proposed and these models can be used in EDA tools like SPICE. The aim of the research is to survey the existing major models of memristors and find the most suitable and appropriate memristive device circuit. The underlying outcomes of the report will help to improve performance of the memory devices. D. Layout In the first chapter, a brief introduction about the memristors has been discussed along with the purpose of the research which is to survey the existing major models of memristors and find the most suitable and appropriate memristive device circuit. The next chapter discusses the literature review. The past work of researchers in this domain will be discussed to understand in depth about this devise. The third chapter is the implementation phase where the settings and the configuration statistics are defined. These parameters were used to conduct the experiment whose results are displayed in chapter four. The fifth chapter is the conclusion and future works. II. STATE OF THE ART A. Development In 1960, Bernard Widrow and Ted Hoff from Stanford University developed three terminal circuit element called ‘memristor’ where resistance was controlled by electric charge