Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
I do not think there are any reliable numerical models for memristor-type devices (solid state electrolytes, phase change memories, etc.) that have been implemented in any commercial device simulator (Atlas, Medici, Sentaurus-Device, etc.).
Remember Silvaco has a C interface to allow yopu to describe your own physics models. For solid electrolytes this would no be too difficult but would require detailed knowledge of the Fab process characterisation that was doing the memristor.
The C interface allows to define some physical models (like mobility etc.) within a rigid set of equations - drift-diffusion equations (current continuity equations for electrons and holes and Poisson equation) or quasi-hydrodynamic equations (plus maybe heat transport equation for lattice temperature). If your model does not fit within drift-diffusion - there is no chance to make use of Atlas or similar tools.
The biggest problem with "memristors" is the lack of first-principle models or theory. There are only empirical models, and the "chemical" or "process" related effects are very difficult if possible at all to capture in models that would be predictive.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
