Close

Well needed update

A project log for Memristor Computing

Exploring memristor fabrication and design techinques

andrewandrew 09/22/2015 at 15:530 Comments

It's been a while since I've had anything to say. There are a couple of good reasons for this, but more on that later.

First, I'd like to make some observations on memristor technology in general. I took a very long break from working with memristors because most of the research is either blocked by an academic paywall or going straight into military applications.

As you can see in the profile picture for this project, I tried my hand at memristor fabrication on my own. Following the advice of the nice guys in the beam robotics mailing list, I oxidized copper and connected it with an aluminum foil terminal. All of the devices show worked for about a week, until I burnt them out. One interesting effect that observed was that each memristor device would oscillate at a descending frequency when attached to an amplifier. Aside from demonstrating the memristor effect, these devices were useless. Every professionally made memristor that I've seen (I am no expert, I am probably incorrect with this information) doesn't have a solid reliability rate. Several papers have asserted that the true memristor has not been found yet, and may be physically impossible. I'm not sure about impossible, but I can't imagine this technology diffusing into the public sector until several improvements are made to the memristor fabrication process. This hampers my own research, as I can only make designs on paper and simulations in LTspice. I'm confident that the field will expand and mature, much as transistors did in the early 60s.

Finally, I realize that I've yet to make any substantial publicaitons of my own designs. This has been partially because of my busy schedule in college, but mostly out of sheer laziness. Additionally, I can never find a satisfactory settling point for my own work. I published the digital timing set mainly because my own design is based around ternary logic. More recently, I've been researching reversible computing and see a lot of potential for it to mesh with my own work. In this week I will publish an overview paper of my work.

Until next time!

Discussions