This page is no longer being updated and remains online for informational and historical purposes only. The information is accurate as of the last page update.
For questions about page contents, contact the Communications Division.
Electrical and computer engineering major writes about his work with Todd Wey, assistant professor of electrical and computer engineering
Serdar Benderli ’09 (Antalya, Turkey) is researching possible new uses for memristors, electrical devices which can “remember” changes in current passing through them by changing their level of resistance. This helps improve the performance of devices as they are scaled down to smaller and smaller sizes. Benderli, an electrical and computer engineering major, is working on his honors thesis under the guidance of Todd Wey, assistant professor of electrical and computer engineering.
Memristors are a new breed of electrical devices. They are essentially resistors with memory. The well-established electrical theory acknowledges three basic elements: resistor, capacitor, and inductor. The memristor was theorized in 1971 to be the fourth circuit element. However, it stayed only a theory until HP Labs physically discovered it in 2008 thanks to advances in nanotechnology.
So far, engineers are working on two suggested uses for memristors: digital and analog. The difference between these two domains is how data is processed. In the digital domain, data are processed discretely as 1s and 0s (i.e. computers). The analog domain data are continuous and can have any value.
In the digital domain, memristors are going to be used as switches in computer memory, much like the transistors we have now. However, the advantage of memristors is that they retain their memory even after power is turned off and this may ultimately result in instant-on computers. HP engineers have already announced that they built prototypes of such memories and expect these to hit the market within five years.
However, the real breakthroughs will happen in the analog domain. For the first time, we have access to memory there. We can now design new types of topologies (the shapes of local-area networks or other communications systems) that were impossible to create before.
Professor Wey and I are exploring new topologies where memristors may prove to be useful. During the fall semester, we created a model of HP’s memristors that we are now using to simulate and analyze other circuit topologies in a software environment. All this analysis has to be in a software environment, as we do not yet have access to the physical devices.
We already have sent a letter about our model to The Institution of Engineering and Technology and expect to be published. We hope to come up with some useful topologies this semester and publish more of our results.