Vanadium Dioxide & Associative Processing


Vanadium dioxide has some unusual properties, and it’s actually able to transition between being a semiconductor and a full conductor. That has begun to open up some interesting possibilities as far as material engineers are concerned. Penn State engineers recently developed an oscillating switch that uses the compound. Due to its ability to fluctuate between the two states, the substance was perfect as far as making a switch was concerned.

One of the doctoral students involved in the project then used a bit of an engineering trick to add a resistor in series with the switch in order to bring some needed stability to the reaction. Stabilizing the circuit still took a great number of cycles, but it was eventually accomplished.

A second system was installed alongside the first and it was suggested that this composite device could serve as the central processing unit for a non-Boolean computer system. Similar systems are seen in various types of organisms. For instance, lightning bugs flash at somewhat random rates but eventually become stabilized. Biological synchronization has existed since time immemorial, but thus far engineers haven’t really been able to harness it.

As a result this opens up an entire new field of associative processing. Instead of using digital data encoding, information could be stored via an array of infinitesimally small oscillators. The oscillators themselves then would align themselves in a certain pattern to store data. When someone wishes to re-sequence them they would eliminate their existing data in favor of the new matrix provided by a software algorithm.

In this respect, these circuits wouldn’t act too differently from existing magnetic and semiconductor based storage systems. However, they would be considerably smaller and able to process things in way that digital circuits have never been able to. As a result these devices could instead handle a variety of unique problems.

Even though it will be quite a while before associative processing makes a splash in the consumer market, it could be impressive as far as solving mathematical equations is concerned. Environmental models could certainly benefit from this kind of technology, and the complex calculations involved in tracking spacecraft might also be an attractive application. Since the computer functions would actually be closer to biological ones, some people might immediately start to think of some interesting cybernetic applications that are outside of the original researcher’s scope.

Considering the manner in which these systems work, it might be possible to additionally construct impressive learning structures as well. These devices would emulate neural networks far more accurately, since digital computers merely simulate logic. Non-Boolean devices could actually think in a logical fashion, and that’s a huge step forward as far as solving mathematical equations is concerned.

Post Navigation