Sign In

Communications of the ACM

ACM TechNews

Memristor Minds: The Future of Artificial Intelligence


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
slime mold

Slime mold feeding on the surface of an almond. These organisms could be the missing link in memory circuits.

Credit: Eye of Science / Science Photo Library

The lack of a rigorous mathematical foundation for electronics impelled engineer Leon Chua to develop one, which led to the formulation of the memristor — a theoretical fourth basic circuit element in addition to the resistor, capacitor, and inductor where electric charge and magnetic flux come together. Since then the creation of memristors has been achieved, and their novel abilities might unlock key insights about the human brain that would be a tremendous step forward for the field of artificial intelligence.

Advantages of memristors include rapid, nanosecond writing of data using a very small amount of energy, and retention of memristive memory even when the power is turned off. The most immediate potential application for memristors is as a flash memory replacement, while durability improvements should make memristors ideal for a superfast random access memory, says Hewlett-Packard (HP) Laboratories Fellow Stan Williams.

a memristor never forgets

The discovery that a slime mold was behaving in the manner of a memristive circuit in that it could memorize a pattern of events without the aid of a neuron inspired a physicist at the University of California, San Diego to construct a circuit capable of learning and predicting future signals. Much earlier, Chua had noticed a sharp similarity between synapse behavior and memristor response, leading to speculation that memristors might help engineer an electronic intelligence that can mimic the power of a brain.

The U.S. Defense Advanced Research Projects Agency has embarked on a project to create "electronic neuromorphic machine technology that is scalable to biological levels." HP's Greg Snider has envisioned the field of cortical computing that focuses on the potential of memristors to imitate the interaction of the brain's neurons. He and Williams are working with Boston University scientists to devise hybrid transistor-memristor chips that aim to replicate some of the brain's thought processes.

From New Scientist
View Full Article

 

Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA


 

No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account