Navigate Up
Sign In

The University of Tennessee, Knoxville

Big Ideas for Brain-Inspired Nanocircuits

Garrett RoseEngineering

Can you imagine machines with the computational power of the brain (some brains at least) implemented at a scale smaller than the diameter of a human hair? While such extreme examples are unlikely in the near future, a path is unfolding toward brain-inspired computers built from electronic devices fabricated at nanometer scales. Specifically, nanoelectronic memristors (or “memory resistors”) are considered for new forms of low-power, high-density computing. In this talk, we will discuss research aimed at implementing brain-inspired systems where the memristors act as artificial synapses, a key functional element in any neural network. Results thus far have been promising in that we can model and simulate networks trained for application areas such as pattern recognition and even basic control. While we aren't building brains, not yet at least, such power-efficient brain-inspired systems are expected to enable a range of applications not easily possible with conventional computers.

Watch Presentation