Scientific Lectures //
Neuro-Inspired Computational Engines: Beyond Von Neumann/Turing Architecture and Moore’s Law Limits
Murat Okandan, Ph.D. - Principal Member of Technical Staff, Microsystems S&T, Sandia National Laboratories
Presented: January 23, 2015
ABSTRACT: There is an evolving convergence among seemingly disparate disciplines: neuroscience, microelectronics and computing. Moore's Law has been at the core of 4 decades of incredible advances in technology and productivity, however, after 3 distinct saturation trends, brute-force microelectronic device scaling appears to be no longer a viable path for continued computing performance improvements.
Neural systems have provided inspiration for computing systems from the very early days, but the technology development efforts very quickly found a more deterministic and tractable path in conventional computing machines.
This talk will present an approach that has been attracting more attention (for example, see NICE workshop, http://nice.sandia.gov): neuro-inspired computational systems.
I will highlight why we need to seriously consider this new path, how we might be able to architect these systems and what they might allow us to do that we cannot efficiently do with conventional computing systems.
To view this presentation please click here.