Communications of the ACM
Nice! The Brain as a Model For Future Supercomputers
The Sandia National Laboratories-sponsored Neuro-Inspired Computational Elements (NICE) workshop is aimed at finding ways to tap the brain's power to transmit signals along massively parallel channels.
Credit: agsandrew/Shutterstock
Finding ways to tap the brain's ability to transmit signals along massively parallel channels, with multiple convergences at downstream nodes, to accommodate rapidly changing, high-volume data is the purpose of the Sandia National Laboratories-sponsored Neuro-Inspired Computational Elements (NICE) workshop.
NICE is envisioned as a way to continue the advancement of computing power once the barrier for shrinking circuits is reached.
Workshop participants proposed isolating brain tissues that govern aspects of behavior, analyzing the shape and behavior of the neurons sending the signals, and copying that configuration using conventional hardware and software or, most likely, a new solid-state substrate.
Possible domain intersections cited by Sandia's Rob Leland include tissue-based and in-vivo sensors, optical nanosensors for chemical analysis within cells, regulated nanoassembly of circuits, digital antibodies, and virus-sized logic chips.
"Brains are highly parallel, can reconfigure themselves dynamically in a few minutes, and use molecular signal transduction [to pass messages]," notes George Mason University's Jim Olds. "In message-passing they use little power and finesse around bottlenecks [that would slow silicon-based] parallel-computing systems."
Still, Sandia CIO Mike Vahle cautions that pattern-matching brain function to computing function may give rise to a host of ethical, cultural, and security issues.
From Sandia National Laboratories
View Full Article
Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA
No entries found