Communications of the ACM
Next-Gen Computing: Memristor Chips That See Patterns Over Pixels
Researchers at the University of Michigan say they have developed a memristor computer circuit prototype that has the potential to process images, videos, and other complex data faster and using less power than today's most advanced systems.
Credit: Wei Lu
University of Michigan (U-M) researchers say they have developed a memristor computer circuit prototype that has the potential to process complex data, such as images and videos, faster and using much less power than today's most advanced systems.
The prototype uses pattern recognition to shortcut the energy-intensive process conventional systems use to dissect images.
The system uses an algorithm that relies on "sparse coding" to instruct the 32-by-32 array of memristors to efficiently analyze and recreate several photos.
Memristors can simultaneously perform many operations without having to move data around, which means they could enable new platforms that process a vast number of signals in parallel and are capable of advanced machine learning.
The U-M system is designed to detect patterns very efficiently, and to use as few features as possible to describe the original input.
The researchers also note it could be scaled up to process and analyze video in real time.
From University of Michigan News
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found