IBM Scientists Say Radical New 'In-Memory' Architecture Will Speed Up Computers

DooKey

DooKey

[H]F Junkie
Joined
Apr 25, 2001
Messages
13,552
IBM research has announced a new computing architecture called "in-memory" computing. The concept is to use one device for both storing and processing information. They believe this new prototype technology will enable ultra-dense, low-power, and massively parallel computing systems that are especially useful for AI applications. Sounds to me like the neural net processor might become a reality. T-800 cpu anyone?

"This is an important step forward in our research of the physics of AI, which explores new hardware materials, devices and architectures," says Evangelos Eleftheriou, PhD, an IBM Fellow and co-author of an open-access paper in the peer-reviewed journal Nature Communications. "As the CMOS scaling laws break down because of technological limits, a radical departure from the processor-memory dichotomy is needed to circumvent the limitations of today’s computers."
 
Sounds great. Hope they can actually make it happen so we can argue about its ability to crank out frames.
 
Last edited:
And batteries that charge in 30 seconds and last 20 years... These announcements are great and all but a lot of this stuff never hits the consumer level or gets released. Intel's Optane sounded like the greatest thing to happen to SSD's and it's been pretty boring so far.
 
Micron has had a similar solution for years (though I don't know of anyone actually using it, which is a big area of consternation for MU investors) and recently (and quietly) spun the tech off into its own company, http://www.naturalsemi.com
 
You must log in or register to reply here.
Back
Top