Featured Articles

Intel releases tiny 3G cell modem

Intel releases tiny 3G cell modem

Intel has released a 3G cellular modem with an integrated power amplifier that fits into a 300 mm2 footprint, claiming it…

More...
Braswell 14nm Atom slips to Q2 15

Braswell 14nm Atom slips to Q2 15

It's not all rosy in the house of Intel. It seems that upcoming Atom out-of-order cores might be giving this semiconductor…

More...
TSMC 16nm wafers coming in Q1 2015

TSMC 16nm wafers coming in Q1 2015

TSMC will start producing 16nm wafers in the first quarter of 2015. Sometime in the second quarter production should ramp up…

More...
Skylake-S LGA is 35W to 95W TDP part

Skylake-S LGA is 35W to 95W TDP part

Skylake-S is the ‘tock’ of the Haswell architecture and despite being delayed from the original plan, this desktop part is scheduled…

More...
Aerocool Dead Silence reviewed

Aerocool Dead Silence reviewed

Aerocool is well known for its gamer cases with aggressive styling. However, the Dead Silence chassis offers consumers a new choice,…

More...
Frontpage Slideshow | Copyright © 2006-2010 orks, a business unit of Nuevvo Webware Ltd.
Wednesday, 08 February 2012 11:22

Boffins work out how to make GPU and CPU work together

Written by Nick Farrell



Ebony and Ivory


Researchers from North Carolina State University have developed a new technique that allows graphics processing units (GPUs) and central processing units (CPUs) on a single chip to collaborate. They claim that the technique boosts processor performance by more than 20 percent.

Dr. Huiyang Zhou said that chip manufacturers are now creating processors that have a ‘fused architecture,’ meaning that they include CPUs and GPUs on a single chip. While this approach decreases manufacturing costs and makes computers more energy efficient, they still have not got the CPU and GPU cores working on the same functions. Chips rarely collaborate to execute any given program, so they aren’t as efficient as they could be.

GPUs were initially designed to execute graphics programs, and they are capable of executing many individual functions very quickly. CPUs have less computational power but are better able to perform more complex tasks. His method is to allow the GPU cores to execute computational functions, and have CPU cores pre-fetch the data the GPUs will need from off-chip main memory.

This is more efficient because it allows CPUs and GPUs to do what they are good at. CPUs and GPUs fetch data from off-chip main memory at approximately the same speed, but GPUs can execute the functions that use that data more quickly. So, if a CPU determines what data a GPU will need in advance, and fetches it from off-chip main memory, that allows the GPU to focus on executing the functions themselves – and the overall process takes less time, he said.

Nick Farrell

E-mail: This e-mail address is being protected from spambots. You need JavaScript enabled to view it
blog comments powered by Disqus

 

Facebook activity

Latest Commented Articles

Recent Comments