Featured Articles

Analysts expect ARM to do well next year

Analysts expect ARM to do well next year

British chip designer ARM could cash in on the mobile industry's rush to transition to 64-bit operating systems and hardware.

More...
Huawei and Xiaomi outpace Lenovo, LG in smartphone market

Huawei and Xiaomi outpace Lenovo, LG in smartphone market

Samsung has lost smartphone market share, ending the quarter on a low note and Xiaomi appears to be the big winner.

More...
Intel Broadwell 15W coming to CES

Intel Broadwell 15W coming to CES

It looks like Intel will be showing off its 14nm processors, codenames Broadwell, in a couple of weeks at CES 2015.

More...
Gainward GTX 980 Phantom reviewed

Gainward GTX 980 Phantom reviewed

Today we’ll be taking a closer look at the recently introduced Gainward GTX 980 4GB with the company’s trademark Phantom cooler.

More...
Zotac ZBOX Sphere OI520 barebones vs Sphere Plus review

Zotac ZBOX Sphere OI520 barebones vs Sphere Plus review

Zotac has been in the nettop and mini-PC space for more than four years now and it has managed to carve…

More...
Frontpage Slideshow | Copyright © 2006-2010 orks, a business unit of Nuevvo Webware Ltd.
Monday, 30 April 2007 09:58

Radeon HD 2000 supports Vec5 superscalar

Written by Fuad Abazovic

Image

320 SPs up to 40 percent faster than G80 (in theory)


Until today we believed that the R600 is a Vec4+1 scalar chip, which sounded very inefficient to us. We had a chance to take a peak at some DAAMIT documents and learned that ATI wants to present the R600 as a superscalar chip with 5 instructions per clock.


Previous ATI hardware, including the Radeon 9600, 9700, 9800 and X series could handle two instructions per clock with 3+1 or 4+1 components.

The Radeon HD 2000 series supports Superscalar marchitecture and it means that with Vec5 or Superscalar, ATI can process 5 scalar instructions per clock. It has 64 Unified Shaders and by multiplying these two numbers you end up with the amazing number of 320 Stream processors.

The Nvidia G80 GTX can only handle Vec4 instruction in a scalable way, or four independent instructions, and it looks like ATI might have an advantage here. Nvidia can process 128 Stream processor instructions, but at the much higher clock speed of 1.35 GHz, while the R600 can do 320 Stream instructions per clock, but at close to half the speed or 750 MHz.

In raw numbers, the G80 at GTX clocks can handle 172.800 Millions of instructions or Shaders if you like, while the R600 can handle 240.000 Million Shaders per clock. If this turns to be right, ATI could run Shader instructions - especially the unified ones - up to forty per cent faster.

Well, this will only be noticeable in next generation games and benchmarks, so you still have some time to wait for this revolution.   

Last modified on Monday, 30 April 2007 11:21
blog comments powered by Disqus

 

Facebook activity

Latest Commented Articles

Recent Comments