ATI's next cards this autumn?

[GheeTsar]

According to Charlie, the roadmap for the 6000 series cards has been brought forward and we could be seeing a launch of the new cards as early as this autumn, rather than at the end of the year/start of 2011 as initially thought.

In short, as the road map to get to 28nm has slipped/looking uncertain now, the idea is to take the uncore elements of the Northern Islands GPU that are planned for the 28nm process and combine them with the Evergreen shaders and put it on the existing 40nm proces. This hybrid is to be dubbed the Southern Islands GPUs.

Anyway, you can read the article here:

http://www.semiaccurate.com/2010/03/...eration-outed/


Once upon a time I disregarded most of the stuff Charlie wrote, but after getting most of his predictions right about Fermi, I take his writing with slightly fewer pinches of salt.


Probably won't allow a mega jump in performance if it is still on 40nm, but it might be enough to take back the fastest single GPU crown.

[matty-hodgson]

So while NVIDIA will be getting their 4xx series up to speed, ATI are going to be moving onto the next set of cards?

This is just stupid.. They're gonna utterly destroy NVIDIA and be left with no competition .

[Funkstar]

This is just stupid.. They're gonna utterly destroy NVIDIA and be left with no competition .

Nvidia still have the highest performing low profile passive card out there, what's up with that??

ATI might out perform them chip for ship, but it's not a universal truth.

[raceway99]

Once upon a time I disregarded most of the stuff Charlie wrote, but after getting most of his predictions right about Fermi, I take his writing with slightly fewer pinches of salt.

i've heard lots of people (on other forums) say he's biased and cant be trusted, but pretty much all of hisarticles i've read have turned out true

so should I trust him?

[GheeTsar]

i've heard lots of people (on other forums) say he's biased and cant be trusted, but pretty much all of hisarticles i've read have turned out true

so should I trust him?

I'd suggest he has some very well placed sources and that his articles are not based on hot air, but then the sources can be wrong. He's also, very, very anti NVIDIA and somewhat pro ATI, so bear that in mind when you consider what he will write about. Ie: if he hears something dodgy about ATI he is less likely to report it.

[TooNice]

This is just stupid.. They're gonna utterly destroy NVIDIA and be left with no competition .

Remember the 9700 Pro (8/2002)? And the GeForce FX (02/2003)? Too late (deja-vu), too slow, nVidia got owned pretty badly that round. I dare say worse than the slap Fermi is getting. What happened afterwards? nVidia caught up. To be honest, the lead AMD had was more or less lost from the time the 6800 was released. Not the the X800 or X1800 series were worse than the 6800/7800 respectively. They simply didn't dominate like the 9700 did over the FX series. I doubt nVidia is getting buried yet, though they can't really slip up again. And if this round allows AMD to balance up the (discrete GFX card) market share from 70/30 to 60/40, it won't be such a bad thing (but I am hoping that nVidia will be able to get the refresh up closer to AMD's next thing).

[DanceswithUnix]

I'd suggest he has some very well placed sources and that his articles are not based on hot air, but then the sources can be wrong. He's also, very, very anti NVIDIA and somewhat pro ATI, so bear that in mind when you consider what he will write about. Ie: if he hears something dodgy about ATI he is less likely to report it.

Generally agree as he seems to have some *very* well placed sources. However this article is about ATI abandoning designs, so he does report when bad stuff happens to ATI.

The thing is, with his usual bias he still manages to end the ATI article with a dig at Nvidia