AMD: R9xx Speculation

From Gibbo at OCUK
Hi there

All I can say is they are bloody fast.

For the money I am amazed.

I can see this launch being hugely sucessul for ATI, easily on par with 58xx series but better because I literally have enough stock to prevent running out which will also mean our pricing will be excellent and I should be able to keep the price low low, no price hikes.
 
He's trying to sell them, he's hardly going to say 'they suck, don't buy them so we are lumbered with stock we will have to clearance at below cost'. :/
 
I would say that SLI is worth 10-15% over Crossfire which is why it is not a done deal for Cayman. Nvidia dual chip cards tend to be better than their ATi ones, but I don't really follow the sector as much so my knowledge may be outdated.
 
He's trying to sell them, he's hardly going to say 'they suck, don't buy them so we are lumbered with stock we will have to clearance at below cost'. :/

This makes no sense to me. If you want to artificially increase the demand, you scream that your stock numbers are low, to create anxiety. You don't say that you have tons of cards, leaving enough time to wait for the reviews to decide.
 
It isn't.

ATi's driver support is still as bad as ever, people say it is getting better, but honestly they are either lying to themselves (fanboys) or are being deceived. SLI is better, but I would always get a single GPU card. Seriously, every time.

It seems to me like crossfire 6850's have the beating of sli'd 460's. This appears to also be the consensus over most of the reviews.

Maybe you should tell us how you know "it isn't".
 
This makes no sense to me. If you want to artificially increase the demand, you scream that your stock numbers are low, to create anxiety. You don't say that you have tons of cards, leaving enough time to wait for the reviews to decide.

Good point. He does mention that there will be no gouging, which helps customer goodwill, unlike 5870 where we saw prices 20% above the recommended price. :(
 
I would say that SLI is worth 10-15% over Crossfire which is why it is not a done deal for Cayman. Nvidia dual chip cards tend to be better than their ATi ones, but I don't really follow the sector as much so my knowledge may be outdated.

No empirical data backs up that 10-15%.
 
Caman is 195 watts. 2x Cayman will also have to have some parts disabled and also lower clocked to fit into the power and thermal limits of a single PCIe card. So yes I believe nVidia will have an answer for the 6990.

Do you believe the 6970 will be slower than the 570? We already have information that suggests the 6970 uses considerably less power than the 570.

I don't see how you can math this out as working unless you believe the 570 is actually significantly faster than the 6970. AMD is going to win on any multichip solution. Nvidia aren't competing that well on a per watt basis.
 
Last edited by a moderator:
How long it's been since Fuad got something right in regard to GPU?
I removed Fudzilla from my bookmarks a while ago, lot of people should do the same.

More on topic the 15th can't some soon enough. The sooner the cayamns are launched the sooner we will more in regard to Kurt :)

Don't know why you have them bookmarked in the first place :LOL:
 
Caman is 195 watts. 2x Cayman will also have to have some parts disabled and also lower clocked to fit into the power and thermal limits of a single PCIe card. So yes I believe nVidia will have an answer for the 6990.

195W doesn't mean a dual card is 2x195W

In fact, I see a 5970 scenario for the 6990.. 2 x Cayman XT's @ Cayman Pro clocks
 
I'm starting to get the feeling that with the failure to advance processes as time goes on, AMD's small die strategy is going to start to backfire because they're aiming too low

Shouldn't this be the other way around? I mean NV chips are about as big as they can be. And they're already having issues with yields. But because AMDs chips are smaller than have more room to increase the die size. With their better perf per watt/mm, I can't fathom how fast AMD can make a chip with the size of the 580.
 
Shouldn't this be the other way around? I mean NV chips are about as big as they can be. And they're already having issues with yields. But because AMDs chips are smaller than have more room to increase the die size. With their better perf per watt/mm, I can't fathom how fast AMD can make a chip with the size of the 580.

You may get a chance to find out if 28nm is really not until 2012.
 
RV970 is still at a 25% die size advantage against GF100b

So does that mean it is 25% smaller than Gf110(err Gf100b sorry ;)) which if we assume GF100b to be 530mm2 is ~400mm2 for Cayman(Dont let Dave catch you say RV970 again)
OR
Does it mean GF100b is 25% larger than Cayman, in which case assuming the same 530mm2 for Gf100b, we get ~425mm2 for Cayman

As for the specs, right let me take a jab at it. How does this sound?

6970 = 1536 SP's, 880 mhz core, 5.5ghz mem

6950 = 1408 SP's, 800 mhz core, 5 ghz mem


Also with regard to the guy who went on and on about 28nm delays, can you please point to a source, otherwise stop talking nonsense. Edit: Meant credible source!
 
Shouldn't this be the other way around? I mean NV chips are about as big as they can be. And they're already having issues with yields. But because AMDs chips are smaller than have more room to increase the die size. With their better perf per watt/mm, I can't fathom how fast AMD can make a chip with the size of the 580.

Yes, i acknowledged that if a < 400mm^2 < 200W card can perform ~GTX 580 levels, then that's quite impressive, esp. from an engineering standpoint

However, the bigger issue is that if these process nodes are being pushed back, ATI increasingly has to resort to Cypress/Cayman sized cards to compete, in which case its less sweet spot and more borderline monolithic - and there's issues with that. That is, not quite good enough to take the crown, but also not quite cheap enough to be sweet spot. So it'll be interesting to see where Cayman performs at
 
I fully expect the GTX680 to come in August of next year on the 40nm process node and be a massively optimised with high clocks, better density, better thermals and all round better performer.
With all due respect, I don't think even Charlie considers nvidia to be that stupid.
 
Back
Top