AMD: R9xx Speculation

You obviously didn't read the thread.

Generally speaking EU prices include 20% VAT and then the usual translation for consumer electronics is not base on the dollar-to-euro exchange rate but basically 1:1.

So if $500 is VAT inclusive that's 500=price*1.20 so price=500/1.2=$416


no, doesn't work that way most of the time. At least in my experience.

US$500 (without VAT) = €500 (with VAT)
 
I don't understand you... In fact they are aiming higher and higher.

HD3870: ~65% performance of 9800GTX
HD4870: ~75% performnace of GTX280
HD5870: ~85% performance of GTX480
HD6970: ......... performance of GTX580 (please fill in by your own)

I know what you're saying, and AMD still has room to expand within 40nm of course, but I'm just saying they need to start pushing against the limits if processes really do take longer to get to. If 28nm really is delayed to early 2012, this is the 2nd node in a row with major issues, in which case they'll need to look at yet another 40nm product possibly next year and with changes already done here, they might have to scrap small dies for another gen

And I would actually say that the 5870 is more of 80% of the 480 so we'll see where 6970 ends up, esp. w/r/t die size and power


U forget the dual card. If the new cards will show same scaling as the 68xx cards in crossfire than nvidia is second again.

Also the money is not in the high end cards. And game developers usualy dont care a shit about nvidia 580gtx or radeon 5970 cards.

Yes I'm well aware of that, but it's not stopping Nvidia from releasing the 570GTX and eventually the 560's which the 6800's will probably have a tough time with

And who's to say Nvidia won't have a 2x card this time either? Dual 570's might well fit within the realm of doable
 
NathansFortune: Check my post in a week :)

ZerazaX: ATI/AMD has exactly the same room as nVidia. In fact they are in better position. They are under 400mm² while nVidia is over 500mm². ATI can increase die-size to achieve better performance. nVidia can't, they are already near the upper limit.
 
It depends on the amount of redundancy, but I will go out on a limb and say a 2000 ALU GPU is significantly harder to get decent yields on than an 8 core CPU. Which is why Fusion designs have been delayed on the 32nm High-K process and the vanilla CPUs haven't.


Well CPU's tend to have a much longer validation time then GPU's, the rules are alot more stringent, so with a GPU/CPU hybrid I wouldn't be suprised the delay is caused because of validation times are going to even greater!
 
NathansFortune: Check my post in a week :)

ZerazaX: ATI/AMD has exactly the same room as nVidia. In fact they are in better position. They are under 400mm² while nVidia is over 500mm². ATI can increase die-size to achieve better performance. nVidia can't, they are already near the upper limit.


That is true, AMD is in a much better position with current designs, if process tech doesn't evolve in a timely fashion.
 
NathansFortune: Check my post in a week :)

ZerazaX: ATI/AMD has exactly the same room as nVidia. In fact they are in better position. They are under 400mm² while nVidia is over 500mm². ATI can increase die-size to achieve better performance. nVidia can't, they are already near the upper limit.

You're right, but again, the question is... will going bigger force ATI to encounter the same issues Nvidia has had? And will they be prepared/able to take the financial aspect of it? After all, Nvidia has other business lines they can sell those cards in to make up for it

And I don't think Nvidia is tapped out just because of it - the 580, while some don't consider 20% as impresive, certainly shows Nvidia is still quite able to revamping cards to perform better and take less power draw - who knows whats still up their sleeve

That being said, ATI is in a better position with smaller die - no doubt about it, but part of ATI/AMD's advantage in the past few years was being one step ahead on process, so come 2012, all bets are off
 
You're right, but again, the question is... will going bigger force ATI to encounter the same issues Nvidia has had? And will they be prepared/able to take the financial aspect of it? After all, Nvidia has other business lines they can sell those cards in to make up for it

And I don't think Nvidia is tapped out just because of it - the 580, while some don't consider 20% as impresive, certainly shows Nvidia is still quite able to revamping cards to perform better and take less power draw - who knows whats still up their sleeve

That being said, ATI is in a better position with smaller die - no doubt about it, but part of ATI/AMD's advantage in the past few years was being one step ahead on process, so come 2012, all bets are off

Good point, there are a lot of advantages to going smaller and ATi will have to deal with the same issues Nvidia had to deal with for GF100 and have now solved with GF110. Maybe that is why there is sudden talk of lower performance, ATi are just encountering the same issues as Nvidia did with Fermi, and the lack of a process shrink has killed their yields and thermals.

The third gen 40nm products are going to be very interesting, I would be surprised if Nvidia bothered with Kepler and ATi will have optimised any issues with larger die sizes by then so hopefully both will be super competitive. :D
 
Die size, performance is anywhere between 20% above Cypress to 20% above GTX580. The most realistic rumours seem to be in the 30% above Cypress which would put it 10% below GTX580, allowing a lower price but a better price/performance ratio.
 
Sounds like CaymanXT performance sentiment relative to the 580 is shifting on an almost daily basis. ATI/AMD have done a remarkable job keeping everyone in the dark this close to launch, but after the inevitable buildup of expectations (even amongst the conservative crowd here at B3D), they may be setting themselves up for a major letdown.
 
Sounds like CaymanXT performance sentiment relative to the 580 is shifting on an almost daily basis. ATI/AMD have done a remarkable job keeping everyone in the dark this close to launch, but after the inevitable buildup of expectations (even amongst the conservative crowd here at B3D), they may be setting themselves up for a major letdown.

My performance prediction for next week is 5970->580->6970->570->6950.
March will be 6990->580->6970->etc as the 6990 replaces the 5970.
 
Back
Top