Still means the R290X has more horsepower.
I think after seeing the specs most of us expected it to do well at 4k, the problem with 4k is it doesn't matter.
Hopefully the card is well balanced and continues this trend downward to resolutions that actually matter.
That said I'm still very elated that AMD has seemingly finally caught Nvidia on a node, sad it isn't going to create the revenue GK110 did but it's still good to see. Hopefully AMD continues this mentality going forward.
Considering how little market share 1440/1600 have, and the price difference (1440 is gaining with cheap Korean panels) it's unlikely $3,500 panel performance numbers will matter in the typical life cycle of a high end product such as this.
A similar situation to CPUs performing very close when in GPU limited situations while in lower resolutions differences are bigger.
I wouldn't call that similar at all, more like a Haswell processor might destroy a Ivy Bridge processor in AVX with support for AVX2 and FMA3, but is only 10% faster in legacy code.
Apples to Oranges.
Design differences may get AMD a greater advantage at 4k, but the only reason that would shrink is because of the design itself that gave it the advantage in the first place.
Which has nothing to do with why the R290 might be faster at 4k but not quite as fast at 1600, there will be no cpu limitation to slow down R290 at 1600 for most tests because most testers are using 4.6-4.8GHz processors which are not bottlenecking the gpu in either case.
Your example is pointless and has nothing to do with what causes the i7 to be faster at low res than it is at high res.
So what would be limiting it then?
AMD wants to make R290X a 4k card, because of its design. Its advantage in ROPs and Bandwidth are seen clearly at 4k, most likely the waters will muddy at lower resolutions where the bottleneck becomes the TMUs/Shaders.
But the performance will still be more than acceptable at the lower resolutions. It may lose a couple % or trade blows at the worst.
Indeed, I don't see it losing to the stock 780.
Depends, and we need to find out at some point. There's those who buy one high-end card every 5 or so years in the hope the next one lasts an eternity, and there's others buying a handful high-end cards per year.Considering how little market share 1440/1600 have, and the price difference (1440 is gaining with cheap Korean panels) it's unlikely $3,500 panel performance numbers will matter in the typical life cycle of a high end product such as this.
)
Excuse me? I can't understand anything you wrote.all I have to say the last few posts re high res don't matter is base on what 5 year old console ports used in the benches .
come ps4-xbone for my money I think 4 k high res. is a good indication of a cards horse power that might be useful in the next gen of ports or maybe not depending on the port gods.
Excuse me? I can't understand anything you wrote.
Still means the R290X has more horsepower.
I am rubbing my hands with Glee at the chance to use SSAA.
As others have stated the Uber mode is likely for user tweaking. the clocks are likely the exact same as the ones in Quiet mode which is why there is negligable difference in the benchmarks. I would imagine that the few benchmarks that do show an improvement are due to the card not throttling in uber mode to maintain thermals.Possible Crossfire performance leaked out. If true, that scaling is disgusting in its efficiency.
Also difference between Uber/quiet modes looks negligible.
http://videocardz.com/46990/amd-radeon-r9-290x-crossfire-performance-leaks
As others have stated the Uber mode is likely for user tweaking. the clocks are likely the exact same as the ones in Quiet mode which is why there is negligable difference in the benchmarks. I would imagine that the few benchmarks that do show an improvement are due to the card not throttling in uber mode to maintain thermals.
Also difference between Uber/quiet modes looks negligible.
http://videocardz.com/46990/amd-radeon-r9-290x-crossfire-performance-leaks