Amd Volcanic Islands details

[GoldenTiger]

@1080p, its really a shame to benchmark those cards at such a low resolution. However, in that review it trades blows with the Titan, sometimes being quite a bit faster, other times being slightly slower. This is likely just the Nvidia/AMD difference in specific games that we see with all cards. But either way, if the price is $650 and is slightly faster or equal to a Titan (not counting power/heat) why would you be disappointed?

Because the 780 has been out at the same price as you're quoting, for several months at this point, and performs around the same. In short it's basically nothing new for the market, other than hopefully causing a small price war. Also, the reviews with 20million-X antialiasing that no one would actually ever want to game at since you literally couldn't see the difference even in screenshots without 400% magnification let alone in motion, are definitely a shame to benchmark in. Give us some actual benchmarks of how people play games, i.e. max settings for example with 2x MSAA + FXAA resulting in a framerate of 80, rather than 8x MSAA with no FXAA leaving tons of motion shimmering and no transparency/alpha antialiasing, while also leaving shader aliasing untouched and crippling your framerate to unplayable 30's.

Absolutely not...that benchmark? is insane to the point I won't even use it for stability testing, Its just a vicious loop that will burn up cards with inadequate cooling.

While I agree Furmark is not a great test due to its unrealistically high load, the cards are on even footing there so you can definitely draw some info out of that testing.

[Stukov]

Well, from some of the document slides, it appeared the hardware runs higher resolution without as much penalty, the whole eye-infinity thing. From the benchmarks I saw, it was faster than the 780. However, I don't know if I trust that review site, we will see in a few days though either way.

[vario]

Few days ?
So when is NDA ending ?
First ists supposed to be 15th, but now people were sayin about BF 4 premiere, and thats in two weeks.

[PedantOne]

Few days ?
So when is NDA ending ?
First ists supposed to be 15th, but now people were sayin about BF 4 premiere, and thats in two weeks.

i told it many times here.

10/24 - 290X and 10/31 290 reviews

[vario]

From the PCOnline Review it looks like R9 290X is 0%-10% slower than Titan with the exception being Metro Last Light.

Thats odd, maybe we read different articles.
PcOnline on 1600P performance
AMD Radeon R9 290X 1080p performance summary

R9 290X vs R9 280X: +37%
R9 290X vs NVIDIA GTX780: +11%
R9 290X vs NVIDIA GTX TITAN: 1%

And as for 1080P performance, it was considerably faster in metro and sleeping dogs ,so putting it into "0%" faster to 10% slower is kinda.not true.At least if you base opinion on that slides.
If you want factually correct statement it is between 18% faster to 10% slower.

i told it many times here.

10/24 - 290X and 10/31 290 reviews

Thanks, makes sense from everything i read.
Will keep score if your right again.

[LightSpeed]

While I agree Furmark is not a great test due to its unrealistically high load, the cards are on even footing there so you can definitely draw some info out of that testing.

No, they are not on level footing. Nvidia detects furmark and artificially caps the power so they look so much better than the amd counterparts. Run games and they are on level footing where that 70w of power difference will come down to a few watts. Just see the 280x review vs the gtx780.

Its amusing that so many people are mislead by furmark numbers thinking that they represent relative real world power consumption between nv/amd. Well, they just dont.

[Particle]

While I agree Furmark is not a great test due to its unrealistically high load, the cards are on even footing there so you can definitely draw some info out of that testing.

I don't really think it could be called equal footing given that both AMD and nVidia write their drivers to recognize and throttle applications like FurMark which they refer to as a "power virus". All you're seeing is how much power they've decided to let the card consume under those situations which is some arbitrary point not related to the maximum power each die would necessarily want at a true 100% load.