nop... nvidia made a just made a BAD turbo boost...
unlike intel and amd turbos, nvidia pushed the turbo to the card's limit
the problem is that diferent gpus have diferent ASICs... and clocks can't be reached in a pattern
so...a every kepler is will have a diferent max performance than others
Value and Conclusion
After testing two GeForce GTX 680 cards in SLI one day after the official launch, with the first public driver I'm quite impressed. We see good SLI scaling in all our games, except for Skyrim which seems to use a quirky rendering engine that doesn't play well with multi-GPU setups. Of course you'll only see decent scaling at high resolutions, 2560x1600 in this case. Everything below that feels like an insult for this two GPU monster setup. However, I can imagine users of 3D Vision with 120 Hz will see some nice improvements at lower resolutions like 1920x1200.
What kind of biased review is this? Along with many of the comments.
You claim the 680 beats the AMD cards in every category which is complete nonsense. Even from your own benchmarks, only half of them did it beat the non overclocked 7970. And the other half the benchmarks were the same or LOWER.
The Nvidia card also did not beat the 7950 in heat or power consumption. Yet Nvidia wins? What a joke.
I'd also be interested in knowing which of those games are using the proprietary graphic drivers for Nvidia. Because that in itself makes it a false benchmark.
^ This guy keeps posting the same thing in every thread, hoping that someone will buy it.
hey janoo i disagree with keys 9 times outa 8, but hes right about temps and power. not saying you are wrong, but i wouldn't immediately discredit his statement.
1.2v gpu at 30C liquid is going to use less power than the same 1.2v gpu with the block off on an air cooler running at 90C.
This is what I believe to be true, every single card has a different GPU boost limit depending on ASIC quality. Thats the only explanation as to why my GPU2 speed is consistently lower than my GPU1 speed.
Therefore 1 persons GTX 680 could perform better than another GTX 680 depending on luck of the draw (non overclocked). I'm not 100% sure of this but I can't think of any other reason why I have GPU that consistently has a lower GPU boost than another.
This is what I believe to be true, every single card has a different GPU boost limit depending on ASIC quality. Thats the only explanation as to why my GPU2 speed is consistently lower than my GPU1 speed.
Therefore 1 persons GTX 680 could perform better than another GTX 680 depending on luck of the draw (non overclocked). I'm not 100% sure of this but I can't think of any other reason why I have GPU that consistently has a lower GPU boost than another.
That is highly unlikely. You'll probably just be getting good availability on the 680 by May.
My guess is with gtx680, more often than not memory speeds are going to have to overclock as well as core speeds (percentage-wise) to get the best scaling. In other words, I think there will be more instances with the gtx680 when overclocking only the core won't produce scaling as good as when overclocking only the core on hd7970.
So, in yet again other other words, gtx680's overclock scaling will rely more on memory overclocking than hd7970's does.
GTX680 SLI review...
And they don't compare it to HD7970 Crossfire.
Pretty useless really unless you knew 100% you were going to buy NV no matter what.
A review that doesn't give a comparison to the competition...
i have a question
does the SLI have micro sutter?
True which is why cooling/temps on the memory could matter more than usual with the GTX 680.
Neither 7970 xfire or 680 sli have microstutter, I play with vsync on. I fiddled with adaptive vsync and didn't notice much difference, I used Adaptive vsync in crysis 2 and Dragon Age 2 (with vsync off for both ingame) and it simply turned the ingame vsync on for me. LOL. Not sure what that was about, didn't seem impressive to me. Anyway
GTX680 SLI review...
And they don't compare it to HD7970 Crossfire.
Pretty useless really unless you knew 100% you were going to buy NV no matter what.
A review that doesn't give a comparison to the competition...
A multi-GPU HD 7950 CrossFire setup is certainly not a cheap thing to get. With $900 for the two graphics cards alone you could get a large TV screen, a console and a couple of games for it. We see good performance scaling in most titles of our test suite, except for Batman Arkham City and Elder Scrolls Skyrim. Battlefield 3 at 2560x1600 also fell back to single GPU performance, even though lower resolutions scale just fine. We noticed the same problems in our HD 7970 CrossFire review with an earlier driver, so it looks like things are not going as fast as they should with AMD's driver development.
The thing is the cards are so close it doesn't even matter. Hell, the extra vRAM on the 7970 is enough to make up the wattage difference. I'm pretty sure an overclocked GTX 680 is going to have a tough time competing against a 7970 @ 1300MHz+. Does that mean the 7970 is the best card? No, just under those conditions. However, you'll see fanboys run around declaring "wins" and "my side has xyz" and all this other nonsense rather than compare the cards completely for what they offer. The best part is most of the people that cause the biggest stink haven't even bought either, so go figure.http://www.computerbase.de/artikel/g...stung_mit_aaaf
10% faster at 1080p and 3% faster at 2560 without AA and 9% and 1% faster with AA might not be meaningful victory but technically it is faster, less power hungry and has lower msrp than reference 7970.
The thing is the cards are so close it doesn't even matter. Hell, the extra vRAM on the 7970 is enough to make up the wattage difference. :
You can't do +50%. You can do 132% of TDP with EVGA precision X, which AFAIK is the only app that works with the 680...
So far i'm using 132% power limit, with 100mhz offset and 100 mhz memory offset. If I go higher than that I get tons of TDPs.
Afterburner doesn't work (even the latest beta)
I agree, to see how similar their designs are, it's really amazing. I'm sure there's some, but not a whole lot, of cross-talk, company secrets and all that, but the evolution over that time frame really does show "survival of the fittest" as far as designs go.Frankly any performance difference under 10% is not noticeable and thus doesn't really matter. For me 7970 and GTX680 are about as evenly matched as they could be with divergent architectures. It's rally amazing that they started those designs 4 years ago without knowing competitor's design and came out with the same performance. If I were buying today I would buy the cheaper card.
Also, I bolded how most consumers are going to feel. In general I think I'm going to be putting in or recommending the GTX 680 to friends/others over the next year (or whenever/if Big K comes out) simply because it's easier and cheaper. If the 7970 eventually gets cheaper than the GTX 680, I'll throw the 7970 in instead.
If all your friends are from the US then it's true but in europe 7970 is cheaper. In my country you can buy 7970 for 10% less than GTX680. Obnoxious price gauging on the GTX680.
You guys have it worse then we do, at the very least GTX580 is 20% cheaper than 7970 where I live.Pretty much everywhere except North America, the 7970 is going to be cheaper. Here in the Middle East, the 580 retails for the same damn price as the 7970, I haven't seen any 680 yet, but it will be higher by at least 60$ than the 7970.
You guys have it worse then we do, at the very least GTX580 is 20% cheaper than 7970 where I live.
Does anyone have an explanation why nvidia cards are so much more expensive in other parts of the world compared to the USA but not AMD cards? Is it only because of greed since nvidia has a stronger brand or is something else at play?