Printable View
Looking at the pic, all the values seem to be bolded, with an exception that the Min FPS and Max FPS values of the ATI run do not seem so bold. But for example the ATI Score's 9 is identical to the 9 seen in Nvidia's FPS. So are the 2's compared to Nvidia Max FPS.
Why the Min FPS and Max FPS seem less bold? They're clearly bolded for Nvidia...
btw, it strange how point of view chages logical conclusion on subject. Take these two:
"it must be fake as numbers are not of same size in score and fps"
"it must be real as numbers are not of same size in score and fps"
First one suggest that it was poor copypaste, second one suggest it was not copypaste for same reasons..
Altough it would be logical that score is bolded more than intermediate results, I cant say for sure if it is the case, so could someone post real unigine run pick?
EDIT: yes it seems that in nVidia screen all the numbers are of the same bold, so likely fake.
Here's my special faked pic. You can take a look at all the 2's all you want, all day.
http://img408.imageshack.us/img408/7316/unigine2.jpg
Conclusion: Those scores are faked, unless they purposely edited the html to use a different font.
So for the ATI runs Min and Max FPS numbers have been altered with 99% certainty. Makes no sense. :confused:
True, the scores and the FPS seems to be using the same font as an unedited ones.
http://img408.imageshack.us/img408/7316/unigine2.jpghttp://i185.photobucket.com/albums/x...860315_000.jpg
Side by side... well not really accurate, but it's good enough. No visible browser zooming there.
Okay.. interestingly, everything on my screenshot is bolded. EDIT: He's using IE. Still, the scores are still shopped (difference in bolding).
Or he deliberately wanted us to believe it's fake even if it wasn't. xD It would be quite a noob if he really wanted to make a convincing effort by not copying the real font.
One doesn't even have to copy the fonts, or even open any image manipulation software to do this. Notepad should do the trick?
Yawn. So the 3DMark scores could be fake aswell? Since without a doubt, these are.
It's pretty annoying. The fonts are perfectly aligned too. Either that's some dedication in photoshopping (checking pixel by pixel), or he simply edits the html file, change the font, and do what you just said. I'm betting on fakes. We know how secretive AMD can be.
Any way to check if at least his FPS & score matches. Like does any1 know how to calculate the "score" from the fps result.
yeah, true @unigine screenshot. it was edited obviously. the fonts look the same for avg fps, score and max/min fps on a legit screenshot. however, who comes up with the idea to fake this screen in photoshop/whatever instead in the html file itself - no one would have noticed if it was done in html, lol.
Hehe yeah. I could totally fake one screenshot RaZz! :P:
Radeon HD 6999 :D
Well, the FPS and Score numbers seem unaltered still.. Maybe it was ran on a HD 5970 which gives lower min and max FPS, so someone had to uppp them a bit?
First of all - Scores on the screenshot of "6870" for the shown "Avg fps = 36.6" - is correct. Secondly - it sure as hell was not edited in photoshop. It really looks like someonne has messed with Min and Max Fps in .html file but forgot to set text to "bold" as it is with "Scores" and "Avg FPS".
Anyway - this is just my thoughts.
More like, someone REMOVED the bold tags from the file. They are there by default. Again, makes absolutely no sense. For me it seems as if someone ignorant just messed with photoshop without knowing that there's a HTML file.Quote:
Originally Posted by Daimler
Why can't there be a checksum to validate the results... :shakes:
Results are valid. AFAIK Scores=AvgFPS*25,18. It doesn't take min and max fps into account. Maybe someone knows better. :confused:Quote:
Originally Posted by Calmatory
And it would take hell of a time to fake this in photoshop with all those alignments... and not to notice the difference in boldness of the text?? no, it's just impossible. He edited .html and removed [ b ] [ /b ]" - 100%. by mistake. IMHO :)
ATI run: 922 / 36,6 = 25,1912568
Nvidia run: 743 / 29.5 = 25.1864407
So they seem valid, theres small rounding error(i'd guess at least) though.
However.. I believe that run wasn't ran with HD6xxx, but with something like HD5970, and the MinFPS and MaxFPS values were modified to make it look like better performer...
This.Quote:
Originally Posted by Daimler
It might be fakes, it might not, but i tend to hope that they're real & be optimist, because that would keep the technology bandwagon moving forward & pushing the prices down one way or another. 229 US$ GTX 460 1 GB seems nice, but it would be nicer if there's a strong competitor from ATi (Bart based cards) that adds choices and perhaps knocks competitor price down a notch or two. :)
so.. that 68xx card is faster than two 5850's ?
fake or they did some magic for high tessellation..
so many assumptions, i give up worrying. when are we do for some real info strait from amd?
They are fakes or at the very least modified based on the evidence in this thread and other threads.Quote:
Originally Posted by spursindonesia
Keep prices down? AMD is selling whatever quantities it has of 5xxx. When the 6xxx gets released. There not going to be a lasting price drop on 5xxx products because they are going to be to or close to out of stock already.
If these scores are true, AMD will jack up the price of their card another 100 dollar up the line at the very least to reflect their performance increase. They are already selling out at their current prices. And there's nothings wrong with this, but AMD is not the saint of a company people make it out to be, nor is nvidia the devil.
But its going to be pretty scary for the consumer. 6870, 499 MSRP 600+street price. 6970 750MSRP 850+ dollar street price.
This is almost a certainty to happen in regards to pricing because these card are so underproduced and cards supply are so constrained that the consumer will pay way more than MSRP and along with the increased pricing, its going to be the most expensive generation of cards ever.
If this generation from AMD has taught us anything, supply can screw with the consumer just as much as any company.
The scary thing is if Nvidia doesn't have anything to respond with, which they won't unless they have something underwraps which is unlikely, the consumer is going to be overpaying till Nvidia get a new generation going. The worst part is a solution might not be in sight till perhaps even longer than 28nm considering how unscalable fermi seems at this point considering the size and performance for a new generation at this point. Considering Nvidia may be selling gtx 470-480 at cost, this is really bad for the consumer when a company has to sell a product below cost. It does horrible damage to a company as seen with AMD after the Core 2 duo generation.
http://i35.tinypic.com/1zg5ij9.jpg
Comes bundled with a copy of Duke Nukem Forever! :p:
fermi is fine. almost every chip in the last 10 years has been designed for process scaling. i wouldnt be surprised if they could get a 1024sp fermi on 28nm. they may do an Si spin and get a considerable improvement on 40nm too. i think GTC will reveal what they are planning.Quote:
Originally Posted by tajoh111
AMD's losses were not only from inferior products but buying ATi and the TLB bug fiasco. nvidia doesnt have those issues.
amd leaked those screens, the 6870 is even more powerful than that.
everyone knows that.
Let me chime in with some rumours ive heard from somewhere. Believe me if you want to, else dont :)
Cayman will be the 6810/6830/6850/6870 (yes, four based on Cayman) to be released in November. The 6870 will be around 10 - 15 % faster than the GTX480, have clocks of 900/1500 and have 2GB memory.
Barts will be very similar to Cypress but on a new PCB and different clocks, maybe some SP's disabled and have 1GB memory.
Turks will be very similar to a Juniper but again on a new PCB and with different clocks (possibly higher).
Hmm.