GT300 "Fermi" Game Performance Figures Leaked

[bboynatural]

Well you can get the best GPU for a performance of 10 fps over ati.
If you have the money, you can pay 1000$ for that SUPER gpu.
As the father of taoism said:

One that pays 200$ more to get "only the best" obviously has no brain.

-Lao tzu

 

[soundefx]

Damn.. 'Leaked' on youtube.
It is either the NDA is really tight on ALL the review sites
OR
This is BS..

I wonder....??????

 

[bboynatural]

Obvious BS man.
NO nvidia card at this point had 30fps more then ati's flagship.
Why would it change??
That would be crushing ati to the bones.

 

[yannifb]

Well the most obvious way to tell this is fake is by looking at the Crysis fps for the 5870 in that vid.... As i recall it got more than 40 fps...

And btw to that John__ who seems to be a super nvidia fanboy - im not biased, my rig is like the UN of computing parts.

 

[TheViper]

Gotta love how the

"Nvidia Test Lab
Confidential - internal use only"

pops up after 20 seconds in. Had it been real, that watermark over the graph would have always been there.

Also funny is that "Confidential - internal use only" data sheets don't come in the form of nicely made up bar graphs with background GPU images in them. They tend to just be data. The graphs get made for PR.

It's also comparing a broad named product series (GT 300) as opposed to a specific GPU (GTX 380) against a specific GPU (HD 5870).

And what's with the crazy looking SLI bridge (especially on a Tesla card)?

 

[IzzyCraft]

Obvious BS man.
NO nvidia card at this point had 30fps more then ati's flagship.
Why would it change??
That would be crushing ati to the bones.

8800GT vs 2900HD anyone? Those are about the same time and if i remember 8800GT crushed the living hell out of the price/performance index and the 2k series all together ofc you could always look at 8800GTX which came before it and was still a better performer and the performance gap was quite big vs the horrible R600 chips.

Anyways at 1680x1050 30fps isn't that impressive of a gap a bais game would do that.

Benches are probably pure ***. haha still leaked on youtube and the best *** they can come up with is graphs? Not a video of a test bed doing the benchmarks. just screams i wasn't tring.

 

[ominous prime]

95% chance its BS, but thing is, thats around the performance I was predicting...

And BTW, 1680x1050 is a better testing metric then 1980x1200, where the CPU comes much more into play. Really, to trully test a GPU, 800x600 should be used to ensure the CPU isn't a factor.

:lol: Should have put your sarcasm hat on gamerk, I think some people thought you were sneerious.

 

[Homeboy2]

Alright, I've never hit a kid before. I mean that's like asking who Gandhi is.

Gandhi????

 

[Jaysin]

This is like a 9/11 truther video, completely lacking in the reality department.

 

[randomizer]

It's not from a random chinese website so it can't really be leaked. Not even Fuddo thinks it's worth posting about. It's amazing that people can post up crap on youtube and it spreads around all the major forums when there's clearly nothing credible about it.