Quote Originally Posted by NapalmV5 View Post
guys dont get your skirts up.. 64 texture units/half shader clock makes fermi mind boggling efficient and still beats 5870 (44gigatexel vs 68 gigatexel)

whitepaper doesnt state half/full or anything the shader clock.. only states "higher clock"

whitepaper @ v1.4 it wont remain @ v1.4 lol

only 64 texture units @ half shader clock cannot coexist.. one or the other will get whitepaper update.. but then again it still beats 5870s texturing power
For someone who kept saying in the other threads that he didn't know what he was talking about, you sure love claiming "to know!"

Kind of like how you said that the GT200 was stomping the RV770 because 512-bit > 256-bit... cept for that inconvenient fact about GDDR5 which is exactly what Fermi is doing huh!



Anyways, at B3D, they analyzed the numbers and it makes sense - games which use a lot of geometry / tri-clock (such as HAWX), Fermi was quite a bit faster than the GTX 285. Those that didn't though, it was close to 5870 performance... meaning there are games where the Fermi changes will improve it's performance greatly, and others where it won't be much faster if at all more than the 5870. Hence the average performance values put it above the 5870 but not faster than the 5970.

And if you want to talk efficiency... how about comparing the card to other cards in its power envelope???