Quote Originally Posted by cegras View Post
Yeah, but if ALL cards are run with the SAME benchmark, it still gives a RELATIVE scale of performance. I don't see what water your argument or [H]'s argument holds.



Considering games are built on the nvidia architecture? Go figure.
It's not my argument, I merely see their point and would like to see it all further tested.

It does hold water though for the following reason...

You see, using cut-scenes and fly-bys don't use half of the effects actually used in gameplay. Why does this matter? Because a card can ROCK at some effects, and suck at others, and without testing the actual gameplay you wouldn't know which is better overall.