Come on, try to be objective to some extent and not illogical. Rounding 27.7 up to 30% is pretty significant and not taking into account that Metro 2033 score is bolstering things up(117) big time with only 7 gaming benchmarks(with no min and max) and doesn't represent a realworld practical setting because both cards are rendering the game unplayable at that setting. It is likely both cards are hitting a memory bottleneck and the 5870 is taking it harder. That average is decreased 10-11% if we take it out of the equation.
Any card can look ridiculous powerful once you start playing both games at unplayable settings, as one will barely run because of a memory bottleneck or another bottleneck.
I can still see a 6970 being faster than 18%(and I expect it) better than a 5870, but not the way your accounting(really really spinning it) the evidence. I think reviews will be substantially better than these results, but atleast with this evidence, I wouldn't say cayman xt is 30 percent faster yet.
Look at the chart below.
With DOF on like the person here. Look at the frame rates when DOF is on in the 1gb 5870 it cut down really really badly.





)
Reply With Quote
Bookmarks