any improvements over the 10.3 w/ W7 & 5770?
Except that according to phoronix there are multiple performance regressions in the 256 beta drivers. Now, I know that's linux, opengl and beta drivers -- but the important bits use the same base code for windows and linux. I'm sure the nv blog praises them to high heaven but so far end user reports of the 256 drivers have not been awe inspiring.
I knew those improvements were too specific. Down to resolution and AA settings? That was just too weird
Nope, I knew they shot there load with the 10.3's
Now we have to watch Nvidia move farther ahead in performance with there new 256 drivers which are great even in beta form.![]()
http://blogs.nvidia.com/.a/6a00d834515fca69e20134818119df970c-pi
Remember the previous drivers gave the gtx series a 5% boost over at Xbit labs review.
Thats 10% and there official gtx 4xx driver comes out with the release of the gtx 465.
The gtx 480 should be a good 25% faster then a 5870 by then, and the gtx 470 about 15% faster then the 5850. With of coarse the better minimums across the board.
What a load of rubbish.
To get back on target. I'm using the 10.4's with no issues, but have read a few quirks among other message boards with users running the 10.5's.
Anyone upgraded from 10.3's or .4's with positive results? I think i'm going to stick with the 10.4's for now.
To get back on target. I'm using the 10.4's with no issues, but have read a few quirks among other message boards with users running the 10.5's.
Anyone upgraded from 10.3's or .4's with positive results? I think i'm going to stick with the 10.4's for now.
Now, I know that's linux, opengl and beta drivers -- but the important bits use the same base code for windows and linux.
i love how now every thread we open its about that one guy blathering about "ATI shooting its load in 10.3" damn near verbatim. the moderation on this site has become a lop sided joke
I feel a thread derail coming up. anyway, gonna get these drivers for the BC2 fix
I never had a problem with Bc2?
The long loading times in DX10/11
SNIP
