Originally posted by OTE
I remeber when Anandtech used to be, lets say more supportive of Nvidia. However his review on the 9700 changed all that, or so I thought. I just read his 9600Pro review, in which he uses a 9500Pro, 9500, 9600Pro, GFFX 5600, Ti4600 and Ti4200. Now correct me if im wrong but wasnt it that the 9500Pro beat the Ti4600 hands down in nearly every test, although not by the 9700 (Pro) margin. Yet in the UT2003 benchmarks the Ti4600 beats everything used to test, after seeing that I closed the window and came here to *****
not quite true
the ti4600 I believe has higher clock speeds and also is optimized... or rather some games are optimized... to run a bit better on nvidia cards... coding specific... by game devs...
the 9500pro performs around the same level as the ti4600 but is targeted @ the ti4200 which it beats handily... the 9700 vanilla (non-pro) is targeted @ the ti4600... which it beats handily..
the 9500pro competes @ around the same level with the ti4600.. but once eye candy is turned on... the 9500pro takes off and keeps highly playable frame-rates while the ti4600 falls behind...
also the 9500pro is a dx9 part.. the ti4600 is a dx8 part...
1 more point... various reviewers do things differently... hardocp at the moment is the only reviewer out there... along with perhaps b3d... doing the reviews the way most enthusiasts would like to see it... trying to do an apples to apples review...
this basically means that BOTH cards are run with pretty much the same settings...
nvidia has retuned their control panel to sort of match ati's... BUT there is a difference... while ati's quality setting == trilinear filtering... nvidia's quality setting (previously called BALANCED) does not do trilinear... hence the whole issue about image quality...
this is how nvidia wants the reviewers to test... stating their mostly bilinear filtered QUALITY setting is the equivalent of ati's trilinear QUALITY setting..
nvidia's application on the other hand DOES do trilinear filtering... and that is how all teh reviewers should really do this.. trilinear to trilinear... looking @ sean peltiers review on
www.hardocp.com you can see exactly what kind of hit the nvidia product takes when using trilinear..
there is a further problem... the gf4 ti core and the nv3x core do things differently AFAIK and the control panel settings work differently... again AFAIK when it comes to the 2 cards... (my description of trilinear v/s bilinear relates to ati dx9 cards v.s nv3x cards)
in light of my rambling above... I hope you give anandtech another decko... they are not so bad...
and I am sure you will see from their conclusion that they are not nvidia biased..
The release of the Radeon 9600 Pro marks a very important step in ATI's growth as a company, this is the first time that we've been able to recommend not only a high-end ATI GPU but also a mainstream GPU over NVIDIA's offerings. It's clear that ATI as a company has truly turned around from the days of the Radeon 8500 and we can only hope that the job-well-done will continue. ATI should not rejoice just yet however, the NVIDIA camp is definitely quite frustrated and will not take kindly to this sort of competition - vengeance is definitely brewing, only time will tell if it's enough to fuel reclamation of the the crown.
If we tally up ATI's wins (Radeon 9700 Pro, Radeon 9800 Pro, and now the Radeon 9600 Pro), the scoreboard is ATI - 3, NVIDIA - 0; but the game is far from over.