ATi interviewed on shader 3.0 & benchmarks among other things

[Force Flow]

What is your opinion on the use of 3dmark03, other synthetic benchmarks, and the driver optimizations debacle? Obviously the point of a benchmark is to measure performance. The problem with benchmarks is that everyone obviously wants to get the absolute highest score possible, and sometimes this can lead to unfortunate situations. The whole point of a benchmark is to try and measure the real-world performance of a graphics card. So if you're doing anything specific for a benchmark then you're defeating the whole purpose of the benchmark. Personally, I would be more concerned with how a graphics card performs on a large selection and variety of games. Until Half-Life 2 comes out right? I'm quite certain ATI will perform very well in Half-Life 2, as we do in all other games available today. ATI's belief is that we must perform well in a broad spectrum of games, and not just focus on one or two ?key? benchmarks/games. It's been about 2 years since you've started the Catalyst drivers and they have really evolved since then. Where would you see the Cats in 2 years? I'd like to see the CATALYST driver as the #1 selling point for all ATI graphics card ;-) We're going to stay focused and continue delivering the most stable, high performance driver in the industry. We're also going to put significant effort into enhancing the usability and features available to the end-user in the coming months. There is a lot of talk over the lack of Pixel Shader 3.0 support in ATI products. What do you think of this argument? We actually do support some 3.0 shader stuff ? our hardware is capable of supporting Instancing for example, but more importantly we won't be seeing games that even make heavy use of 2.0 shaders for another year or two, and by that point in time the graphics hardware architecture available will be a few generations ahead of what's available today.

~From vr-zone.com

http://www.vr-zone.com/?i=1016&s=1

[pam123]

Thanks F_F, a good read.
I think ATI guessed wrong about the 3.0 shaders ( and they've been hearing the howls about Open GL for awhile now ) but so long as the don't pull any NVidia moves ; lie, optimize for benchmarks, that sort of thing; they should be back on top by spring.
I'm thinking about a new graphics card by year's end, in the $200 range, and they're very much in the running ( if NVidia comes out with a good mid-price card though...).