hmmm... interesting... looks like the empire (oops) ATi Strikes back...
seems to me to be a good reply by ATi in light of the d3 benchmarks...

eagerly waiting for the X-Bit Labs article... that would throw real light on the performance...

Sorry not at all believing the performance differential they are showing between the 6800 ultra and the X800XT, just as I don't buy it when you reverse those numbers in some other benchmark.

But oh well.

Either card performs superbly in a system - people ought to be out having fun with whichever they chose...

I have a 9600XT! And the HL2 coupon.

Where can i get this game free?

but look at the settings, only 16x12 4AA/8AF was benched. Here I really expected a more clearly Ati-Lead, due to the fact that the x800 leads in almost all games at these settings. Probably the 6800's winning all 10x7 w/o AA/AF benches.

That's pretty misleading... there's no mention of the resolution or level of AA and AF being applied in the graphs. I run the game at 1024x768 with 4XAA 8XAF with a 6800GT with vsync on and I never notice ANY slowdowns even in intense battles and driving the buggy where everything's moving very fast.

Oh wait... there it is in EXTREMELY small print...

I bet we will see FP16 on this game for nvidia cards in a future driver update.

Nvidia will say you will get premium performance at a cost of a slightly image quality decrease.

Doom3 runs what FP12, FP16?

> Doom3 runs what FP12, FP16?

Doom's shaders are programmed with the fastest flag which will mean they should use FP16 (although it is up to the driver and early drivers used FP32).

These benchmarks show what xbit showed earlier with the preview, ATI runs HL2 better at high resolution/AA settings. In xbit's engine stress test preview at medium settings it was a tie and low settings the nv4x won. I assume this is still the case as they only showed the high resolution settings xbit already showed ATI winning in.

So we've just got marketting here, not news.

The Inq's:

http://www.theinquirer.net/?article=19707

ぬるぽ

HI,

hmm... wo bekomm ich den benchmark da denn her ??


mfg

I don't think these results are true. Percentage instead of fps count looks suspicious to me... 140% does not mean anything actually, does it? I'm not going to buy any of the cards benchmarked;D, I'm rather interested if the 'old' GF FX 5600 will run the HL2 well;)... does anyone know how much of RAM is needed to play it comfortable? Will 320mb be enough (I suppose not, but I still hope) :)