UNREAL TOURNAMENT 2004
Although a DirectX 8.1 title, UT2K4 is a mainstay so I recorded my own timedemo
in the BR-IceFields map with 15 bots for a two-minute duration using the
console's timedemo feature. I also used 3DCenter.org's
highly stressful Primeval timedemo.
3DCenter.org - Primeval Timedemo
As expected, Unreal Tournament 2004 shows its CPU bottleneck as overclocking the
BFG 6800 Ultra OC yields little result. The X800XT PE handles Primeval a bit better, especially
at 4xAA/16xAF again attributed to the X800XT PE's higher fillrate.
BR-IceFields - My Custom Timedemo
Here, the BFG 6800 Ultra OC does quite a bit better in non AA/AF modes and 1024x768 with
those on. The X800XT PE pulls slightly ahead once again with the higher fillrate demands.
The X2 Rolling Demo
makes some use of DirectX 9.0 and so we'll see how these cards performed at
various settings within the demo.
X2 The Rolling Demo
The BFG 6800 Ultra OC beat the X800XT PE in every
X2 test. Although, the margin of victory is still relatively
small (just like the X800XT PE's margins were in some previous results with
UT2K4 and Far Cry). The two cards are essentially the same regarding
performance. It's how they'll fare with soon-to-be released games and being on
even yet faster systems that will tell us more about their potential.
Emil Persson
(a.k.a. Humus) has been a well known independent contributor to the 3D gaming community
for some time now. Over the past few years he has developed a wide array of interesting and informative 3D
demos. He was hired by ATI Technologies, Inc. back on May 4, 2004. Humus has released three new demos, Dynamic Branching
being the one which I benchmarked below with FRAPS.
Here are some
comments from Humus regarding this demo and how it runs on ATI and NVIDIA
hardware:
Notice that the default settings are different
between the ATI and NV cards. I simply choose the path that's
fastest on each card. For ATI cards it's faster to just zero stencil
when the stencil test passes, while on NV cards this seems to
interfere with early stencil rejection, thus they are faster when I
clear the stencil buffer between lights instead. That's still quite
fast on ATI, but not as fast as zeroing.
Humus' Dynamic Branching Demo
The results
appear to fall in line with what Humus stated above. Naturally, a card would be
directed to operate at the optimal settings given that no image quality
degradation would occur. There were no IQ differences at all that I could
detect. We're talking about a stencil test here though and not textures, AA, AF,
etc...so it's not a surprise that IQ remains constant. What I found to be most
interesting is that, given each card's best approach, the BFG 6800 Ultra OC was 10%~15%
faster than the X800XT PE. The worst-to-best comparisons are quite different but
they're also quite irrelevant.
Some of you like to see
3DMark2003 scores so here you go. The BFG 6800 Ultra OC gets a nice ~8%
performance gain when overclocked to break through the 13,000 point mark.
