1800XT CF vs. single 1900XT ?

[WarlordBB]

I've been so focused on CPU's and RAM that I wasn't really paying attention to my vid setup. Just settled in on my CPU and RAM so I fired up my 1900XT and my 1800XT CF cards for a few rounds of 3dMark05 and look what I found out:

System Config:
CFX3200-DR
Opty 165 @ 300x9 = 2.7GHz, stock volts
G.Skill 2x1GB kit at 245

GPU's are all stock settings with no Windows tweaks, CAT 6.4's:

Single 1900XT avg. of 3 runs = 11983

Dual 1800XT CF avg. of 3 runs = 12941

(quick runs with ATI Overdrive OC settings of 690/800 for both get me about another 200 points in both situations)

So, I know this isn't exhaustive or even real-world testing but if my other testing shows results similar to the above...

I've been out of the vid forums for so long, is the 1900XT's just that much better or is it simply that 3dMark05 can't show me the real world benefits of CF? Or, do I just need to do a _lot_ more testing?

 

[Drexion]

A quick search on google shows that your 1800xt crossfire score is pretty much dead on target.

Its pretty common to see something like this in DirectX apps, cause the cards are not being stressed at that resolution, so your CPU is being the bottleneck.

If you run the tests at 1600x1200 or something like that you should not be bottlenecked by the CPU as much.

 

[Croak]

And the single 1900xt STILL has more shader pipes than your pair of 1800's, which matters in 3dMark05

Run 3dMark06 with the two setups and compare results, I think the 1900 will still hold the lead, and maybe widen it.

But long story short, yeah, the 1900 is just much better, period.

 

[Blacklash]

Two X1800 are nothing to sneeze at in actual game play:

Single card:

Crossfire: