I think you are wrong and just cocky,, I think there is a way but, have yet to stumble across the thread with the info...
Sorry, heheh... but you did say another ~4000 points
Yes I know what you're talking about but, there's really not that much there!
The XTX had high-spec memory compared to the XT (otherwise identical) and I remember the bandwidth increased from ~52 to ~54GB/s or whatever - I forget but as a percentage it was a very small increase, and I thought not worth the extra $100 bucks at the time.
The memory timings are proportional to the memory speed in MHz though; in other words the XTX memchips could run much tighter timings if clocked at XT memchip speeds.
But the XTX memchips have been selected for speed, and run at higher speed, so will be no easier tightening the timings than it would be on XT, or any other vidcard.
Your best 3DMark scores will result from a fresh-booted rig, running lean/clean and tight. Vidcard OCing is easier when increasing vidcard memchip speed; timings however are more 'obscure' and must be safe enough to deal with heavily OC'd vidcard!
So, you're in tough there...
But look, maybe ATI Tool or whatever can fiddle with your settings - rock on! - but take good care, promise me you won't flash the vidcard with new 'miracle' BIOS or overly aggressive settings... don't trash it
I used to OC in ATI CCC but stability is #1 most important thing - Oblivion doesn't care how many 3DMarks you get, when it
crashes - so on the latest CCC I now just leave it non-OC'd.
Anyway: look to your overall rig if you want the big 3DMark numbers. Working with vidcard memchip timings is fraught with deadly perils! LoL
L8R