After getting to play the game all weekend, thank u best buy

I decided to get down to biz & run some benchmarks. Let me start out by saying the game is everything I expected it to be & more. It is def living it to its hype for me & amazed by it scalibilty. It runs great on my system & upgrade is not needed. Here is my work in progress
My System:
All Default Voltage
Mobile 2600 Barton @ 2.2Ghz
NF7-S 200x11
Radeon 9800 NP 325/290
2x512 Hyper-X pc3500 2-2-3-11-1T
WD Raptor 36g
Audigy 2
Windows XP Pro SP2 RC2 (2612)
To benchmark D3, bring down the console by pressing CTRL + ALT + ~. Then type timedemo demo1.demo You should run the test atleast twice as the second score will be higher since it gets cached in the memory.
All benchmarks were run with all advanced options checked & vsync checked, AA disabled. Drive CP settings were High Quality for both, Vsync always On, Truform Off. I ran each benchmark 3 times & then would defrag & reboot. Used driver cleaner & ghost image for driver changes. agp is agp latency. I found setting my agp latency to 32 (same as bios pci clks) helped solved some stuttering issues for me with games/movies. It did not affect any synth/game benchmark scores, just solved stuttering problem. I found some very intersting results & have more drivers to test. Here are my results so far:
1st test.. compare High Detail vs Medium Detail w/ 8pAF
4.8 Beta
32 agp 1024x768 No AA/AF High Detail Triple Buff On = 36.1/39.8/39.9
64 agp 1024x768 No AA/AF High Detail Triple Buff Off = 30.6/32.4/32.4
128 agp 1024x768 No AA/AF High Detail Triple Buff On = 37.3/39.8/39.9
32 agp 1024x768 No AA/8pAF Medium Detail Triple Buff Off = 26.6/27.7/27.7
64 agp 1024x768 No AA/8pAF Medium Detail Triple Buff Off = 27.0/27.7/27.7
64 agp 1024x768 No AA/8pAF Medium Detail Triple Buff On = 32.1/33.0/33.0
Very suprised by this test here. I found High Detail to look alot better & scores were higher (5.2 fps

). I wonder what kind of AF D3 uses on High Detail? You also get better textures on High Detail, so I recommend using High Detail over Medium w/ 8pAF. Test also shows Triple Buffering gave me a 7.4 fps gain on High Detail & a 5.3 gain on Medium w/ 8pAF

. So I will now keep that on for the rest of my benchmarks. I also found when the timedemo runs the 1st time, It slows down very badly & almost pauses on certain spots like when gettings hit with the 1st fireball & switching to your flashlight goin around the corner to kill 2 zombie soldiers. I only have the slowdown with the 1st benchmark for each as I read after 1st benchmark, it becomes cached in the memory. This would explain why 2nd & 3rd benchmark scores are higher. I found as I increased the agp latency, it would help fix the slowdowns in the 1st benchmark & also scored higher. Once it was cached in memory, it didnt affect scores. I will now try raising the agp latency higher & compare more results. (btw.. agp lat default is 255)
-------------------------------------------------------------------------------------
2nd test... Overclock 2.2ghz vs 2.5 ghz
4,8 Beta
200x11 1.575v VCore
32 agp 800x600 No AA/AF High Detail Triple Buff On = 47.4/54.2/54.1
128 agp 800x600 No AA/AF High Detail Triple Buff On = 49.7/54.2/54.2
255 agp 800x600 No AA/AF High Detail Triple Buff On = 49.7/54.3/54.1
32 agp 1024x768 No AA/8pAF Medium Detail Triple Buff Off = 26.6/27.7/27.7
64 agp 1024x768 No AA/8pAF Medium Detail Triple Buff On = 32.1/33.0/33.0
200x12.5 1.675v VCore
32 agp 800x600 No AA/AF High Detail Triple Buff On = 48.4/55.3/55.3
255 agp 800x600 No AA/AF High Detail Triple Buff On = 51.1/55.3/55.3
32 agp 1024x768 No AA/8pAF Medium Detail Triple Buff On = 31.7/33.1/33.1
255 agp 1024x768 No AA/8pAF Medium Detail Triple Buff On = 32.2/33.0/33.0
I only seem to gain 1fps from 300 mhz oc. I think I will chose to run it @ 2.2ghz as D3 engine is sensitive like nothing we have seen b4. I did not notice any graphical probs @ 2.5 but why even risk for 1fps. I found I could not overclock my vcard at all (white dots & weird textures). I normally run it 380/325 w/ no problems in any other game. Now I really can't wait to get my new card

-------------------------------------------------------------------------------------
3rd test.. 4.7 Omega vs 4.8 Beta ( all test are back to 2.2ghz, 200x11 1.575v Vcore)
4.7 Omega
32 agp 800x600 No AA/AF High Detail Triple Buff On = 40.1/44.0/44.3
255 agp 800x600 No AA/AF High Detail Triple Buff On = 37.6/43.4/43.5
32 agp 1024x768 No AA/AF High Detail Triple Buff On = 28.1/30.1/30.3
255 agp 1024x768 No AA/AF High Detail Triple Buff On = 27.2/30.0/30.2
4.8 Beta
32 agp 800x600 No AA/AF High Detail Triple Buff On = 47.4/54.2/54.1
255 agp 800x600 No AA/AF High Detail Triple Buff On = 49.7/54.3/54.1
32 agp 1024x768 No AA/AF High Detail Triple Buff On = 36.1/39.8/39.9
128 agp 1024x768 No AA/AF High Detail Triple Buff On = 37.3/39.8/39.9
WOW... I had a 10.2 fps drop at 800x600 & 9.6 fps drop at 1024x768. I did not notice dif in image quality, but did not take screenshots to compare. But nietherless, I will not be using the 4.7 Cats w/ D3. I have always used omega drivers over original catalyst, so I will not be tesing those. I also found that lower agp latency seems to scores higher on 1st benchmark w/4.7. Opposite of the 4.8 Beta.
I am going to do some more testing with older sets, but I need to get some sleep. I will post some more results later.