I benched this system at 15,275 at default settings, but this is at the bottom of the heap of machines with very similar specs, and the discrepancy is huge. Top machine is over 30,000, and yet the processor component of the benchmark really is scaled accordingly, and seems about right since that person is running the QX at 5.7Ghz. It seems to be in the video area, with huge diffs in frames/sec. There are about 15 machines with virtually identical specs as mine, and scores go up in the 25K area. What possibly could explain this? The hardward specs can't explain it, for example the FSB on a 25K score is only about 60 more than what I'm set at. Any clues? Are these bogus scores? Liquid nitrogen cooling? I set my CPU at 4.32GHz, and 800GT at 700mHz.QX9650 w/ Retail HSF|8GB Muskin PC-6400|ASUS P5E|EVGA 8800GT @700|Seagate SATA 2 x 4|Seagate Cheetah 15K.x|XP Pro|Vista 64--soon to be installed

Yes, those high scores do make one think his/her system has major issues. I thought the same thing but I understand those systems with really high scores are overclocked to the max and, in some cases, way beyond safe OC levels. I benched my system at 16020 at default settings. That's the highest I have reached so far. Your system appears to be well-tuned but you should be able to squeeze a few more frames as I think my system is not as powerful as yours. The only thing overclocked in my system is the CPU.My system: (Dell 720H2C) Intel Core2 Extreme X6800 CPU (Overclocked to 3.73GHz)/4GB DDR2 SDRAM at 667/Dual 768MB NVIDIA 8800GTX video cards in SLI mode/X-FI PCI Sound Card/1 KW PSU/Dell 3007WFP LCD Monitor/Dell H2C Cooling Unit/NVIDiA 680i SLI MotherboardBest regards,Jim

I think quite obviously it's mostly the video card. They are showing FPS of 100 where I am seeing them around 55 or 60. When I looked at the "Processor Scores" my machine was exactly where it should be vis a vis the overclock I am uisng. I would assume the only possible way to see 25,000 is with SLI highest end nV cards coupled with QX processors, super cooled or something like that. Your 16K score would be from your GTX I would assume. I am using a GT at 700mHz. QX9650 w/ Retail HSF|8GB Muskin PC-6400|ASUS P5E|EVGA 8800GT @700|Seagate SATA 2 x 4|Seagate Cheetah 15K.x|XP Pro|Vista 64--soon to be installed

"Your 16K score would be from your GTX I would assume. I am using a GT at 700mHz."I did another benchmark this PM because I hadn't done one for quite a while and your post got me interested in seeing if things were better or improving. They aren't. My latest 3DMark06 score is 14992, a little over 1000 below my best benchmark of 16020. Quite a difference. I had the Nvidia WHQL 169.25 drivers installed. I got my best benchmark with the 163.71 (beta) drivers. There has been no change in hardware or other software so suspect the drivers as well as periodic updates to Vista. I ran 3dMark06 w/o SLI enabled too and the score dropped below 12000 so SLI makes a big difference in FPS. At least on my system. Oh well!Best regards,Jim

Are you sure video card controlled AA/AF is off for the 3DMark06 runs? If enabled, it can make this much of a difference.Gary

Look at the SM2 and SM3/HD scores. Those scores are single index number representations of the FPS results that occurred for those two Video benchmarks. For instance, I have a 8800Gt clocked at 700mhz too. I'm getting:SM2 = 4890 (43 FPS-Battle /38 FPS- Firefly)SM3/HDR = 4837 (41-FPS-Leaping fish-lizard thing/53 FPS-Arctic base)I expect these same scores would occur with any 8800GT clocked to 700mhz. These video benchmarks really need to occur separately from any CPU result. Otherwise, what's the point?Liquid Nitrogen cooling is not just limited to CPUs. There is a whole community out there that focuses on massive Gpu O/Cs. If you are seeing huge SM2/SM3 scores - or their directly representative frame rate results - , it's likely some massive GPU O/Cs are occurring.

Although my system isn't really comparable, I can add that my 3DMark scores are always improved by setting the GFX card to "Performance" (ie, no AA, AF etc) and also by removing all but the essential background processes and USB devices - which I do for FS9 anyway...For comparison, the machine in my sig gets 9507 in 3DMark06 (free edition) but this can drop by as much as 1500 points if I don't "clean up" first!Geoff

Good point. I had nHanser setup with 8xs AA and 16x AF. I shut off both and went for another 3DMark06 run. This time, I got a 15% boost and beat the 9800GTX SM2/3 scores. These SM scores can give the general idea, but are pretty basic without a little extra info. The 3DMark's Headline Score is totally meaningless.