Name: jarhead123 Date: March 18, 2008 at 14:11:16 Pacific Subject: low 3d mark 06 score? OS: windows XP CPU/Ram: Q6600 @ 2.8ghz
Comment:
Hi everyone i have a question. i have 2 8800 GTS 640mb SLIED, Q6600 @ 2.8ghz, 4gb ram overclocked, 7200 RPM hdd, EVGA 680i SLI board and i get around 12000-13000 is that a low score?
use 3d mark 06 free version so what ever the setting are on. i am actually at 2.9ghz and the memory frequency is at 967MHZ all I did is keep memory and CPU linked and adjusted the FBS a little at a time
jam, he's probably talking about his combined 3DMark06 score (the number that 3DMark spits out at the end), not just the SM 2.0 portion of the benchmark that was compared on Tom's.
jarhead, that score does seem a little low for a q6600 @ 2.8, especially since 3DMark 06 is nearly as sensitive to CPU speed as it is to video performance. My system, a 3.2GHz dual-core Opteron with SLI'ed 8800GTS-640 cards, produces around 13, 900 in '06. Of course my video cards are running way above the reference clock rates. But your faster, quad-core CPU should easily make up for your slower, reference-clocked video cards.
I'm aware that you listed WinXP as your OS, but I've learned not to trust what users write up there. :P
If you're running Win Vista, then lower scores can be expected. Vista likes to chop 500-1000+ points off of a high-end system's 3DMark score. My 13.9K score was generated by XP; under Vista-64, I get 12.7K.
Other things to consider: -Did you run at the benchmark's default display settings? -Did you make sure that no background tasks were robbing CPU cycles during the benchmark? -Does your motherboard provide each video card with 16 PCI-E lanes? Reason I ask is that some budget motherboards only provide 8 PCI-E lanes to each video card while running in SLI mode. High-end GeForce 8 cards are known to suffer slight performance penalties running at PCI-E x8.
EDIT: Ignore what I said about the PCI-E thing. I just re-read your post and noticed the i680-based mobo. That mobo definitely provides 16 lanes to each video card.
The creme de la creme of Socket 939: Opty 185 @ 3.2GHz SLI'ed GTS-640s, both flashed to 625/1458/1950 4GB PC3200 Blu-Ray/HD-DVD, X-Fi A8N32-SLI Deluxe 3DMark06: 13896
With default settings, my Q6600 @ 3.2GHz and single 8800GT scored 13106, so I am thinking along the same lines as jam, exactly what have you got and how did you get to 2.8GHz.
"i am actually at 2.9ghz and the memory frequency is at 967MHZ"
OK, let me see if I can get this straight. Your CPU runs at 9 x 266MHz by default. If you didn't change the multi, the CPU freq is approx 322MHz. And by leaving the RAM linked at it's default speed, it went from 400MHz (DDR800) to 483MHz (DDR967).
Try this...leave it linked but manually set the RAM to DDR533 (266MHz). Leave the CPU multi at 9.0x & increase the CPU freq to 333MHz. That will put the CPU at 3.0GHz/1333MHz FSB with the RAM running as DDR667, giving you a 1:1 ratio.
when i put in your settings, i got a score of 13013, i did what u said put my ram to 667 (while i had the FSB to 333) then i lowered it to 1295 FSB so the temps would go down
How do you "get a spike in volts"? You're certainly not gonna fry the CPU at 60C+. The C2D is thermally protected anyway...it cannot be overheated to the point of failure. It will auto-shutdown long before it will fry. Regardless, if you say you can't do it, I'm not gonna argue.
Try lowering your RAM speed. The Core 2 performs best with the CPU:DRAM freq ratio at 1:1. If you're gonna keep the CPU clocked at 9 x 322MHz, lower the RAM speed to 322MHz (DDR644)
I might be steering you wrong on the linked/unlinked issue. Just do what you gotta do to make the CPU & RAM frequencies the same. The keyword is "frequencies"...don't worry about the QDR & DDR numbers. In other words, if the CPU is at 322MHz, the RAM should be at 322MHz (I still think you should be able to hit 333MHz).
Also, make sure you're locking the PCI-e at 100MHz, the PCI at 33MHz, & disabling ALL Spread Spectrum options.
The information on Computing.Net is the opinions of its users. Such
opinions may not be accurate and they are to be used at your own risk.
Computing.Net cannot verify the validity of the statements made on this site. Computing.Net and Computing.Net, LLC hereby disclaim all responsibility and liability for the content of Computing.Net and its accuracy.
PLEASE READ THE FULL DISCLAIMER AND LEGAL TERMS BY CLICKING HERE