Dell 390 memory question / stability with PC2-6400

makeit

Distinguished
Jul 17, 2008
2
0
18,510
Purchased a Dell precision 390 about a year ago and am currently experiencing some stability problems and could use some guidance. I bought it because it was inexpensive and supposedly very stable. I upgraded it to the last available BIOS, a G0 Q6600 with proper cooling, 4GB A-Data DDR2-800, a 850 watt PS and a BFG 8800 GTS OC2 (640MB) – When it works, the machine seem capable, but I will get application faults and the occasional blue screen if I really stress it (i.e Crysis, Codemaster’s Dirt, etc).

From reading a zillion posts it seems like a possible memory issue, though memtest 3.4a runs fine with no issues – I updated the 1GB of DDR2-553 which came in the machine from the factory to the DDR2-800 1.8v because the i975x chip set says it supports it, however in any memory configurator where I specify a Dell 390, it never offers anything above DDR2-667 – I’m wondering if the memory I purchased may be good, but the timing or whatnot causing an issue? The Dell BIOS is striped of any useful information and wondering if there is some definitive way to test the memory compatibility with the machine?

CPU-z says my q6600 running at 9x, 266 bus speed, and 1064 FSB. My memory appears to be running at 332.5MHz, at 5.5.5.15/21 – In cpu-z when I look at SPD, only 200, 266 and 400MHZ are listed for the installed memory – Where does 332.5 come from?

Apologize if this is simple stuff for some of you but I’m confused – There must be some tool to verify what I have is actually compatible???

Thanks,

Bob
 
The memory is DDR, so the data is transferred at twice the base clock rate. So, 332.5MHz = 665 MT/s. So, your memory is running as 667, not 800. The 400MHz is for the rated 800MT/s speed. Sounds like your memory is running fine at below the rated speed.
 

makeit

Distinguished
Jul 17, 2008
2
0
18,510
Thanks for the input. dug further in cpu-z and PC Wizard and see that the memory is reported currently at PC2-5300 - I was actually hoping this could be a cause of the issue of program faults so I could stop looking - I guess its back to the drawing board...

Thanks again,

Bob