• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

4 dimms causing system instability

N

NukeULater

Gawd
Joined
Sep 12, 2006
Messages
917
Lately I have been having some major instability problems with my pc, they seemed to be getting progressively worse until it culminated in the system being unable to boot into the bios. After clearing the bios and reseating the ram and any other components it was able to boot again.

Logically, at first I thought it was the 3.5GHz overclock I have on my Q9450 (2.66Ghz stock). Yet, after backing down the FSB timings and prime95 testing it for 9 hours straight I started to look elsewhere. Strangely, the system would run for hours with prime95 small FFTs for testing the CPU, yet in blend mode the test would fail instantly. This led me to run memtest86 to check the ram. Needless to say it failed.

Now this is where I need advice, because it only fails when all four dimms are installed in the machine. If I run two at a time they pass, even after six or more hours of memtest86. Thus, I thought it could be the dimm slots so I tested the ram in each slot. It passed with no problems.

After some forum searching, it seems that higher northbridge and ram voltages might be needed. I bumped up the ram voltage to 2.2V, from 2V. The northbridge was raised to 1.28V from 1.25V. Still the machine isn't stable. :confused:

Has anyone had problems like this before? It seems odd that the ram would test fine with two dimms and fail with four dimms.

Thanks for any help. I'm stumped :)

System specs:
Intel Q9450
Abit X38 QuadGT
8GB G.Skill DDR2 1000 CAS 5-5-5-15
Sapphire Radeon 7950 3GB
x2 Seagate 500GB (raid 0)
x2 Seagate 1.5TB (raid 0)
Highpoint 1690 raid card
x4 Seagate 3TB (raid 10)
PC power & cooling 750W
Windows 7 Pro


Link to the ram I have.
http://www.newegg.com/Product/Product.aspx?Item=N82E16820231145

Screenshot from the 4 dimm memtest run.
http://www.newegg.com/Product/Product.aspx?Item=N82E16820231145
 
Small FFTs doesn't touch the IMC (or FSB/NB in pre-IMC era chips) at all :p
I thought this was common knowledge - and I why at chuckle when any experienced overclockers suggest that passing small FFTs = 24/7 stability haha.

After some forum searching, it seems that higher northbridge and ram voltages might be needed. I bumped up the ram voltage to 2.2V, from 2V. The northbridge was raised to 1.28V from 1.25V. Still the machine isn't stable.
Click to expand...

And that would be the right course of action. You probably need to bump NB further than that, 0.03v is not much of a bump at all. I'd try something closer to 1.35v. You may want to consider directing a fan towards your NB, but with good airflow I wouldn't be too concerned. I wouldn't bump vDIMM at all from stock unless you're overclocking the memory itself.
 
Last edited:
Once you load all the dimm slots increase the voltage a good bit. You can always go down and fine tune it but. But a .2v & .02v increase isn't doing anything. Then add on the faster you're ram is the more voltage you need. Factor in how good/bad a imc you have also will effect the voltage.

* Try around 1.35 volts on both.

I miss my old Abit X38 QuadGT. Was a great board and a great company.
R.I.P Abit
 
Mav451 said:
Small FFTs doesn't touch the IMC (or FSB/NB in pre-IMC era chips) at all :p
I thought this was common knowledge - and I why at chuckle when any experienced overclockers suggest that passing small FFTs = 24/7 stability haha.

And that would be the right course of action. You probably need to bump NB further than that, 0.03v is not much of a bump at all. I'd try something closer to 1.35v. You may want to consider directing a fan towards your NB, but with good airflow I wouldn't be too concerned. I wouldn't bump vDIMM at all from stock unless you're overclocking the memory itself.
Click to expand...
Well that's good to know. Everyone seemed to use prime95 small ffts back when I was overclocking it in 2008. Do you recommend using the blend test or another program for testing?

I'm glad I was on the right track regarding uping the voltages, I just didn't think it would take so much more juice to drive the northbridge.

KILLER_K said:
Once you load all the dimm slots increase the voltage a good bit. You can always go down and fine tune it but. But a .2v & .02v increase isn't doing anything. Then add on the faster you're ram is the more voltage you need. Factor in how good/bad a imc you have also will effect the voltage.

* Try around 1.35 volts on both.

I miss my old Abit X38 QuadGT. Was a great board and a great company.
R.I.P Abit
Click to expand...

Well I raised the voltage across the board. MCH to 1.35V, ICH to 1.15V and VTT to 1.2V. I also have the ram at 1:1 so it's a bit slower than before. So far prime95's blend test hasn't failed. So that's a plus. I'll report later after it runs a while longer.

I agree I miss abit boards as well. The X38 QuadGT was a replacement for my old IN9-32X. That thing was only let down by its appalling 680i chipset...
 
You must log in or register to reply here.
Back
Top