• 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.

Q9550 overclocking woes

Exeodus

Exeodus

[H]ard|Gawd
2FA
Joined
Sep 26, 2008
Messages
1,124
I am at my wits end. I just purchased a Gigabyte EP45-UD3P and a Q9550 to replace my Q6600 that just died on me. I am trying what seems to be a very mild overclock for this chip, 3.4ghz (400x8.5). I can't seem to get this stable to save my life. I have used bios F6 and F8 and have the same results. The Q9550 is a E0 chip, and I have upped to voltage all the way to 1.34 in the bios, about 1.29-1.3 in windows according to CPU-Z.

I use OCCT for testing, and it never gets past 18 mins without rebooting. I never get a error in OCCT, the computer just reboots. I am using 4gb of GSkill (2x2gb sticks), and I am running them at 1066 spec, 2.1 volts. After one of the reboots, I ran Memtest with these settings, and the ram did not produce any errors. I have tried leaving the voltages to auto, turned VTT up from 1.2 to 1.26, MCH from 1.1 to 1.3, but none of these helped. I tried turning on loadline calibration, but this causes the motherboard to reset when posting, and it tells me that my overclocking or voltage settings fails.

My power supply never drops under 12.3 volts when loaded, it is a OCZ GameXstream 700watt.
I am running Vista64, and the OS seems fine. no data corruption from what I can tell.

Also, the hottest any core got was 52, according to realtemp, so heat is not a issue.

What could be wrong, I have seen plenty of post in other forums about guys hitting 4.0ghz stable with less voltage, I am only trying 3.4
 
I tried turning on loadline calibration, but this causes the motherboard to reset when posting

Hmm, very very odd. If you turn on Loadline Cal. with stock settings, do you also get any resets/failed posts?
 
MisterSparkle said:
I tried turning on loadline calibration, but this causes the motherboard to reset when posting

Hmm, very very odd. If you turn on Loadline Cal. with stock settings, do you also get any resets/failed posts?
Click to expand...

Interesting point, I have not tried that. I did try different voltage settings, then tried to enable it, was a no go each time. But I was using overclocked settings at the time.
 
Exeodus said:
I am at my wits end. I just purchased a Gigabyte EP45-UD3P and a Q9550 to replace my Q6600 that just died on me. I am trying what seems to be a very mild overclock for this chip, 3.4ghz (400x8.5). I can't seem to get this stable to save my life. I have used bios F6 and F8 and have the same results. The Q9550 is a E0 chip, and I have upped to voltage all the way to 1.34 in the bios, about 1.29-1.3 in windows according to CPU-Z.

I use OCCT for testing, and it never gets past 18 mins without rebooting. I never get a error in OCCT, the computer just reboots. I am using 4gb of GSkill (2x2gb sticks), and I am running them at 1066 spec, 2.1 volts. After one of the reboots, I ran Memtest with these settings, and the ram did not produce any errors. I have tried leaving the voltages to auto, turned VTT up from 1.2 to 1.26, MCH from 1.1 to 1.3, but none of these helped. I tried turning on loadline calibration, but this causes the motherboard to reset when posting, and it tells me that my overclocking or voltage settings fails.

My power supply never drops under 12.3 volts when loaded, it is a OCZ GameXstream 700watt.
I am running Vista64, and the OS seems fine. no data corruption from what I can tell.

Also, the hottest any core got was 52, according to realtemp, so heat is not a issue.

What could be wrong, I have seen plenty of post in other forums about guys hitting 4.0ghz stable with less voltage, I am only trying 3.4
Click to expand...

Stable is a very "subjective" term, don't trust everything that you read on the forums. That said some chips don't overclock as well as others. I had an e8400 that needed 1.4+ VCC for 445x9 stable, while my Q9650 only needs 1.344 and they have the same VID.

Just to start out with run your ram at 1:1 with your fsb and set the timings at 5-5-5-15-55trfc-pl10 untill you get your fsb stable then tweek from there.

What is the VID on your chip? What core voltage have you tried?

Also keep in mind auto DOES NOT mean stock.
 
Have you tried other stress tool? like Prime95 or Intel Burn Test?

The fact that OCCT has never failed but only reboot the system is quite odd. Usually CPU errors do not cause instantaneous reboots, but memory errors do. try lower multiplier with the same FSB, see if you can replicate the error.
 
Bakku said:
Have you tried other stress tool? like Prime95 or Intel Burn Test?

The fact that OCCT has never failed but only reboot the system is quite odd. Usually CPU errors do not cause instantaneous reboots, but memory errors do. try lower multiplier with the same FSB, see if you can replicate the error.
Click to expand...

Thanks for your input, it was very helpful. It passed 20 passes of IBT, with error reporting on, set to maximum stress level, and had over 3gb of memory available for testing. I set it to run Prime before I left for work this morning. I was just surprised that OCCT was doing that, I have never had a stress program just reboot to pc without a error. :confused:

Thanks again for your input.
 
I was having ram problems, I think I have the same G.Skill (5-5-5-15 at 1066), and I have a UD3R.

I changed the 3 voltage settings after DRAM voltage to reflect the 2.1V, so termination, CHannel A, Channel B, I set to 1.04 instead of Auto (normal is 0.9V).

I also changed all the Auto voltages to what it said for "normal" - in case Auto was setting some weird voltages. Those 2 changes helped - haven't narrowed down yet which - had to go to work. :(

Also, did you turn off the features that regulate CPU speeds, C1E etc?
 
Exeodus said:
Thanks for your input, it was very helpful. It passed 20 passes of IBT, with error reporting on, set to maximum stress level, and had over 3gb of memory available for testing. I set it to run Prime before I left for work this morning. I was just surprised that OCCT was doing that, I have never had a stress program just reboot to pc without a error. :confused:

Thanks again for your input.
Click to expand...

Ive had IBT cause a BSOD and not even error out on my Yorkfield which was 10+ hours Prime stable. My VCC was too low. I have found Prime to me much more stressful than IBT on my Wolfdale (e8400), though. I was passing IBT (20+ runs) then failed prime small ffts in under 45 minutes. Tweeking my GTLs fixed that. Thats why it's a good idea to use more than one stress testing app. If you are getting BSODs on any stress test your oc is not stable. Set your ram to 1:1 with loose timings, get your fsb stable, then tweek your memory.
 
I think I may have it finally!!!! The GTL needed to be tweaked, finally passed OCCT!!!!!!

Thanks for all your help guys.
 
You must log in or register to reply here.
Back
Top