Q6600 stuck at 3.2

herpsuk

herpsuk

herpsuk

Associate
Joined
19 Feb 2009
Posts
10
Location
Preston, UK
HI guys

So im currently running :
Q6600 @ 3.2 - Stable for 1hour of OCCT (I will be running an overnight test aswell)
Ultra 120 eXtreme
EP35-DS3
1 x 2GB XMS2 8500 (1066)

CPU set at 9 x 356.
PCIE - 100
Memory Multiplier - 3 (1068)
vCore - 1.375
100% Load - 50/51c

I upped the FSB to 361 and the computer will boot but i cannot get it past 4 Mins of OCCT. The screen went blank and it rebooted.I think it also disabled the editing of the FSB and put it back to 266 but kept the vcore as it was.

To counter this i upped the vCore one notch in the bios (i think it was 1.38 or something like that) but it would not boot and kept booting until it disabled the editing of the FSB again and reset it to 266 and kept the vcore as it was.

What am i doing wrong here?

any help would be appreciated

Jon

 
Last edited:
What are the rest of your voltage settings in the BIOS? Just changing vcore will not get you a good overclock, there is a lot more to it than that.

Also, you may need more vcore, depending on how much it drops under load.

What are you using to measure the temps? Make sure to use RealTemp, CoreTemp, HWMonitor or Open Hardware Monitor for good accuracy. The latter 2 should also show you board temps and maybe even PWM temps.
 
Thanks for the reply.

Ive adjusted no others. Voltages are as below:





Software wise i use CoreTemp, CPU-Z, OCCT, SuperPi, PCWizard. So have a wide variety of software for Showing Temps and such.

Regards

Jon
 
Last edited:
I cant coose a voltage on the fsb. Can only up it by 0.1, 0.2 & 0.3.

Just up it by 0.1?

Never tried linx, might give it a go.

Thanks :)
 
FSB and MCH will both likely need a little bump, although moreso the MCH I would think. Raise MCH 1 notch and try again. If it fails, raise one more, then if it fails again, raise FSB and vcore 1 notch.

Try it again and report back
 
Thanks a lot for that, got it past 3.2 (just lol) :D.

8 x 410 (3.28ghz)
Memory Multiplier - 2.5 (1025)
BIOS vCore - 1.42500v
FSB OverVoltage - +0.1
(G)MCH OverVoltage - +0.2
Core 1 & 2 @ 100% - max69c
Core 3 & 4 @ 100% - max63c
Idle - 36-38c

Ive been using OCCT but i found the standard CPU test not as intense as the LinPack test so ive started using LinPack, usually crashed within minutes if their is something wrong.

Stable for 1 Hour of LinPack.



Just trying 8x415 with linpack as i type.

What are peoples views on voltage? ive read that +20% is the max i should look at. So thats 1.59v? seems a tad high to me. Whats best for daily usage?

Regards


Jon :D
 
Last edited:
I wouldn't go much over 1.5v unless you are water or sub zero cooled. Preferably closer to 1.45v for 24/7 use with air.

Prime95 Small FFTs, or blend are also good stress tests.

And if you REALLY want to heat that CPU and test it just about as hard as it gets, you could use Intel Burn Test (IBT). It gets my CPU quite a few degrees higher than Prime95, and will very quickly crash if not 100% stable.

In fact it is so picky that my CPU (which is 6+hrs Prime stable) crashes within 3 tests of IBT. It never crashes under anything else though, so I left it as it is. I think another tiny bit of vcore would stabilise it, but I am quite high already.
 
Last edited:
Indeed, LinX is a GUI for the same Linpack algorithms as (the original) Intel Burn Test. It seems the back-end of the newest IBT (version 2.5) are more up to date as the LinX guy has given up. You may want to switch to IBT 2.5. Forget about p95, it's old hat.
 
Thanks again for the help. Ill try upping it some more tonight and see what i get. Think ill need to look at some new thermal paste as well (about 3 years old :rolleyes:). Probs going to be restricted by the temps though, contemplating upgrading from the TRUE but i dont know if ill benefit much from a Pro Megahalem or not?

Ran IBT this morn and it is stable under standard for 5 runs and stable under Maximum fo 5 runs. Temps similar to that of LinPack. How many runs should i look at doing of IBT?
 
You must log in or register to reply here.
Back
Top Bottom