• 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.

Low-temp Voltage Question

mezz

n00b
Joined
Oct 1, 2007
Messages
25
I have my Q6600 G0 watercooled and it's running great at
3.6ghz
1.46875v bios/1.408v idle/1.344v load.

I am trying to raising it to 3.8ghz or higher, but it needs MUCH more voltage to be stable.
I am still unstable (prime 95 testing) at 3.8ghz 1.568/1.504/1.424v
I have my memory timings loose (5-5-5), overvolted 1:1 ratio memory +0.2v ,and my chipset +0.2v just to eliminate those variables.

I don't want to go much higher in voltage because Intel says that operating voltage is up to 1.5v, and the threshold-of-doom I read about in the technical paper is 1.6v.
My temps are extremely low though.
My overall temp is 47C, and the cores are 59/54/54/59 after a long long session of prime95. (I know they're uneven and I'll lap it later, but that's another day ;) )

In addition, I never get the temps nearly as high as prime95, even when I'm rendering fractals for days using 100% CPU, so I really want to run the processor faster to the point that the temps are at a good medium (60, up to 70 on cores) temperature. My fans are all fixed low speed, so it won't be loud.

My only concern is the voltage. What happens when your voltage gets too high? Will my processor simply be unstable and restart like when it's undervolted? Or will something much more subtle, drastic, and unsafe happen? The cruel reality is probably that my chip simply wont run at that FSB, but I am hoping I can safely up the voltage to at least see.

So far prime95 has never had single-core errors in all my messing around. The computer either crashes completely in a few hours from undervoltage or is totally stable. To me this is a sign that I can push it further until I find my weak core.
 
What happens when your voltage gets too high?

The CPU is damaged and no longer works.

C2D;s have internal temp protection and will just shut down, cool off, and probably no harm done. There is no such provision for voltage, if you exceed the breakdown voltage of a transistor junction and it fails, that transistor junction is gone, forever.

Intel Absolute Max Vtt on a C2D is 1.55 volts, be aware that at the high voltages you are running the normal voltage spikes from the cpu voltage regulator as it overshoots the target voltage tying to maintain a stable voltage when the load current to the cpu rapidly changes (like firing up Orthos or Prime and the cpu goes from Idle to near 100%) could put your CPU at risk. The amount of overshoot just depends on the board design and we are not given that information. Assume 50mv so an actual Vcore (Vtt for the cpu) of 1.50 volts could be deadly. Relying on software to measure voltages at the top end of maximum voltage ranges can also be deadly.

References:

Voltage Regulator-Down (VRD) 11.0 Processor Power Delivery Design Guidelines - Intel Corp.
 
The CPU is damaged and no longer works.

C2D;s have internal temp protection and will just shut down, cool off, and probably no harm done. There is no such provision for voltage, if you exceed the breakdown voltage of a transistor junction and it fails, that transistor junction is gone, forever.

Intel Absolute Max Vtt on a C2D is 1.55 volts, be aware that at the high voltages you are running the normal voltage spikes from the cpu voltage regulator as it overshoots the target voltage tying to maintain a stable voltage when the load current to the cpu rapidly changes (like firing up Orthos or Prime and the cpu goes from Idle to near 100%) could put your CPU at risk. The amount of overshoot just depends on the board design and we are not given that information. Assume 50mv so an actual Vcore (Vtt for the cpu) of 1.50 volts could be deadly. Relying on software to measure voltages at the top end of maximum voltage ranges can also be deadly.

References:

Voltage Regulator-Down (VRD) 11.0 Processor Power Delivery Design Guidelines - Intel Corp.
Thank you very much! I think I will stick with my much safer 3.6ghz.
I think that Vcore of 1.50 volts would be safe since the voltage droop is there specificially to combat overshooting. However, I'm not particularly crazy about pushing for an extra 5% at the risk of destroying my hardware, and my 1.46875v bios/1.408v idle/1.344v load setup should be much safer.
 
Back
Top