high Vcore worse for me?

[ketchup318]

I'm on my last phase of OCing my chip. It's pretty much settled into the setting that i have on my sig (below). Just for giggles, i was thinking what if i bumped the vcore to 1.95 and see how much more i can squeeze out of it...

well it turns out that at 1.90, 1.95 the max freq. that the chip will accept is somewhere along 2200Mhz. The temps skyrocket to 62C loaded.

I was wondering at these high vcore settings, is it because the temps are so high that it's causing prime95 to fail, or is it current leakage or something esoteric like that? Thanks

 

[greenman100]

could be:

PSU can't handle the voltage at that much current and stay stable

heat causing problems, directly related to higher vcore

 

[Gautam]

Vantecs are generally high quality supplies.

Looks more like something could be up with your loop. Is your waterblock mounted well? No kinks, tiny leaks, etc? I went up to 2.1v on water myself, always stayed below 50. Nothing special either.

 

[greenman100]

Vantecs are generally high quality supplies.

Looks more like something could be up with your loop. Is your waterblock mounted well? No kinks, tiny leaks, etc? I went up to 2.1v on water myself, always stayed below 50. Nothing special either.

80mm rad can't handle that heat

but I bet the TIM is crappy, too

or something else

 

[ketchup318]

so i guess it's a temperature related issue. well that's something i can easily fix. i've already ordered a black ice extreme with a good 120mm fan. i guess if i use that in series with my little 80mm rad and fan, it should lower temps a good deal.

 

[Josebmw]

The temps could not be accurate either, the NF7 2.0 I have puts out at least 10 C.+ than its supposed to sometimes, so don't trust the readings. 1.95v is ok with water cooling. Yes, the black ice is pretty good, since you can have a relatively lower fan noise and have higher performance.

 

[K4mui]

The temps could not be accurate either, the NF7 2.0 I have puts out at least 10 C.+ than its supposed to sometimes, so don't trust the readings. 1.95v is ok with water cooling. Yes, the black ice is pretty good, since you can have a relatively lower fan noise and have higher performance.

No, it's not reading 10+, it reads just bulls*it.
My CPU's at 46°, my HDDs at 34- yeah, sure. It's the same water running through the coolers, so something seems to be stupid, right?

The only way to find out how "good"/ precise it reads is letting the temperature reach a critical level- my PC will shut down if it reaches 65°.

 

[ketchup318]

so i take it that hardware doctor and motherboard monitor aren't too accurate? that's kinda surprising to me. i mean i was figuring plus or minus a couple of degrees, not like 10....

 

[Josebmw]

It's not that the motherboard monitor, or speedfan are not accurate, it's the motherboards sensors that give the temperatures to them that are not accurate. The NF7 has this issue, sometimes it reads 36 C., sometimes 54 C. both idle @ 20 C. ambient. It doesn't have to be necessarily +/-10 C., but do not trust them, to get something more accurate place a probe and stick it next to the core (never on top) and that should yield something closer.
K4mui, even though you let the cpu get upto 65 C., it might be shutting down because of the protection in the BIOS that you can set to shut down at a certain temperature, it doesnt mean that the actual CPU is overheating.

 

[Gautam]

80mm rad can't handle that heat

but I bet the TIM is crappy, too

or something else

Aha! Missed the 80mm rad part. I agree, that's too tiny to handle the heat. Upgrading to a 120mm would really be worth your while.

 

[K4mui]

K4mui, even though you let the cpu get upto 65 C., it might be shutting down because of the protection in the BIOS that you can set to shut down at a certain temperature, it doesnt mean that the actual CPU is overheating.

Now we come to the actually interesting point- for the emergency- shutdown the internal diode is used
At last I can tell that for sure

 

[greenman100]

No, it's not reading 10+, it reads just bulls*it.
My CPU's at 46°, my HDDs at 34- yeah, sure. It's the same water running through the coolers, so something seems to be stupid, right?

no

HDDs make 15w heat
CPU makes 100w

c/w of HDD block...0.15 C/W?
c/w of CPU block, best case scenario, 0.15C/W

so

expect CPU 15C over water temps
expect HDD 2.25C over water temps.

water temp of 32?

HDDs at 34
CPU at 47

Physics says your mobo isn't too far off

 

[KKAT]

try a bios flash. same thing on my nf7-s v1.1! thought they would have fixed it by now.

 

[Josebmw]

KKAT, a bios flash could fix the temp thing on the 2.0 boards?