OCing help please

[SpateD]

Okay, so I decided to finally take it on since I have downloaded the things I need on my HDD to run everything I need... Since I'm on Vista 64-bit, I can't run CoreTemp, so I'm running SpeedFan, and I'm monitoring it to not go over 50C.. I have pushed the Vcore to 1.4575v. It's been running Prime95 for over 15 minutes now. It's staying at 48-42. At test 13, Core 0 & 1 both hit 50C. While Core 2 & 3 were at 43C.

So my questions are as follows...
1. How come my 4 cores are in such a difference in temps, when all the screws have even pressure on them?
2. Does this mean, I need to move my Vcore down from 1.4575v because it hit 50C on 2 cores after 15 minutes?
3. If so, how low do you suggest I turn down my Vcore?
4. What next shall I ben doing in order to make sure, this is a safe PC to be gaming on for multiple hours?

Thanks in advance everyone.

[SpateD]

Dang, all night and nothing?

[MpG]

A seven temp degree difference between pairs of cores is actually quite typical for quad-cores. Have you configured Speedfan to add 15 degrees to your temps? I'd normally expect higher temps then yours, with that kind of voltage. 65 degrees is a good limit to work with as far as temps are concerned. If Prime doesn't push your temps beyond that, likely nothing else will.

As far as system stability is concerned, you'll want your computer to be able to run Prime for at least eight hours straight without it crashing. 12 or 24 hours is even better.

As for Vcore, you want it as low as possible, while still being able to maintain stability at your clock speed. No two processors are exactly alike, unfortunately. It'll take some experimenting and time to see how high you can clock yours, with how little Vcore, and all without overwhelming your cooling system.

[SpateD]

I set it to 1.40v last night, with 380 FSB, and it ran Prime 95 for 32 minutes until I stopped it to try and lower the Vcore a bit. The temps were at 44-45C on Speedfan at full load.

[Mars]

Originally Posted by SpateD

1. How come my 4 cores are in such a difference in temps, when all the screws have even pressure on them?

Temperature sensors in PCs tend to suck, they're just not calibrated well. The actual temperature is probably not very different between cores, if you're loading them equally.

Software temperature readings are just a guideline in any case.