post #11 of 11
Thread Starter 
Originally Posted by RonindeBeatrice View Post
Shouldn't it not work this way?

It seems that temps (and energy consumption) rise gradually with clock speed, but virtually exponentially with voltage.

Couldn't you get yourself in a situation where you simply have the voltage too high, and there'd be a little more room for clock speed if you backed it down a bit more?

It seems your method would require another step, where you decrease the voltage a little after you've achieved what you thought was your best overclock, then attempt to inch it forward a bit more.
Now that's a valid point. I only stress test it when I believe I'm at my final overclock though, so temps are never higher than they are at idle. My final overclock is usually something like 200mhz lower than whatever overclock will not let me boot into Windows.

I'm an imprecise and impatient overclocker, I'll admit it.