Thanks Pitch for posting cos you gave the other side:
Overclocking is not something like gaming...
It is dangerous if you don't follow some basic steps....
AMD gain CPU speed estimator
As you can see my
CPU burn-in guide at XtremeSystems after burn-in the current 939 3200+ Venice had MSF=2670 MHz with
Vcore=1.663 (32M SuperPi).
Now it passed 2680 MHz!
Room temperature is six degrees lower!
That's why my
CPU gained 1 MHz FSB!
Of course
Vcore is the same.
Note:
This program based on same desktop CPUs (not mobiles).
See help file for more info.
Let's explore more what is going on...
Multiplier is 10X MSF was 2670 MHz and temperature (some days ago) 48C.
Note:
MSI monitor shows anything.
I measured
CPU temperature using thermistor and it is 8-10C lower anyway I am bored to put always probe in heatspreader that's why these temps based on MSI's monitor.
The main parameter is temperature difference not the actual values.
Temp now is 42C...
But as you can see the closest value for temperature is 41C (see second combobox).
Cheat program!
I edited value and I input 42!
Results:
The second region has 33% possibility but lower limit is 2663 MHz < 2670 MHz (initial MSF)..
That's why we refuse this area.
First area has 22% possibility and upper limit 2683 MHz (almost 2680 MHz) --> predicts OK regardless small possibility value.
Here is autumn and winter is coming...
Then ambient temperature will be eight degrees smaller (I keep notes for anything)
so
CPU temperature will be 42 - 8 = 34C.
If we estimate new MSF for Venice in winter fourth area shows 2708 - 2728 MHz --> +40 MHz gain at least.
But it is not impossible that first or second/third area can predict better than fourth regardless smaller possibilities.
All possibilities are < 100.0% --> when an event has 100.0% possibility it occurs always.
I will see!
Note that the screen capture may differ than current version cos any new edition has new options, capabilities etc....