when is an overclock not an overclock?

Wigwam

Senior member
Dec 26, 2002
943
0
0
sorry if this has been discussed before - i did a quick search and didnt see much.
just built a 2100+ system. i have only kept the stock hsf at the moment.
what i want to know is what factors contribute to increased cpu temp if all else stays the same in the rig.
for example 12x166 is about 2mhz and 200x10 is 2mhz.
even though i didnt intend to o/c at the moment, i thought it would be rude to my machine not to up the fsb from 133 to 166 [though i did drop the multi from 13 to 12 as i was being cautious!] and now i have a 2400+ and no increase in temp.
i know the higher the fsb the faster the machine, so would it be same to up the fsb if i drop the multi without worrying about heat and having to up the vcore, or not?

i would be very interested to know more about this aspect so i can plan for informed o/c decisions in future [or maybe tonight ;)] but to do so safely.
all advice or links appreciated
 

MrEgo

Senior member
Jan 17, 2003
874
0
76
12x166 = 2GHz.. not MHz.. I'm sure that was just a mistype.

The Voltage increases the temperature more than anything.. I'd actually try putting your multiplier back to 13 if you say there is no increase in temperature. If your system keeps staying stable, then there is no reason to increase the voltage. But when you notice some instability, it's time to increase the voltage by small amounts (0.025, 0.05 V incraments).

Be careful with the fsb vs. multiplier situation. I would personally think it's better to have a 2.167 GHz machine @ 133 MHz fsb than to have a 2 GHz machine @ 166 MHz fsb. But if you placed two identically clocked CPUs (example: each clocked to 2 GHz), one with 133 MHz fsb and the other at 166 MHz fsb, the one with the 166 MHz fsb would definitely be the better CPU.

Hopefully your motherboard has AGP/PCI lock, or a 1/5 divider.. or else you could be in some trouble. Try to have your temps be under 60C if you can.

Edit: I typed @ instead of 2.