OC'ing my Pentium D 820 2.8ghz

MicRaDoT932

Member
i am trying to OC my cpu and in my bios it has 3 different Voltage settings. the first one is .05 then .10, then .15 i was wondering wut one i should have it set at wen OC'ing? i have my cpu at 3.0ghz right now and i put the voltage to the .05 should i have it at that or a higher one. i plan on OC'ing higher wen i get these voltages figured out. and my mobo is the Abit IP-95
 

oregon

Active Member
I would assume those are the increments that it raises the voltage at. So .05 raises it just a bit, and .15 raises "alot". I believe 1.35 is the stock. I used to have the Pentium D820 (2.8GHz), and I got it up to around 3.15 using just the performance enhancer tool. So I would imagine if you overclock manually you should be able to hit 3.25. The max temp for it is probably around 70 under load. So I would up the FSB until it doesn't boot, then up the voltage by .15, check temps under load, then repeat. And once the temps get above sixty I would go down to .1 or .05.
 

fade2green514

Active Member
the lower the voltage the better. voltages will burn your computer up.

overclocking raises heat but it will become unstable and refuse to boot before it burns anything. stick to overclocking, instead of overvolting if you dont have after-market cooling.
 

oregon

Active Member
ummm....i think your confused here. overclocking does not raise heat, overvolting raises heat. overclocking decreases stability, and overvolting increases stability. so you want to find a good balance of stability and heat. essentially the max voltage at which it stays under the safe temperature, and the max stable frequency at that voltage. know that voltage does increase wear and tear, though generally it will decrease the lifetime from 10 years to 5 years, or something like that which won't worry most of us.
 

fade2green514

Active Member
All electronic circuits produce heat generated by the movement of electrons. As clock frequencies in digital circuits increase, the heat generated by overclocked components also increases. Due to increased heat produced by overclocked components, an effective cooling system is necessary to avoid damaging the hardware.
source
 
Last edited:

oregon

Active Member
The movement of electrons described in your quote refers to voltage. When an electron moves, it faces resistance, thereby losing energy in the form of heat. Increasing the electron flow (voltage) will increase heat. Overclocking increases the work the CPU is doing, and so it requires more energy (higher voltage). If that higher voltage is not supplied then the stability decreases resulting in crashes usually. I think wikipedia is grouping overvolting into the process of overclocking (for most people it is).
 

TrainTrackHack

VIP Member
ummm....i think your confused here. overclocking does not raise heat
Higher frequencies ALWAYS cause temperatures to go up, whether you bump the voltage or not.

EDIT: ALso, wikipedia has separate articles on overclocking and overvolting.
 

fade2green514

Active Member
yea, that article was about overclocking and had nothing to do with overvolting... they clearly stated that overclocking raises heat.
 

oregon

Active Member
i did read the quote.... i just don't understand how overclocking alone would raise heat. i understand that you guys are probably right on this, but can anyone explain the why?

i always thought if the electron flow and resistance remained constant, then the heat produced would stay the same as well. what does raising the frequency do that causes the heat to increase?
 

TrainTrackHack

VIP Member
I don't know if this is correct, but it would kinda make sense if increased frequency meant pulses of current being shoved through more frequently, hence more electrons going through = more heat...or something along these lines?
 

hermeslyre

VIP Member
Each clock cycle produces (or fails to adequately dissipate) small amounts of heat as a byproduct of Transistor on/off cycling, and such other things. The wattage (heat) required to do logic operations is small, but together can be noticeable. Increasing the clock signal thus creates more heat, by what amount is variable. Some claim to see very little change in temperature based on overclocking alone (So i've gathered from reading on different forums); I myself don't see noticeable increase in heat on a .8ghz overclock on my X2, but my chip runs hot as it is, and software sensors might not be the best of judge. :rolleyes: (this is how I understand it)

Now increasing the voltage is a much faster and efficient route to increased temperature. It certainly is the primary culprit when dealing with increased temperature IMO.

Tl;dr: What hackapelite said.
 

fade2green514

Active Member
basically clock frequency increases heat linearly (2ghz = 2x heat, 3ghz = 3x heat) whereas voltage increases heat exponentially.. the higher it is, the higher then next .05 volts will increase heat.

clockspeed increases frequency.. as in how frequent the processor calculates numbers or instruction sets. the more frequently it does this, the more often it uses its electricity, and the more often it has to give off the used electricity as heat.
 
Last edited:
Top