Overclocking a P4 2.53/PC1066/Intel 850E

[SanTropez1971]

I am getting my new system put together today that consists of a P4 2.53GHz processor and a Intel 850E chipset motherboard (Gigabyte 8IHXP). From a previous post I have learned that the only way you can overclock the P4 is by increasing the FSB. Is this true? If so then why do most of the new board have settings to increase voltages for different components (RAM, AGP, etc). Are these voltage settings here to help overclock? If so what procedure should I use to test the different settings?

 

[Cooj]

I don't know much about O/C but here is a link that was posted earlier in the forums that explains the benefits/reasoning behind increasing the vcore.

http://www.hardwareanalysis.com/content/article/1482.1/

An excellent read for all you engineers out there

Cooj

 

[Crashman]

The more work your CPU does, the more energy it consumes. The more voltage you use, the more current you can pass through. So the demand for more power is met by an increase in voltage that allows more current, when overclocking.

<font color=blue>You're posting in a forum with class. It may be third class, but it's still class!</font color=blue>

 

[Crashman]

The thing that I find worst about such accedemic articles is that they downplay the significance of increased voltage in overclocking, saying things like it "may add a little stability" or "it may help you squeeze a few extra MHz out", etc. I've seen instances where I could DOUBLE the clock speed by increasing the voltage. For example, I owned a PIII 700, I wanted to see what the lowest heat output I could get out of it was. At 1.50v, I could get it to run 466 (66MHz bus) and up to 546MHz. At 1.65v it would go to around 870MHz. At 1.95v it could go to 1021MHz stable. You might argue that 1.50v was below spec, well I would argue right back that MANY processors have a LOWERED spec. The same core on my Celeron 566 was stock at 1.55v, and it ran fine at 1.85v/978MHz. The same core revision as my 700 that came stock at 1.65v.

<font color=blue>You're posting in a forum with class. It may be third class, but it's still class!</font color=blue>

 

[Cooj]

Well technically yes you can just keep pumping the vcore higher and higher but if you keep your vcore that high it's bound to do damage to your cpu.

As far as I know from what I hvea heard if you keep it above 1.75v you could permanently damage your processor.

Then again I don't really know much about O/Cing just what I read from there =)

Cooj

 

[Crashman]

It depends on the size of the subcomponenets. Early 486's had a 5v core, later version had a 3.3v core. Pentiums had a 3.3v core. Early PII's had a 2.8v core. Coppermines had anything from 1.50v to 1.75v factory stock, the slower processors were effectively "undervolted" to aid cooling and reduce power consumption. These had no problems with 1.95v in my experience, some people claim to have run theirs at 2.05v for years without problems.

These new processors are on an even smaller process, .13 micron, for which 1.75v seems to be about the limit, just as 2.05v was about the limit for the .18 micron Coppermines.

<font color=blue>You're posting in a forum with class. It may be third class, but it's still class!</font color=blue>

 

[JaeSun]

well, 1.75 is the "known limit" or suggested, but 1.6 is prob the best safe limit .... id only go 1.75 if i had better cooling, like upgraded fan, or even better, water cooling ....

_____________
100 Character Limit Sucks!!