Pentium 4 Clock Throttleing

[charliec2uk]

I fell to thinking one dark and wintery evening, huddled around the fire with a good bottle of bran...

Ahem,

What I was thinking was that when the P4 detects its core temperature going outside safe parameters, it throttles its clock speed down until the temp. falls back into safe operating range. How exactly does this work? Does it step down its multiplier or what? But some how the core frequency of the P4 is not as locked as Intel would like us to believe. Therefore it seems that it may be possible for some kind of software utility to control this feature, and , drum roll please, new overclocking opportunity. Obviously this is not possible as some bright spark would have figured this out already. My real question is: Why is it impossible?

Charlie

Democracy Bernad, it must be stopped!

 

[Guest]

It's not that no one else hasn't thought of it. They may have done already - It's more like how do you put it into practice. Try finding an O/C software on the net and see what happens. When it lowers its own temp are you sure its something to do with the multipler - have you looked in your bios - maybe that could be updated to allow the settings that you need to be accessible.

Could be fun. Also remember when tinkering - smoke is NOT a good thing.

Yo Norman - Kick up a STORM.
<A HREF="mailto:techie2000@supanet.com">techie2000@supanet.com</A>

 

[charliec2uk]

I have no clue as to what goes on. It was a topic that was designed to provoke thought.

Democracy Bernad, it must be stopped!

 

[LoveGuRu]

how can you be sute its the miltiplayer and the simple FSB drop?

<font color=green>
*******
*K.I.S.S*
*(k)eep (I)t (S)imple (S)tupid*
*******
</font color=green>

 

[FatBurger]

Simply checking with WCPUID while the P4 is throttling should reveal whether it's the multiplier or the FSB.

<font color=orange>Quarter <font color=blue>Pounder</font color=blue> <font color=orange>Inside</font color=orange>

 

[girish]

Well, get this <A HREF="http://ftp://download.intel.com/design/Pentium4/datashts/24919805.pdf" target="_new">Pentium-4 datasheet</A> and scroll to section 7.3, thats it.

And then <A HREF="http://www.64bits.org/cpuheat/cpuheat1.htm" target="_new"> this</A>

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>

 

[girish]

Damn! THG adds http:// before a FTP link.

read the above link as <font color=blue>ftp://download.intel.com/design/Pentium4/datashts/24919805.pdf[/url]

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>

 

[girish]

Damn! THG adds http:// before a FTP link.

read the above link as <font color=blue>ftp://download.intel.com/design/Pentium4/datashts/24919805.pdf</font color=blue>

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>

 

[charliec2uk]

Oh well, that seems to put the kibosh on that particulary train of thought.

It was a nice idea though.

Charlie

Democracy Bernad, it must be stopped!

 

[charliec2uk]

If I've got this right, the clock speed doesn't actually change. Instead it just restricts the number of operations it actually performs, reducing switching activity and hence heat.

Woke up in Borneo one day, stuck in the pouch of a big marsupial, but the boat was made of marsipan!

 

[girish]

exactly, and as the temp rises, it can even reduce the proportion (duty cycle) of P4 working and being idle.

it can reduce from 100 to 12.5% of duty cycle, i.e. working for only 125 milliseconds in a second in the worst case.

you can see that in the THG heat test video and my report about that, as the frame rate drops from 88 to 11, the same proportion 1/8 is it a coincidence or the P4 did swicth to 12.5% duty cycle?

girish

<font color=red>No system is fool-proof. Fools are Ingenious!</font color=red>