P4 Multipliers - Adjustment theory

[DivideBYZero]

Does anyone here know anything about the way the P4 multiplier is supplied to the motherboard?

I was mulling this over and thought that if you could make a kind of 'intelligent' powerleap style socket converter that provided the multi setting rather than the CPU that was plugged into it.

Also, as the motherboard AFAIK is being told by the proc what the multiplier should be, would it not be possible to write a BIOS that simply ignored the value supplied by the CPU? Obviously, this could not come from the factory like this, but a downloadable update available through 'other' channels.

Wadday think?

 

[CTho9305]

I'm pretty sure that the multiplying is done within the CPU itself, and all the motherboard knows is what string to show the user.

 

[AbsolutDealage]

Wadday think?

I think that this would be better suited to the overclocking / general hardware forums. This is not highly technical.

 

[MikeMike]

Originally posted by: AbsolutDealage

Wadday think?

I think that this would be better suited to the overclocking / general hardware forums. This is not highly technical.

oh yeah? go look at VR zone for their thoughts on unlocking HT and the Multipliers on all P4's!!!

 

[DivideBYZero]

Originally posted by: AbsolutDealage

Wadday think?

I think that this would be better suited to the overclocking / general hardware forums. This is not highly technical.

It could quite easily get technical, as I am not the owner of a P4 and do not intend to overclock one. I am interested in the way that the whole scheme is implemented and how it might be circumvented.

Thanks for you input anyhow..........

 

[CTho9305]

Originally posted by: DivideBYZero

Originally posted by: AbsolutDealage

Wadday think?

I think that this would be better suited to the overclocking / general hardware forums. This is not highly technical.

It could quite easily get technical, as I am not the owner of a P4 and do not intend to overclock one. I am interested in the way that the whole scheme is implemented and how it might be circumvented.

Thanks for you input anyhow..........

According to pm, the multiplier and HT capability are set by fuses on the chip that are blown. It isn't possible to change the multiplier / HT setting once it has been set at the fab.

 

[TerryMathews]

Originally posted by: CTho9305
According to pm, the multiplier and HT capability are set by fuses on the chip that are blown. It isn't possible to change the multiplier / HT setting once it has been set at the fab.

Yes, that is what has been the general knowledge since the P3 days. That being said, it is possible that the P4 is different. To me, microcode on a flash chip makes far more sense for several reasons not the least of which they (Intel) could encrypt the contents of the microcode so that if they run into a group of people remarking their processors, they (Intel) can shut the remarkers down via the DMCA.


Hmmm...

 

[pm]

They are internally fused using one-time programmable electronically blown fuses. Intel has a couple of academic papers on the fusing technology that is used.

As far as encrypted microcode, this scheme would definitely not be ideal due to added cost and complexity. There's no need to go after remarkers using the relatively untested DCMA when remarking is a fraudulent activity currently covered under the existing criminal code. Besides the fusing method makes remarking very difficult which is an even better deterent IMO.

 

[AbsolutDealage]

It could quite easily get technical, as I am not the owner of a P4 and do not intend to overclock one. I am interested in the way that the whole scheme is implemented and how it might be circumvented.

Thanks for you input anyhow..........

OK... post where you like. I was merely trying to suggest that this would have more traffic and more people with personal experience in the overclocking forum.