1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott technology?

[ol' man]

1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott technology?

I acquired a Gigabyte motherboard today and it had vcores to 0.825v so I did a little testing like I usually do. Check them out.

The gigabyte GA-8IPE1000 is a better board for me. I owned two of the ABIT IS7's and both of them had very hot running southbridges. Essentially after a few days of using the IS7's my system became unstable and the mobo's are essentially burnt up. They will not run stable anymore without weird glitchies. Since a couple of the mosfets and southbridge got so hot they about burn you I am going to assume some to alot of the problem is there as also maybe small stresses with the memory.


My gigabyte southbridge is cool to the touch as alsoall the mosfets and the mobo lets me run my ram at 5:4 while my IS7 only allowed 3:2.

So I did some calculations OC'd and at default. This is what I found out.

Default

OC'd

So at intel specs they would be at 28w @ 2.4GHz @ 1.0v and 73w @ 3.2GHz. The higher wattage is calculated from AMPxVOLT which intel specs do not go by.

Suddenly the prescott prolems we have been hearing about seem to be simply BS. Prescott is supposed to run on 1.25v or less. Theoretically at the same speeds processors on different processes should run at lower watts/vcores. The tbredA was an exception

 

[FIZZ3]

Hey nice findings. The system is fully stable at those low voltage settings?

This would be good news if it were to apply to Prescott.

 

[ol' man]

This would be good news if it were to apply to Prescott.

It would be nice if it did. Who knows really? It was just amazing that my 2.4 does 2.4GHz on 1.04v. In essence that means that the 1.525v setting for it is way over kill. 1.25v would probably be more like it for this chip. When dealing with a CPU there is a very big difference between 1.5v and 1v. This is amazing. It makes me wonder why intel sets their defaults so high some times. I remember another guy had his 1.8a run at 1v too but that is a whole 400MHz slower than mine.

It all makes me wonder if intel is using stretched silicon for this last stepping. Or basically 5~20% germanium in the last northwoods.

 

[Mhypertext]

Sweet i backed my 2.60C down to 1.325 and its running fine

 

[Since87]

Re: 1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott technology?

The higher wattage is calculated from AMPxVOLT which intel specs do not go by.

I don't know what this statement means. Intel doesn't go by A * V = W ??????

Suddenly the prescott prolems we have been hearing about seem to be simply BS. Prescott is supposed to run on 1.25v or less. Theoretically at the same speeds processors on different processes should run at lower watts/vcores. The tbredA was an exception

Assume a 3.2 GHz Prescott runs at the same power consumption as a 2.4C at 3.2 GHz, and that the last value in your wattage calculations (85.6 Watts) was the power dissipation of both CPU's.

The 2.4C will draw the following current:
85.6 / 1.525 = 56.1 Amps

The Prescott will draw
85.6 / 1.25 = 68.5 Amps

Now look at this link from Intel describing changes that need to be made to a motherboards Vcore regulator. In particular the following line:

"Change the 500 nH at locations L2B1 and L3B1 from 30 A to 40 A"

The only inductors on the motherboard that would need ratings in the 30 to 40 Amp range would be the switching inductors for a 2 phase Vcore regulator. With a 2 phase Vcore regulator, the current draw of the CPU would be split between the two inductors fairly equally. So with the original inductor on that motherboard, (30 Amp rating) the maximum current the Vcore regulator could safely provide is 60 Amps.

Looking back at the currents calculated for the two CPU's, you can see that the motherboard as originally designed could provide enough current for 2.4C, but not for the Prescott.

Clearly there can be regulator issues even if the power consumption of the Prescott doesn't exceed that of the 2.4C.

 

[Darryl_D]

all i'm saying is

HUH????????????

 

[Fushyuguru]

Re: Re: 1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott technology?

Looking back at the currents calculated for the two CPU's, you can see that the motherboard as originally designed could provide enough current for 2.4C, but not for the Prescott.

Clearly there can be regulator issues even if the power consumption of the Prescott doesn't exceed that of the 2.4C.

Thats an excelent observation. Great digging.

Of course we can speculate all we want, but you cant avert the fact that if the hardware is being changed and prepped for a larger current draw for the vcore (if thats what it ultimately is), and you cant dodge P=iV. Then the power consumption of the new revision will be higher. If the consumption wasnt higher, and was much lower, then why the change to beef up the current ratings of the inductors?

We cant know for sure yet.

 

[Ralf Hutter]

Sweet i backed my 2.60C down to 1.325 and its running fine

What board are you using?

 

[Since87]

Re: Re: Re: 1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott technology?

Of course we can speculate all we want, but you cant avert the fact that if the hardware is being changed and prepped for a larger current draw for the vcore (if thats what it ultimately is), and you cant dodge P=iV. Then the power consumption of the new revision will be higher. If the consumption wasnt higher, and was much lower, then why the change to beef up the current ratings of the inductors?

Even if the power consumption of Prescott is lower, the change to the Vcore regulator may be necessary, because of the lower operating voltage of Prescott.

For example assume Prescott consumes 80 Watts compared to the P4C's 85.6 Watts.

The current draw for Prescott would then be:

80 / 1.25 = 64 Amps

This still exceeds the current rating the Vcore regulator was originally designed for.

 

[jdmcnudgent]

at 1 volt, is it stable at a full load, or at just idle??

 

[Jarlax]

/boggle

I would be worried that during heavy loads...ok, heavy game playing, something would go wrong from a lack of voltage?

Maybe I am just being overly worried here??

 

[Fushyuguru]

Re: Re: Re: Re: 1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott technology?

Even if the power consumption of Prescott is lower, the change to the Vcore regulator may be necessary, because of the lower operating voltage of Prescott.

For example assume Prescott consumes 80 Watts compared to the P4C's 85.6 Watts.

The current draw for Prescott would then be:

80 / 1.25 = 64 Amps

This still exceeds the current rating the Vcore regulator was originally designed for.

Maybe you read my post wrong. I was agreeing with you totally.

 

[Since87]

Re: Re: Re: Re: Re: 1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott technol

Maybe you read my post wrong. I was agreeing with you totally.

I probably did. I may have misinterpreted this part:

Then the power consumption of the new revision will be higher. If the consumption wasnt higher, and was much lower, then why the change to beef up the current ratings of the inductors?

My impression was that you were saying that, 'higher current draw equals higher power consumption'.

 

[gamefoo21]

*newbie q*where do you find that wattage calculator?

 

[Fushyuguru]

Re: Re: Re: Re: Re: Re: 1 VOLT WONDER. Northy 2.4C @ 2.4GHz @ 1v. @ 3.2GHz 1.4v. Early prescott tec

I probably did. I may have misinterpreted this part:



My impression was that you were saying that, 'higher current draw equals higher power consumption'.

Yeah I could have worded it better also. My mistake.

 

[Albuquerque]

Yeah, my 1.8A will do 1800 at 1.2v (the lowest my Albatron will go), and will do 2400 at 1.3v, and 2700 at stock 1.5v. Not like I'm lucky enough to have a "C" class chip, but I'm along the same lines as you. Runs great, runs cold, love it!