Undervolting - differences between RM Clock and PHC

Discuss about PHC here. For problems please use the PHC Help forum
Post Reply
Geralt
Posts: 6
Joined: Tue 14. Oct 2008, 15:15

Undervolting - differences between RM Clock and PHC

Post by Geralt » Sat 18. Oct 2008, 13:09

Hi,

there's one thing I'm wondering about: in RM Clock I can undervolt my Core 2 below 1.000V. But here I've read two or three times, that the VID 19 is the lowest possible value for my cpu, which is 1.000V iirc.
At least it's a higher value than the lowest possible value in RM Clock. Now I'm wondering why RM Clock can set lower values (and at least it also displays lower voltages in its plot) than PHC.

And then I suppose that in PHC I have to set higher VIDs than I can set with RM Clock, in RM Clock I have no voltage value higher than 1,000V (actually even one below, I think it's 0.9875).


Geralt.



P.S. If exact values are important I'll post them later, at the moment I don't have them.

the-fallen
Administrator
Posts: 346
Joined: Wed 9. Jul 2008, 19:57

Re: Undervolting - differences between RM Clock and PHC

Post by the-fallen » Sat 18. Oct 2008, 21:19

I am not sure that I understand you correctly.

But I can set up a Value of 0.95V as lowest Voltage to my Core2 T7200. Lower ones can be set but they are not reached because the CPU does not use them.

With PHC you can also set the lowest possible voltage (in my case 0.95V) to each speed step but the CPU may no be stable then in higher frequencies.

Geralt
Posts: 6
Joined: Tue 14. Oct 2008, 15:15

Re: Undervolting - differences between RM Clock and PHC

Post by Geralt » Sun 19. Oct 2008, 15:56

Maybe I'm wrong, but to be sure I need a way to find the voltage associated with a VID and which VID is currently set. Can I be sure that the set VID shown by PhC for my Core 2 is the real one?

the-fallen
Administrator
Posts: 346
Joined: Wed 9. Jul 2008, 19:57

Re: Undervolting - differences between RM Clock and PHC

Post by the-fallen » Sun 19. Oct 2008, 20:04

What do you mean with
"Can I be sure that the set VID shown by PhC for my Core 2 is the real one?" ?

Do you mean the VIDs shown in /sys/devices/system/cpu/cpu0/cpufreq/phc_vids ?
Yes that are the ones Linux reads from your CPU's registers.

You can verify if the VIDs you wrote to this file are really applied by using the tool read_msr.py downloadable here in the forum.

User avatar
DavidG
Posts: 180
Joined: Fri 18. Jul 2008, 11:25
Contact:

Re: Undervolting - differences between RM Clock and PHC

Post by DavidG » Sun 19. Oct 2008, 20:12

I noticed that the vid to voltage translation is not always the same for all processor lines, at least for AMD processors. It sounds logical that this does not apply to Intel processors either.
RM Clock seems to be very well informed, but at least for AMD processors it's quite hard to get all information on this topic. But I won't rule out that it may be the vid-to-voltage conversion of PHC that needs improvement.

the-fallen
Administrator
Posts: 346
Joined: Wed 9. Jul 2008, 19:57

Re: Undervolting - differences between RM Clock and PHC

Post by the-fallen » Mon 20. Oct 2008, 10:41

The PHC patches does not use any voltage expression. Because of the fact that the calculation of the voltages depends on the CPU we use the VIDs. That makes it more save because any calculation error could cause wrong values to be set to the CPU.

The tool PHCTool calculates Voltages from VIDs. I can not guarantee that those are always correct because -like DavidG already wrote- it is hart to get the neccessary data.

How ever there is no reason to use voltages. It is only more pleasant to the human mind because people know "voltage" from somewhere ..

The only thing that really matters is to find the lowest usable value, no matter what it looks like, don't you think so?


But I promise to keep the VID to Voltage calculation in future PHCTool releases even if they might be incorrect :)

Post Reply