Are there any other programs than RM Clock that can undervolt P8400 CPU using Vista?
The thing is RM Clock won't go beneath 1.0375V, I'd like to see if other programs do.
I am aware of NHC but don't know if it will work.
-
You can check out Crystal CPUID. (Its good for ES CPUs, but if your CPU is retail, I guess only the unlocked/available VID range will be accessible)
-
I might try a RMClock alternative on my T8300...RMClock settings keep getting reset for me =/
is Crystal CPUID the only alternative?
Messing with the menus, how to undervolt seems pretty self explanatory with CPUID -
You can try doing the registry tweak to unlock the voltages but it probably wont work. I believe the cpu itself blocks any requests of voltages lower than default.
NHC wont work either.
Only if you could drop the idle voltages from .988v down to .700v like on Pentium M's. -
How did you get yours to go down to .900 though... mine stops at .925 -
-
How did you get your CPU to go down to 0.9V??
Mine allows 1.0125V to 1.2V only?!?
I have it in a X200 Notebook. Nontheless RMClock detects and handles it as aDesktop CPU. When I switch to Mobile it allows me 0.9V to 1.0875V, BUT if you check the Voltage with CPU-Z for example, he only sets the values from above again according to which step you choose.
Can you confirm with another software that you actually run on 0.9V?
Greetings Xaverin
P.S.: It also bugs me that I loose 133 MHZ or half a multiplier, but I did not find any chance to solve that one as well -
Mine allows the lowest too(0.9v), I'm sure it could go lower, theres just no setting on RM to allow it.
Its down to the processor, it's the exact one in your laptop that determines how low a voltage you can go. The new lower wattage P series are good at undervolting though. -
Have you actually checked the Values with CPU Z or do you just look with RM Clock? Because if I set to Mobile I can go to 0.9V without a problem, RMClock also claims to be there, but in fact he stays on 1,0125 (according to CPU-Z).
The Processor Specs on Wikipedia also told me that P8400 should go from 1V to 1,25V...
Is there a way to realy make sure which Voltage is set (except using a Multimeter)? -
-
I think most likely it is not possible by design to go lower than 1V. Still it would be nicht to see a support of halt multipliers.
greetings Xaverin -
I have a P8400 here and on it's lowest multiplier (6x) RMClock can't go below 1.0375.
Crystal CPUID does give lower options, but it does not seem to work.
Any ideas? Any new programs available? -
John Ratsey Moderately inquisitive Super Moderator
1V is high for the SLFM voltage. It should be in the range 0.75V to 0.95V. Did you select Mobile on RMClock's advanced CPU settings page?
John -
Now I am down to 0.9250V.
I'd like to go lower though. Any options? -
John Ratsey Moderately inquisitive Super Moderator
I think it is actually leakage current, rather than voltage, which makes the difference in power consumption when you are down at that voltage (which is the same as mine). One of my P8600's cores is 5C cooler than the other when on idle and 9C cooler at full load (undervolted to 1.05V).
John -
With undervolting: Core 1 idles around 28 C, Core 2 at 35 C
How does that compare to your P8600 John?
I'd really like to limit the fan noise as much as possible, but pin-modding is a step too far. -
John Ratsey Moderately inquisitive Super Moderator
John
PS: 20 minutes later, fan off and core temps 31 and 35C. -
Weird. What is the lowest voltage shown by Crystal CPUID ? Works perfect for my T5200, but provides the same options as for the voltages as RMClock.
Try running the multiplier management side-by-side with CPUz and note any changes when setting voltages in CCPUID. -
I'm trying that out now and when I set RM Clock at 0.925V, CPUZ says the Core VID is 1.038V.
any idea why that is? -
Use CPUz v1.41....
And I guess you already have done the useless "unlock voltages" hack in RMClock ? -
John Ratsey Moderately inquisitive Super Moderator
Yes, the older CPU-z versions agree with RMClock voltages and CPU speed, but with HWmonitor, V1.11 agrees with RMClock temperatures and the older V1.10 has core temperatures that are 5C higher.
John -
-
Ok, you know about it ?
Anyway, go into Regedit > HKEY_CURRENT_USER > Software > RightMark > RMClock > Look for "UnlockVid", if not there, create a 32-bit DWORD value with that name and assign the valuedata as 1. [0x00000001 (1)] -
Ok thanks, will try that.
I'm thinking maybe some Sony utility is preventing changes?
Right now RMclock says VID = 0.9250V.
While CPU Z and CPUID say current is 1.038V.
Who is right?
-
Even CPUz 1.41 says 1.038V ?
Uh, yeah the hack is useless. I get values upto 0.7125V, but none work below 0.9500V -
Yes CPUz 1.41 says 1.038V, while RMClock says 0.9250V.
CPUID (in intel speedstep > current) says 1.038V -
John Ratsey Moderately inquisitive Super Moderator
I use CPU-z 1.40.5. It has just told me 798MHz and 0.925V.
1.47 says 1598MHz and 1.038V. Temperature is is only sure indicator.
John -
I think it can't sense the 6x multiplier too on the P8x00 series
(IDA is not recognised)
-
Cpu Rightmark is not the problem it is RMCLOCK.
I have my T5300 undervolted to 0.950 volts on every multiplier.
But when I measure the voltage out of the PVM for the cpu, I get 1.063volts, which is exactly what cpu-z picks up.
I just thought I would let you know that cpu rightmark is what is reading improperly.
K-TRON -
John Ratsey Moderately inquisitive Super Moderator
John -
CPU Z 1.48 just got out: http://www.cpuid.com/cpuz.php
-
Thats intersting John. Maybe cpu-z and myself were reading the voltages from the wrong leads. I will have to check again, maybe tomorrow after class. I will have to also try out cpu-z 1.48. I have still been using 1.44.
K-TRON -
John Ratsey Moderately inquisitive Super Moderator
CPU-Z 1.48 still reports 1.6GHz @ 1.038V when my CPU is on idle (SLFM mode which RMClock thinks is 800MHz @ 0.925V), so no change from 1.47 in that respect. Old CPU-Z (eg 1.40.5) agrees with RMClock.
John -
And going by your battery life RMClock is right?
-
John Ratsey Moderately inquisitive Super Moderator
The better test is to run a stress test when running on battery and look at the power drain. I've done it before. You can try it.
John -
I have set 0.95v in RMclock, CPU-Z shows 1.063v wich i believe is correct, just like some desktops are overvolting the cpu a little just to keep stability, but i have not a P8x00 cpu.
-
I have an P8400 and when i run cpuz 1.4X it identifies the CPU allright, but when i run an orthos nothing changes multipliers volt and everything stays the same... Wierd... the lowest i can get in RMclock is 0.875. and that is the setting it comes with from the start. It's wierd... So is it the software thats not up to date to handle these CPU's, or is it the hardware that is locked!?
-
Sorry to raise this thread from the dead, but I figured I would both share my results and also explain my understanding in response to some previous questions.
I'm using CrystalCPUId, which has the nice benefit that it supports half multipliers unlike RMClock, using RMClock without hacks causes it to top out at 8x, losing some speed. I have undervolted my Penryn P8400 at its top multiplier of 8.5 down from the default 1.250 volts to 1.063 volts. Bottom multiplier of 6 is locked at its base of 1.038.
I'm pretty sure the weird results everyone was getting where they were settings crazy low voltages like .9 which then weren't working is because of the fact Intel locks the CPU minimum voltage to the voltage that the lowest multiplier defaults to.
There's supposed to be a lower voltage setting in SuperLFM (Super Low Frequency Mode) that is essentially a level below 6x, but I haven't seen any evidence of that. Even without running any CPU tweaking program, and just watching CPU-Z while the system idles on battery and is set to maximum power savings by Windows, the voltage never drops below 1.038 at a 6x multiplier. Running RMClock, it shows a SuperLFM multiplier, with the lowest voltage being the 6x multiplier of 1.038. If you reset RMClock into mobile CPU mode (it auto detects to desktop), it just messes up its conversion table of VIDs (which are actually numbers to the CPU and not voltages), and even though it shows a lowest of .925, this is actually 1.038, and the next step up it shows as .9375 is actually 1.050, etc. when displayed by CPU-Z. -
CPU-Z will not show the SuperFLM mode by default. I think you have to hit F5 for it to show (Its one of the F Keys).
-
Man now I need a program with both SuperLFM and half multipliers. Maybe I will look into the registry hacks for RMClock. -
Check out CPUgenie. It's not a free program, but if RMclock and CrystalCPUID are giving you headaches, give it a shot. It's got a VERY user-friendly interface, and it let me undervolt my T5800 to 1v, where RMclock only let me go down to about 1.05v.
There's a 30-day free trial out there. http://www.cpugenie.com/ -
dondadah88 Notebook Nobel Laureate
yeah i'm cheap so thats a problem. that program should be 10 dollars.
-
-
CPUGenie seems to work fine until it gives me BSOD after the test has passed. -
Anyone else have a T5550? My RMClock allows it to go down to 0.7125V!!
-
FWIW
My P8400 settable core voltage is 0.9250V to 1.2000V. The minimum voltage 0.925V is not limited to SLFM. The 8.5x multiplier appears to revert to 1.1375V with heavy cpu usage if set for a higher VID.
For some strange reason I don't seem to be able to select the 6.5x multiplier, all others are okay.
Some CPU-Z versions pre 1.52 read the cpu voltage incorrectly, version 1.52 seems to read correctly.
Hopefully there will be a new release of NHC, the expired Beta had SLFM, half multipliers, and seemed to do an all round good job of adjusting core voltage and selecting multipliers. -
John Ratsey Moderately inquisitive Super Moderator
John -
-
-
RM Clock alternatives for undervolting P8400/P8600?
Discussion in 'Hardware Components and Aftermarket Upgrades' started by Phil, Aug 23, 2008.