[To all newer AMD Turion owners]Strange power handling on hp dv5

I just can't wait! Do you see any benefits from having third PState? I'm a little sceptical about it, I mean I doubt it will reduce power consumption significally.

On what sort of processor do you testing it? You've got Turion? Am I right? I am curious how low can you go with Watts. I measure it with PowerTOP (with disable everything and including hints from it). I was thinking about buying ZM and I would like to know if there would be any difference. On ql-60 it was 16.5 W or so.

 

I got a ZM-80 and have three pstates by default. Using tpc scaler I get a fourth working pstate, so I have p0 at 2100 Mhz, p1 at 1600 Mhz, p2 at 1050 Mhz and p3 at 525 Mhz.
If you got a ql, then you have 2 pstates, and probably it would be nice to get 4 pstates

I'll put the linux version in the first page of the thread as soon as I write some documentation for it. Actually it is in a very initial stage and a bit unoptimized too, but it is working very well.

 

Yeah, I've got two pstates, and the third would be nice and fourth... it's just too much

 

Good news. A newer testing version is available in the first page of the thread. Sorry, for Linux only!

 

Another news: finally I had the chance to compile the program under Windows, so TurionPowerControl reaches version 0.20.
It is available in the first page for Linux, Windows x86 and Windows x64 versions.

 

Mmmm funky. I'll have to test this tomorrow - i'm on my DV9000 atm and will be leaving for two weeks leaving the DV5 at home, but i will install the new version on my DV5 tomorrow and see how it goes.

 

whooo~~~so great!!I can't wait to try the new version.I just wanna know how you choose the voltage when you set the p1 at 1600 Mhz?and could you please paste your .cfg here,it may help a lot

 

New version is already available at first page.
I will soon release a 0.21 version with a correction about -altvidslamtime switch. I managed to obtain a nicer downvolt (from 1.0875v to 1.050v) on my ZM-80 setting AltVidSlamTime to 5 or 6 (default is 2) but version 0.20a and 0.12 have a nasty bug when setting that register.

However, there's an included .cfg example file in the version 0.20a, but I will soon post here my testing .cfg file with fourth pstate (since I'm not using the notebook now )

edit: oh, voltage at 1600 Mhz is around 1.050v. Unfortunately I still don't understand why when I use DID=0 the processor needs a lot more voltage than when I use DID=1.

 

excellent,thax
another quetion:if turn on the scaler model,should i have to leave the cmd window on the desktop all the time?or,how can i keep the scaler model working without the cmd window on the desktop

 

At the moment, yes, you have to leave the window opened.
Actually I need to polish the scaler code, it's still very experimental, then I will find a trick to let the program run in background without creating any window. A shared solution for win and linux would be really good, but I think there's no such solution.

Anyway, this is my test.cfg configuration file with 4 pstates.
Pstate 0 @ 2100 Mhz 1.0875v
Pstate 1 @ 1600 Mhz 1.0750v
Pstate 2 @ 1050 Mhz 0.7750v
Pstate 3 @ 525 Mhz 0.750v

Code:

# TurionPowerControl configuration file

: PSTATESET pstate 0

vid 37
fid 13
did 0

: PSTATESET pstate 1

vid 38
fid 8
did 0

: PSTATESET pstate 2

vid 62
fid 13
did 1

: PSTATESET pstate 3

vid 64
fid 13
did 2

: GENERAL

psmax 3

: SCALER

samplingrate 75
uppolicy dynamic
downpolicy step
upperthreshold 60
lowerthreshold 40

You can notice that pstate 1 (1600 Mhz) has a quite high voltage.

Downpolicy uses a "stepped" behaviour, meaning that it will drop to a lower pstate when cpu usage goes below lowerthreshold.

Uppolicy is using a "dynamic" behaviour, it means that if cpu usage detected is above 80% (half way from upperthreshold 60% and 100%), the scaler will step up by 2 pstate instead of 1 step.

There's also a rocket policy, meaning that there are no intermediate pstates. If the processor goes above (or below) a threshold, it will immediately jump to fastest (or slowest) pstate.


I'm gonna update first page with version 0.21, fixing the -altvidslamtime bug

 

here is my .cfg

and i t make me crush all the time,i don't know why since i did't change anything except adding a new pstate.so sad
ps:i got a zm-86

 

But does the scaler with just 3 pstates crash too?

Maybe try to set pstate 1 (in your 4 pstates configuration) equal to pstate 0 and see if you still have troubles.
Also I suggest to use step policy for beginning and then go with dynamic or rocket policies if you like them more.
Also I can't be sure that the trick always works, it works on my system but unfortunately I have no access to other machines with similar configuration to test it.

 

It's been a long time I'll try new version soon and, I hope I would find time. Keep working. If I make any progress with gui I will let u know

 

Hi guys, I report some news.

First, I investigated more about the pstate6/7 behaviour.
I found and report the fact that I can undervolt my processor much more with pstate 6 anomaly on than with pstate 6 anomaly fixed.

Without touching pstate 6 (so with anomaly on), I can undervolt my processor pstate 0 to 1.0375v for core 0 and 1.050v for core 1, and it's running pretty stable for some days now.

Touching pstate 6 (so with anomaly fixed) I can't go below 1.075v (for both cores) because I experience sudden crashes. The most curious thing is that I can undervolt to 1.050v or even 1.0375v: the computer will work fine for some time, but then all of a sudden I get a crash.

I'm thinking more and more about the fact that pstates 6 and 7 serves as a sort of "trampoline" to pstate0, and they just allows the voltage to ramp up just a moment before frequency goes up too, and so allow better voltage stabilization.

BTW, it looks like they produce a bit warmer processor, but actually I can't say nothing about battery life.

 

Hello again.

Just updated TurionPowerControl to version 0.22.

Changelog:

- Fixed Dual Plane/Triple plane operational mode reporting
- Added some more interesting bit reporting (PSI_L enable/threshold,
clock ramp hysteresis)
- Added capability to control PSI_L bit enable/disable and threshold
- Added C1E state reporting
- Added C1E bit set enable/disable

Available, as usual, on first page.

 

C:\Users\dv6t\Documents\Downloads\Compressed\tpc-0.22\bin\Win x86>turionpowercon
trol.exe -l
Turion Power States Optimization and Control - by blackshard - v0.22
Detected CPU:
Family: 0xf Model: 0x3 Stepping: 0x1
Extended Family: 0x11 Extended Model: 0x3
Package Type: 0x2 BrandId: 0x570
Detected Physical Cores: 2
Detected processor: Turion Ultra ZM processor
Processor has 2 cores
Processor has 8 p-states

Power States table:
-- Core 0
core 0 pstate 0 - En:1 VID:34 FID:16 DID:0 Freq:2400 VCore: 1.1250
core 0 pstate 1 - En:1 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 0 pstate 2 - En:1 VID:60 FID:16 DID:2 Freq:600 VCore: 0.8000
core 0 pstate 3 - En:0 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 0 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 6 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
core 0 pstate 7 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
-- Core 1
core 1 pstate 0 - En:1 VID:34 FID:16 DID:0 Freq:2400 VCore: 1.1250
core 1 pstate 1 - En:1 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 1 pstate 2 - En:1 VID:60 FID:16 DID:2 Freq:600 VCore: 0.8000
core 1 pstate 3 - En:0 VID:48 FID:16 DID:1 Freq:1200 VCore: 0.9500
core 1 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 6 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
core 1 pstate 7 - En:0 VID:28 FID:16 DID:1 Freq:1200 VCore: 1.2000
Processor Maximum PState: 2
Processor Startup PState: 3

Minimum allowed VID: 64 (0.750v) - Maximum allowed VID 28 (1.200v)
Processor AltVID: 64 (0.750v)
Processor Northbridge VID: 52 (0.900v)

SMAF7 is disabled; processor is using LMM Configuration
Registers for Power Management
Core 0 C1E on dual core multi-processing halt bit is disabled
Core 1 C1E on dual core multi-processing halt bit is disabled
DID to apply when in C1E state: 0

Voltage Regulator Slamming time register: 5
Voltage Regulator AltVID Slamming time register: 2
Voltage Regulator Step Up Ramp Time: 0
Voltage Regulator Step Down Ramp Time: 0
Processor is operating in Triple Plane mode
Processor PState Identifier: 0x18
PSI_L bit enabled (improve VRM efficiency in low power)
PSI voltage threshold VID: 60 (0.800v)
Clock ramp hysteresis register: 2 (960 ns)


C:\Users\dv6t\Documents\Downloads\Compressed\tpc-0.22\bin\Win x86>

 

Ok, need help now. I've managed to post the -l of my turion x2 ultra zm, but i don't know what to further type to fix my processor. My temps are around 90-99 after laptop being run for constant 8 hours.

 

If your processor temperatures are so high, you should first check the amount of dust in your heatspreader. Cleaning exhaust can let your notebook being much more fresh. Also I found in my dv5 that there was a severe problem with thermal grease on the processor-heatspreader assembly and also a venting problem: there is not enough fresh air stream for the fan, this is a engineering fault IMHO.

You may check this trick I've done to my dv5 for a better heat management:
http://forum.notebookreview.com/hp-compaq-voodoo-pc/486755-dv5-overheating-little-nice-trick.html

Then you can try to downvolt your processor using TurionPowerControl.

TurionPowerControl has a lot of options, but you may be interested in -pallc switch.

For example if you write:

TurionPowerControl -pallc 0 36 0 16

you're telling your processor to change pstate 0 and use VID 36, DID 0 and FID 16.
VID 36 corresponds to 1.100v (VID 34 is 1.1250v, your processor default).
DID 0 and FID 16 corresponds to a frequency of 2400 Mhz.

So issuing such a command you're going to let your processor use less energy (since you are reducing voltage) for the same frequency.

Check chapter 3 in readme file for some very easy formulas to obtain actual voltage from VID values and actual frequency from DID and FID combinations.

If you raise VID value, you're going to give less and less energy to the processor, so it will produce less heat. Therefore giving less energy may also cause instability. The trick is to find the highest VID value (so the lowest voltage) that give you no system instability and this can be done just in a try and error manner.
For example, I have been able to reach VID 40 for my processor (that's 1.050v) with no stability issues, but all processors behave differently.

I hope I have been clear enough, but if you have more questions just write here

 

to blackshard83:

man you saved my note
here (in Brazil) ambient temp. goes over 30°C and my note (an tx2-1040br w/ an turion rm-72 2200mhz) got over 105°C! At first I though it was gonna fry! then get used to, then found your forum read about the problem, and the SOLUTION!

YOU ARE AN NOTE SAVER!

btw, people in the forum are with trouble to get the command window hidden, try CMDOW, its an simple and small Program that hid CMD windows.

now that i'm down to 95°C in full throttle I'M Happy

 

Glad to hear that you found my program useful

And thanks for cmdow program, I think it will be useful to me too!

 

Hello guys, new update!

TurionPowerControl reaches version 0.29 pre-alpha with new experimental support for Family 10h processors. It means that now are (partially) supported also desktop Phenom, Phenom II and Athlon II processors, and also mobile Athlon II, Turion II and Phenom II processors.

Anyway I need feedback for proper bug discovery, since I had a limited time to work on new supported processors.

You can grab it here.

TurionPowerControl v0.29 pre-alpha

 

New updated version 0.29.1a alpha.

Changelog:

- Added a warning message at startup if program can't access CPUID or MSR registers that asks for administrative privileges and, for linux users, cpuid and msr modules.
- Added the -set switch, which is really useful and much more user friendly than direct FID/DID/VID manipulation. Check the documentation and examples for usage.
- Fixed detection of correct PVI/SVI mode for Family 10h processors
- Updated documentation with all the program switches
- Corrected some minor issues with K10 processors regarding altvidslamtime (unsupported on K10) and psithreshold switches
- Added requirements paragraphs in documentation
- Added the frequency and voltage items in the configuration file definition. They can be used instead of FID, DID and VID values for easier pstate manipulation. (See the included example.cfg file on how to use them).
- Corrected a detection error with Turion II/Athlon II mobile processors.
- Removed AltVIDSlamTime reporting for Family 10h processors (unsupported on them).
- Added CPU Maximum Frequency reporting


Available here:

TurionPowerControl 0.29.1a

 

Does TPC support AMD Turion TL-58? Because whenver I excute 'TurionPowerControl' then I just got error message 'TurionPowerControl: command not found'.
I followed the tutorial in the Doc dicrectory.

 

New version 0.30 available.

You can download it from here: TurionPowerControl v0.30

Changelog:

- Corrected an issue with minimum voltage VID with Parallel VID
implementations.
- Corrected an issue with -set voltage switch reporting wrong rounding
when used on a specific core instead of all cores. The effect of the
command was therefore successful.
- Corrected a problem with cpu usage reporting with K10 processors with
more than 2 cores.
- Removed an annoying debug string when using -set frequency with
K10 processors.
- Added support for northbridge voltage and divisor change for K10
processors. Northbridge voltage can be changed with -nbvid switch or
using -set switch. Divisor can be changed with -nbdid switch
- Addes support for northbridge voltage change in configuration file
- Added some more usage examples, fixed some documentation issues

 

Hi people and blackshard83, I´m new in the forum, I was googling about a problem with my DV5-1114ea... and I finded this thread, and I´m impressed with your app!
great work, I downladed it to try as soon as possible, but... I want to ask you about what you think if is possible that this rare issue with the pstates could damage my turion ZM82 processor?... The machine started to have random shutdowns but suddenly now... when I press the power button, works for about 45 seconds... then shutdown. Press again and works for 15 seconds and shutdown... press again and instantly shutdown... so, I leave it off for 15 minuts, and the sequence is repeated when I turn on again.
I cleaned the fan (anyway was very clean...), I replaced the thermal paste with Artic Silver 5, but nothing... I was able to verify that the Microprocessor raise his temperature very very fast... (in 10 seconds and I can not continue touching the heat sink)...
Now I don´t know if I have a problem in the Motherboard VRM area or I must to remplace the microprocessor...
I can not use your app, windows or anything because the more time that the machine work is 45 seconds.
what do you think? thanks you in advance.
PD: I forgot to say the GPU is a ATI HD... and apparently have not problems, no high temperature compared with the microprocessor.

 

Hi all, just reviving an old but useful thread.

Ive been using black shard's tool on my tx2 since I got it, but unfortunately many severe overheats seems to have compromised it's stability, probably due to the desoldering problem. Anyway. I've cooked it in it's own heat and have done the 1981 copper penny thing out of YouTube, and it's mostly been OK as long as it doesnt hit 90C otherwise I get the black screen and have to cook it again. Sigh.

Thing is its been hitting the 80s despite all the undervolting I've tried. Best case was 1.075 at 2.2ghz pstate 0. The other p states I set to most stable frequency at 0.8v and 0.75v. Worked mostly Ok, but those weird transitions appears to be causing the overheat. It seems the p state 6 transitions are coming more and more often, I don't know why. Seemingly less so when using custom scalers, but I lose the ability to throttle under battery power without user intervention.

Tried undervolting p state 0 to no avail, always crashing even if I under clock at the same time. But p state 1 and 2 have been Ok with aggressive under volt / over clock. So I tried setting the win 7 built in scaler and set it to 50% for both AC and battery and used the fastest stable clock at 0.8v (right now at 1600mhz but testing higher tomorrow) for p state 1, same with p state 2 at 0.75v. Now much less transitions to p state 6/7, so I don't see the 1.2v@1100mhz as often... Temp stable at low to mid 50s in either battery or AC. Battery life at roughly 4 hrs with an 8 cell (DIM display with wifi).

If this proves to be stable i may have found the perfect setup for my tx2!

Now to decide if I want to do SSD... G skill Phoenix pro at 120gb? Or go cheap with a Marvell based plextor? I just need Office, mainly outlook and one note, and a browser...

 

Hello guys, I'm resuming the thread to ask if someone has a brazos/llano processor and wishes to give a hand testing a newer version of tpc with support for those new platforms.

 

hi, Someone could help me how to configure?

AMD Turion X2 Ultra Dual-Core Mobile ZM-84
Ubuntu 11.04 64 bits


Main processor is Turion Ultra ZM Processor
Family: 0xf Model: 0x3 Stepping: 0x1
Extended Family: 0x11 Extended Model: 0x3
Package Type: 0x2 BrandId: 0x540
Machine has 1 nodes
Processor has 2 cores
Processor has 8 p-states

Power States table:
-- Node: 0 Core 0
core 0 pstate 0 - En:1 VID:34 FID:15 DID:0 Freq:2300 VCore: 1.1250
core 0 pstate 1 - En:1 VID:48 FID:15 DID:1 Freq:1150 VCore: 0.9500
core 0 pstate 2 - En:1 VID:60 FID:15 DID:2 Freq:575 VCore: 0.8000
core 0 pstate 3 - En:0 VID:48 FID:15 DID:1 Freq:1150 VCore: 0.9500
core 0 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 0 pstate 6 - En:0 VID:28 FID:15 DID:1 Freq:1150 VCore: 1.2000
core 0 pstate 7 - En:0 VID:28 FID:15 DID:1 Freq:1150 VCore: 1.2000
-- Node: 0 Core 1
core 1 pstate 0 - En:1 VID:34 FID:15 DID:0 Freq:2300 VCore: 1.1250
core 1 pstate 1 - En:1 VID:48 FID:15 DID:1 Freq:1150 VCore: 0.9500
core 1 pstate 2 - En:1 VID:60 FID:15 DID:2 Freq:575 VCore: 0.8000
core 1 pstate 3 - En:0 VID:48 FID:15 DID:1 Freq:1150 VCore: 0.9500
core 1 pstate 4 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 5 - En:0 VID:0 FID:0 DID:0 Freq:800 VCore: 1.5500
core 1 pstate 6 - En:0 VID:28 FID:15 DID:1 Freq:1150 VCore: 1.2000
core 1 pstate 7 - En:0 VID:28 FID:15 DID:1 Freq:1150 VCore: 1.2000

--- Node 0:
Processor Maximum PState: 2
Processor Startup PState: 3
Processor Maximum Operating Frequency: 2300 MHz

Minimum allowed VID: 64 (0.750v) - Maximum allowed VID 28 (1.200v)
Processor AltVID: 64 (0.750v)

 

The following command should work and be safe enough:

Code:

> TurionPowerControl -set core all pstate 0 vcore 1.100 pstate 1 vcore 0.825 pstate 2 vcore 0.785

You may wish to reduce the vcore of pstate 0 to 1.0825 or 1.075 and see if it is still stable.

 

Hi guys, sorry for replying to old thread. My laptop is stuck running at 0.55 ghz, I cannot get it to run higher even though its a 2.3ghz. Windows 7 and 8 downlock it, where as a live Cd or safe mode it will run at 2.3ghz!

this is my output. Is there anyway to get the cpu working as it should and in a way that survives a reboot?


TurionPowerControl 0.44-rc2 (tpc-0.44-rc2-r144)
Turion Power States Optimization and Control - by blackshard

Main processor is Turion X2 RM Processor
Family: 0xf Model: 0x3 Stepping: 0x1
Extended Family: 0x11 Extended Model: 0x3
Package Type: 0x2 BrandId: 0xcd0
Machine has 1 nodes
Processor has 2 cores
Processor has 8 p-states
Processor has 0 boost states

Power States table:
-- Node: 0 Core 0
core 0 pstate 0 (p0) - En:1 VID:36 FID:15 DID:0.00 Freq:2300 VCore:1.1000
core 0 pstate 1 (p1) - En:1 VID:48 FID:15 DID:1.00 Freq:1150 VCore:0.9500
core 0 pstate 2 (p2) - En:1 VID:60 FID:15 DID:2.00 Freq:575 VCore:0.8000
core 0 pstate 3 (p3) - En:0 VID:48 FID:15 DID:1.00 Freq:1150 VCore:0.9500
core 0 pstate 4 (p4) - En:0 VID:0 FID:0 DID:0.00 Freq:800 VCore:1.5500
core 0 pstate 5 (p5) - En:0 VID:0 FID:0 DID:0.00 Freq:800 VCore:1.5500
core 0 pstate 6 (p6) - En:0 VID:28 FID:15 DID:1.00 Freq:1150 VCore:1.2000
core 0 pstate 7 (p7) - En:0 VID:28 FID:15 DID:1.00 Freq:1150 VCore:1.2000
-- Node: 0 Core 1
core 1 pstate 0 (p0) - En:1 VID:36 FID:15 DID:0.00 Freq:2300 VCore:1.1000
core 1 pstate 1 (p1) - En:1 VID:48 FID:15 DID:1.00 Freq:1150 VCore:0.9500
core 1 pstate 2 (p2) - En:1 VID:60 FID:15 DID:2.00 Freq:575 VCore:0.8000
core 1 pstate 3 (p3) - En:0 VID:48 FID:15 DID:1.00 Freq:1150 VCore:0.9500
core 1 pstate 4 (p4) - En:0 VID:0 FID:0 DID:0.00 Freq:800 VCore:1.5500
core 1 pstate 5 (p5) - En:0 VID:0 FID:0 DID:0.00 Freq:800 VCore:1.5500
core 1 pstate 6 (p6) - En:0 VID:28 FID:15 DID:1.00 Freq:1150 VCore:1.2000
core 1 pstate 7 (p7) - En:0 VID:28 FID:15 DID:1.00 Freq:1150 VCore:1.2000

--- Node 0:
Processor Maximum PState: 2
Processor Startup PState: 3
Processor Maximum Operating Frequency: 2300 MHz

Minimum allowed VID: 64 (0.7500V) - Maximum allowed VID 28 (1.2000V)
Processor AltVID: 64 (0.7500V)

Done.