The ThrottleStop Guide

It doesn't matter if its checked on or unchecked, as long as it's grayed out, that means you can't enable dual IDA.

Just checked, the TJmax for T9300 should be 105 C, check the "hwmonitor.ini" file, which should be right along with the hwmonitor.exe. Open the "ini" file and make sure the TJMAX value is 105, not 100.(close the hwmonitor.exe before edit, or it won't save).

Edit: Did you press the "turn on" button on the interface of throttlestop? You won't notice the voltage change if you won't let TS take over control of the CPU. Also make sure you have the latest CPU-Z.

 

Yes, there is a way to fix this. First go to the Intel site and find out what the correct TJMax value is for your T9300.

Intel® Core?2 Duo Processor T9300 (6M Cache, 2.50 GHz, 800 MHz FSB) with SPEC Code(s) SLAPV, SLAQG, SLAYY, SLAZB

Intel says 105 so you need to edit the hwmonitorw.INI configuration file and set that to the proper number.

CPU_0_TJMAX=105.0

Now ThrottleStop is all fixed.

Edit: You beat me to it lidowxx.

I've never had a request for more than 4 profiles. Maybe someday if a few more requests come in. Why do you need more than 4?

If the EIST box is grayed out then this is locked so Dual IDA won't work. You can unlock EIST in many of the Dell laptops by editing the bios.

If you want to use a SLFM voltage then you need to enable the SLFM option. In RM Clock you need to enable the DFFS, Dynamic FSB Frequency Switching, option to access the lowest possible SLFM voltages.

You can run a program like HWiNFO32 and it will give you a nice little table showing you what votlages your CPU supports for SLFM, LFM, HFM and IDA.

Post a screen shot or two if you are having any problems.

 

Ok I'll have to Change that when I get home on my laptop!

More than 4 just so I can have more options to be able to set to, like different multipliers. Since RMclock did more than just one at a time, I'd like to have a few more multiplier settings to be able to easily switch from
but not that big a deal lol

I'll check those other things asap and post back.

 

ThrottleStop 2.90 works great! the old version interfered with my wireless connection for some reason...this one works awesome! My pc- Alienware M15X CoreI7 720q. Before using this using this program, my pc was to slow to use when not plugged into A/C....I was only getting a max multiplier of 6 on battery...such a pain. You brought back mobility to my PC....Thanks!

 

The M15x is not adequately designed so that you can run at full speed while on battery power. You should phone Dell and tell them this "feature" was never mentioned when you bought your M15x. That's usually why Dell throttles your laptop to a crawl when not plugged in. If you use ThrottleStop to try and force full speed while on battery power, it is quite possible that you will damage your battery. It probably won't catch on fire but it might. It might also significantly reduce the life of your battery.

All I can say is be careful. I don't recommend using ThrottleStop while on battery power.

The minimum multiplier for a Core i7-720QM is 7.0. If CPU-Z is showing less than that, it must be combining the chipset clock modulation data with the multiplier data and creating some sort of effective multiplier.

Someone else once reported that they were having wireless connection issues when using ThrottleStop while gaming. They switched from a Dell wireless card to an Intel card and their problem went away. Just curious. Do you know what brand of WiFi card is in your laptop?

 

If you try to force you laptop to full speed on battery your machine will simply switch off. I have tried this

 

Wireless - Intel(R) WiFi Link 5300 ANG

Something that is strange is that when I open your ThrottleStop.exe file and run...I do not have to turn it on (bottom right). As soon as I open ThrottleStop it changes my multiplier from (5-6) to (20-21) without turning on. So I just open and minimize without turning it on...Or is that how the program is designed...soon as I open does this turn it on anyway. If that made any sense.
edit: something i just noticed...when I open ThrottleStop and it changes my multiplier...I can close the program and it still is keeping my multiplier at 20-21. Is there a reason it would do this?

Not sure if dell will do much about it...I have had it a year. Had to do it over...would not have bought an Alienware. My thought was to upgrade to the i7 920 when prices went down a year ago and upgrade myself...but so far they haven't.

 

My original theory was that if a person is taking the time to run ThrottleStop, they are probably doing so because their laptop is running like a slug. Clock modulation and chipset clock modulation are immediately turned off since Dell shouldn't be using these methods to slow down a users computer. It also sets the multiplier to its maximum so even if ThrottleStop is only in monitoring mode, at least the operating system can re-gain control of the multiplier when ThrottleStop exits.

I've heard from more users that are happy with this decision than users that are not happy so I've left it as is so far. I might change this in the future if enough complaints come in.

What software are you using to report the multiplier before and after? A 5 or 6 multiplier physically does not exist in your CPU. That's why it sounds like ThrottleStop is mostly turning off the chipset clock modulation that plagues the M15x.

At idle, the 720QM automatically drops the multiplier significantly lower than 20 or 21 but CPU-Z does not report this. Here is what the programmer of CPU-Z says about that.

Try downloading RealTemp 3.60

Download Real Temp 3.60 | techPowerUp

Included with that download is another program called i7 Turbo GT which accurately tracks the average multiplier of each thread. This program turned into ThrottleStop so the multiplier data should be similar without having to run ThrottleStop.

By running i7 Turbo GT, you will have a better look at what the multiplier is up to before and after running ThrottleStop. At idle, you shouldn't be seeing an average 20 or 21 multiplier on a 720QM.

 

I ran RTemp and attached my log...added notes when I made a change. I am not sure what the numbers mean.

I also attached a screen shot of a process that keeps running...it will at times take a great deal of percentage of CPU while at idle. Have ran checks for virus' and came up with nothing.

 

Blue Screen for me





Will ThrottleStop work on an Atom CPU?

Edit: just tried it and found my answer
Any Atom support coming in the future?

 

Nuttz220: The log file you ran shows that the average multiplier your CPU is using during your tests is hardly changing from one test to the next. The PEAK column shows the peak average multiplier. It wanders around mostly between 11 and 14 which is completely normal for a Core i7-720QM at idle.

As mentioned before, a 5X multiplier does not exist in the i7-720QM. When CPU-Z reports a multiplier lower than 7.0, that is usually a sign that it is combining the clock modulation throttling with the actual multiplier and creating some sort of equivalent multiplier. I guess it is telling you that your CPU is running similar to a mythical CPU with a 5X multiplier. When it shoots up to 20X, it's then telling you that your CPU is using a sky high multiplier when it really isn't. It does this for more consistent CPU-Z Validation purposes.

Anyhow, starting up ThrottleStop immediately turns off the two types of clock modulation. That causes CPU-Z to switch from under reporting the actual multiplier to over reporting the actual multiplier and is why you see a huge difference. The actual multiplier has hardly changed at all. Just the clock modulation throttling has been disabled. You will also find that ThrottleStop does not show any 20X multiplier when the CPU is idle because this is not really happening. i7 Turbo GT and ThrottleStop and RealTemp and Core Temp all follow the multiplier monitoring method recommended by Intel in their November 2008 Turbo White Paper. The programmer of CPU-Z has developed his own method.

As for that rogue file, here's 101 opinions about whether it is good or bad or a trojan, etc.

wmiprvse.exe Windows process - What is it?

My Windows 7 laptop is running fine without this process running.

[>>ahl395<<]: For some reason, Intel disabled the high performance timers in the Atom CPUs that ThrottleStop depends on. That's why TS does not support the Atom. I might add limited Atom support some time in the future. Intel has finally added these high performance timers into the latest Atom CPUs that were recently released.

 

Thanks for looking at the log and clearing up how this thing actually works. I will probably get rid of CPU-Z and check out that running process as you mention.

 

CPU-Z is useful in that it does confirm that something is going on when clock modulation is being used. It just doesn't accurately tell you exactly what it is that is going on. When it starts displaying multipliers that are less than what is physically possible (less than 7 in the Core i7-720QM), you know there's a problem somewhere and in a Dell laptop, it's usually one type of clock modulation or the other.

With ThrottleStop running in monitoring mode, plug your laptop in and then unplug it and switch from AC to battery power and see if the clock modulation is reported in the monitoring area. It probably will be. My older Dell laptop does this but the clock modulation only lasts for about 15 seconds before it is properly restored to 100%. I assume this was a Dell idea to prevent excessive load on the battery when you quickly switch between AC and DC. The only problem with your laptop is it sounds like they forgot to turn this "feature" off so your laptop is always running like a slug on battery power.

 

Ah okay, that sucks lol. It would be helpful to set it at full power all the time, seeing as its pretty low power itself lol

 

ThrottleStop 2.90 beta 4
http://www.mediafire.com/?wyqfkhqg9tx2u8t

I just added a new benchmarking feature to ThrottleStop so it will be easier to get some quick feedback whether your adjustments are improving performance or not.



Edit: Minor bug fix. Updated to beta 4.

 

Would it be possible to make it so ThrottleStop keeps the best benchmarking results? As at the moment it resets if you exit ThrottleStop

 

That's a good idea JKleiss. I'll look into adding that feature in the near future.

 

@unclewebb
Thank you for your work join me.I run "ThrottleStop" on the HDX Dragon with T9300.
Details of the specific way of working on my system - here.Sorry, my English is very bad, so here are just a short thanks!
(Thanks also to Nando4, because without Bios Modd none 2.7 GHz)

Question: Is it possible to integrate a voltage measurement in ThrottleStop?
At the start of "ThrottleStop" Although everything is adjusted properly, but the voltage is occasionally over my attitude.
Sure there are other tools for it, but would be an autorun so impractical.

Example: Dual IDA on Start "on" Default Profiles "AC Profile 1" - "on"
Set Multiplier "13.5", Voltage ID 1.0250 (stable!)

Behavior at the start: about ten times ThrottleStop starts with 1.0250 V.
Then, however, is blocked plözlich to 1.2125 V, or 1.1000 volts.
The Multplier 13.5 but correct.
This is due not to a "ThrottleStop" but you see there just is not there.
(Not a voltmeter)

PS: 1.2125 V original voltage IDA. attempts to work with higher voltage (example 1.1000 V) did not help.

 

Is there an alternative / plans to implement this for AMD processors?

 

RickiBerlin: The CPU needs to have EIST enabled when you change VID values. If EIST is disabled, FID and VID changes do not work. I think that is the problem you are having. I might add a VID display some day.

You also need to enable SLFM if you want to access the lowest possible SLFM VID voltages. Post some screen shots if you need some more help.

jerg: I don't own any AMD hardware so unfortunately I don't plan on supporting AMD.

 

Aw ... Could you suggest any alternative software that could show AMD clocks/throttling? I've been crawling all over NBR today trying to find out if my CPU is getting throttled in games ;_;

 

The only AMD program I know about is K10stat.

 

ThrottleStop 2.90 beta 5
http://www.mediafire.com/?bnz7r2oh0odaj00



JKleiss: added the ability to save the benchmark results from one session to the next. Also added a checksum feature to keep everyone honest and added a Reset button to the TS Bench module so you can clear all of your times and start fresh.

RickiBerlin: added VID voltage reporting for Core 2 CPUs. In the above example, the VID request is set to 1.0500 but because Power Saver and SLFM are enabled, the CPU is using its lowest SLFM VID voltage which is 0.9500. That is now reported in the monitoring area.

ilguercio: added a new INI file option to access a lower CPU speed on the Core 2 CPUs with 200 MHz bus speeds.

SuperSLFM=1

On these CPUs, SLFM mode drops the bus speed from 200 MHz to 100 MHz at idle and then increases the multiplier to 8.0 for a total of

8.0 x 100 MHz = 800 MHz

If you have a Core 2 CPU like a T7250 or T8100, adding the above option to the ThrottleStop.INI configuration file will drop your CPU speed down to this at idle

6.0 x 100 MHz = 600 MHz

This speed is not officially supported by Intel but RM Clock lets you do this trick so now, ThrottleStop lets you do this too.

I also added a few options for myself.

The Battery button has been upgraded with some more information. You can now click on the Batt button and toggle the info from:

1) Battery Percent remaining.

2) Estimated Battery run time when on battery power.

3) System Power Consumption in watts when on battery power and charging rate in watts when plugged in.

4) and finally Maximum Battery Capacity in Wh, aka. watt hours.

If your battery is rated at 50 Wh and draws 20 W when you are using it then you should get about 50 / 20 = 2.5 hours (2:30) of run time out of a full charge.

Almost forgot. Added a DC Exit Time option to the Options window for Nando4. You can use this option to force ThrottleStop to close after a pre-selected amount of time in seconds. Some people don't like running ThrottleStop on battery power so if you set this value to 10, ThrottleStop will automatically exit approximately 10 seconds after you switch from DC to AC. Minimum value is 1.

Lots of little changes so if you see any problems, let me know.

 

Sorry, I know exactly what is to be set. Only my English is poor.
Try to explain it again ...

Start ThrottleStop "with the following settings (always 100% identical conditions) [EIST, SLFM" on "etc] -see this:




And now, behold, the voltage in CPU-Z ! (HWInfo32 shows the same value):


The voltage is not correct. Stand firm on the maximum value. Without my control it would be a high CPU load for a long time.

This happens only every 10th start of "ThrottleStop.
In general, the start is with the pre-making of me voltage of 1.0250 V.
sse this:


Attention: I do not need help! Yes, all combinations have been tested.
(V, 1,100 or others, "BD PROCHOT" on and off, and others)
Believe me, tried everything!

Would like to point out. And thus the need for a voltmeter in "ThrottleStop.

For reconsideration Info: If you do this---> <--- klick!, have not read, here is an excerpt:

BD PROCHOT raises the HDX, the CPU limitation on Battery.
Thus, therefore, to 2.7 GHz on battery possible. (Previously max 1.2 GHz T9300)

Thank you with your work gives me a lot of fun. :yes:

 

The VID Voltage on the left side of ThrottleStop is a VID request. This only works 100% reliably when you have EIST enabled. You MUST HAVE EIST enabled for this to work correctly or the CPU might ignore the value you set.

I've already added VID monitoring to ThrottleStop 2.90 beta 5 just for you RickiBerlin.
Thanks for the good idea.



ThrottleStop 2.90 beta 5
http://www.mediafire.com/?bnz7r2oh0odaj00

Some of the problems you are having have already been fixed so please use the newest version when testing. Thanks for your feedback.

 

ThrottleStop 2.90 beta 6
http://www.mediafire.com/?jdal8a26utytlum

A couple of very minor updates. Moved the new real time VID data over beside where you set the VID and increased the maximum PowerSaver C0% to 100% in the Options window so you can run at full load with SLFM enabled if you want. I won't ask why you would want to do that.

 

Sorry, may I thank you later?
"2.90 Beta 5" almost perfect. Please read my test!
I'll look like once the beta sixth
Man you're fast ...

EDIT: My mistake, this function has already 2.90.

 

Is everything OK now RickiBerlin? I added a little VID monitoring window for you to keep an eye on things. When EIST is disabled, software can not change the multiplier or VID. What ThrottleStop does now to try and get around this problem is that it turns on EIST for a few milliseconds when you change profiles so the FID and VID can be changed and then EIST is immediately turned off again. This seems to work OK but if you have any problems just let me know.

I want to make sure there are no major problems before finally releasing the official version of TS 2.90.

 

very nice

First of all, thank you for reporting VID voltage!
Unfortunately, it was really necessary, because after hours of testing out everything today announced that as "Beta 5 +6" occasionally start with the wrong voltage.
It's really very rare. Something you must have changed.
To me it seems as if the fault is in beta 6 is more common than in Beta 5.
Both versions with exactly the same settings.
Error occurs both at the start as well as the profile change.
Is very difficult to reproduce, as extremely rare. (1 error in 30 x ok)

This function for me without success. (Is not necessary)
yes Could CO% 100 and Power Saver "on."
But as I said only for info. (Yes I "SuperSLFM = 1" was inserted in ini)

 

Super SLFM should work on the T9300. Did you have PowerSaver and EIST enabled? Can you show me a screen shot of ThrottleStop at idle?

EIST should be on for all profiles except your Dual IDA profile. The CPU needs EIST on for FID and VID changes to work properly. I have not changed anything recently with any of the 2.90 beta versions but I will check to see if I can improve this even when you have EIST turned off.

 

yes

yes, but I do not do it. Explanation follows...

Have the explanation for the malfunction of SuperSLFM = 1 found.
As always, it is very simple. One must know:
If you have already finished their profiles (including SLFM on) and adds the command "SuperSLFM = 1" only following one, then it does not work.
Here one must SFLM off and back on again + saven profile.
Result 6x100MHz works now.

Vice versa, it is this: Get back to reach the normal value (8x100MHz), it is not enough to change the value in the ini or delete them.
Even without this ini entry (SuperSLFM = 1) the factor remains at 6x100MHz.
Not even no program restart!
Here too, of SFLM off and back on again + saven profile.
Result 8x100MHz works now.

I decided not to post the pictures already finished it. Even though I have this creates extra each charged with Ini's, it's not really meaningful.

But here is a little picture puzzle for you. No "ini" or other changes.
(No need for assistance)

That just looks good that screenshots can also confuse.
As you can see Power saver of, and the CPU runs at 6x100. The default is 13.5x200. EIST is on! You do not employ so wanted to show it only. Operator error on TS surface are not excluded. All depends on the constellation of settings.
(Will remove this later, he is no help)

Against a lock in low end is talking, but nothing. Or what?
The CPU throttling can not. (So yes ThrottleStop )
It would be important but only before the end of "ThrottleStop" to enable EIST. (Factory setting for Steping)

Can you (or better still, to me) to explain the value of HWiNFO32?
Please note that the four lines of the picture are not to be taken seriously!

Finally, I love your little program! It makes my CPU very large.

 

Oh my lordy

 

Sure, you've identified an ambiguity in this sentence.
Will not hope that it is the same which I think.
Good thing I do not know what I write and read it even less, in that language. See, I can not English. So the best position to chat something on this site.

 

Thanks for your testing RickiBerlin. I hate SLFM and Super SLFM so I never use these functions. My main desktop computer that I use for programming does not support these features so I did not test these options thoroughly. I will do some more testing today to make sure that it works correctly for users that like SLFM.

HWiNFO32 looks like a big bug but I don't know why. Was your real bus speed still 200 MHz when it went crazy? Like your ThrottleStop screen shot shows, all software does crazy stuff sometimes.

It is difficult to understand your German to English translation but I appreciate your help.

That's what I like to hear.

 

Damn you, dude, you made me spray a whole mouthful of water on my screen!!!:laugh:

 

What the funny thing about this set?
If I "CPU" not to abbreviate, otherwise something else does. (Something that can grow in male users)
Please clarify for me! I would also like to "spray a whole mouthful of water on my screen"

 

No problem
Super SLFM's also completely senseless. can save power so you do not.

Not necessary!
1. It works wonderfully.
2. One does not need.
3. If you would you take a bit more time for my text, then you knew Item 1 .(1= It works wonderfully.)
It is only a setting order problem. (Did it accurately described above)

And this is what I wanted to show you with the ThrottleStop image. Your Program includes operating faults. The picture does not do with "Super SLFM.
It is not a program error, but an unpredictable for you at the same time senseless Program setting.
Explanation of my approach to generate the image state.
1. SuperSLFM = 1 entry in ini.
2. ThrottleStop start it and make the following settings.
In options, "Default Profiles on AC Profile" and "Start Dual IDA" on.
Now we create the profile "Performance"!
Full Multiplier, power saver on, on SLFM (EIST does not matter) and save! (Yes, this setting is total nonsense, but it's all about control errors)
3. "Forget not turn on" then quit "ThrottleStop!
4. ThrottleStop start. Wait until the highest multiplier was captured.
5. Now you can switch from the context menu IDA. (Stop Dual IDA)
6. Now you can see my image of this post.
Reason: Start Dual IDA "terminated while" Power Saver ", but" Stop Dual IDA "reactivated" Power Saver "without showing it.

And so it is very similar with "SuperSLFM.
Operating sequence errors.

I can not say more. If you can run HWiNFO32, the CPU bar sometimes appears in red. This happens so fast that it can hardly be perceived. (Under a second) this moment I have been fit and creates a screenshot. The value of 4.1 GHz I saw only on the screenshot and was very surprised.

No problem, I write from now on German.
This would save me a lot of work.

 

I found a few problems. After you select Start Dual IDA mode, if you click on Power Saver or some of the other settings, it can cause some problems like you have found. I'm working on changing a few things so ThrottleStop works more consistently and warns users or blocks Power Saver mode when it is trying to get into Dual IDA mode. Just a few minor changes for the next update. I like software that works consistently with no surprises.

 

I have a question about my X9100, so far i have used it at x12.5/1.2125v, and it works fine. The problem is I want to bump up another notch to x13, the thing is orthos always gives me error after a few min until I increase the voltage all the way up to 1.3v.

Out of curiosity, I tried the x13 multiplier at default voltage 1.2625v(3.46Ghz), I tried some CPU intensive games(dragon age, etc), some CPU benchmarking(cinebenchR10, 3dmark06 CPU test), and also fired up dolphin emulator and did a long session of emulation, which is very CPU intensive, oddly enough, I didn't get any errors and lockups during the usage. Does it mean I can safely use the x13/1.2625v 24/7?

 

If there is a CPU error or two in a game, you might never notice the results and the game might keep playing just fine. If you fire a gun and the bullet misses its target because of a CPU calculation error, you're never going to notice and life will go on. When Orthos or Prime95 starts giving you errors it means it did something like added 2+2 and got 5 instead of 4. That's not a problem for a game but that is a huge problem for a computer that is searching for the next Prime number.

Run your CPU as fast as you like but keep in mind that if it is producing errors, in the long run, you might have to reinstall Windows if it gets corrupted or might see some other hard to explain problems. For me, that's not worth it. I could run my CPU a little faster but no one would be very happy with me if ThrottleStop had some random data in it. TS is dangerous enough as it is.

I found some fixes for the problems that RickiBerlin found. I'll post an update in the near future. I added a couple of safety features like blocking access to the Power Saver feature when TS is already trying to enter Dual IDA mode.

 

That sucks, I thought it would be okay to run it as long as the it doesn't cause programs/games to crash. Guess I will stick with x12.5/1.1125v then. Thanks! ahh, only if I had a E0 x9100!! well, can't ask for much as I only paid 150$ for it!

 

You exaggerate! There are no real problems. ThrottleStop is an excellent program and has a display for all important parameters.
Not to mention how many platforms it must be reconciled.
Thank you for your always quick answers!

 

There is always room for improvement.

Your suggestions have helped make ThrottleStop a better program.
I appreciate the feedback.

 

I did some further tinkering with the x9100, couldn't give up the thought of taking it up to 3.6 Ghz It stay stable at x12.5/1.2125v, but needs 1.3v in order to stabilize the next multiplier!

Funny thing is once it gets passed x13/1.3v, I only need to increase another 0.2v to make it stable at x13.5(3.6ghz)!! Stress tested with orthos for 15 min, and it produced no errors at all! Temp hit a maximum of 90 C during the test. I suppose it is acceptable?

I did some Wii emulation for an hour, here is the screenshot showing the temp after that(about 10 C higher than it at 3.33ghz/1.2125v), should I keep it running at 3.6ghz/1.325v all the time? I don't suppose 1.325v is too high for it to run on daily basis? It's a 0.06v increase from the default voltage(1.2625v).

 

I'm totally new to this kind of programs and i'd like to understand some basical things, 'cause what i'd like to do is some undervolting to lower a little the temps...
When i select a profile (for example battery) i have some options to change and one is "Set multiplier"...
Well...which value should i put in here?
I mean...i tried also rmclock and i had more than one value to adjust...
Maybe the value i use is the maximum that the cpu can use after i setted it?

In you opinion which is the best way to use ThrottleStop to lower the cpu temp?
Sorry for the probably dumb questions...

 

I don't think a VID setting of 1.325 is too high to run on a daily basis. The Intel design includes some vdroop so at full load, that setting should result in a little less than 1.30 volts going to your CPU. The Intel datasheet for a similar T9500 shows that 1.30 volts is the recommended maximum voltage when the CPU is using Intel Dynamic Acceleration (IDA). With the help of ThrottleStop, you are creating your own IDA mode. 3.6 GHz is a little more aggressive than Intel intended these mobile CPUs to be running at but the actual voltage getting to your CPU isn't beyond spec. I've been running my desktop QX9650 with a little more voltage than that without any problems so far. If it ever degrades, I have some Core 2 Duos in my closet collecting dust so it won't be the end of the world if my Quad craps out.

Core 2 Duo processor stability is directly related to the temperature you are trying to run it at. You may not be Prime95 stable for an hour at 90C but if your CPU is stable enough for the tasks you run on it then I guess it's good.

MarkraM: I found when developing ThrottleStop that most CPUs spend very little time at the intermediate multiplier ratios so I decided not to use them. That gives you less control than RM Clock but the advantage is that ThrottleStop is a lot easier to set up.

I also found at idle that most of the mobile CPUs enter one of the deeper C3/C4/C6 sleep states so the CPU automatically uses a lower VID regardless of what you set in RM Clock or ThrottleStop. On the newer 45nm CPUs, a lot of what RM Clock can do really isn't that necessary anymore.

http://forum.notebookreview.com/6369641-post5060.html

My conclusion was the only important voltage to control is the full load voltage so that's the only thing that needs to be adjusted in ThrottleStop. Tell me what CPU model you have so I can help you out.

The latest beta version of ThrottleStop includes battery power consumption monitoring which will make it easier to see what settings are useful and which ones are not.

ThrottleStop 2.90 beta 6
http://www.mediafire.com/?jdal8a26utytlum

 

Thanks, that's kinda reassuring. 90 C is the highest temp during orthos test, normally it hovers around 70 C(with occasional jump to 75-77C) during intensive usage. I know lots of people with core 2 duos and i core cpus usually run into 80 C zone under load, I suppose it's safe as long as the temp is okay, right? Thank you, unclewebb!! This program adds a lot of extra value to my X9100!!

 

First i want to say thank you for the reply, you couldn't be clearer than that.
My cpu is a common Core 2 Due P8600 (2400Mhz), and my aim is to get a little increase in battery life and if possible a little lower temperatures of the CPU.
What settings do you suggest for Multiplier and other options?

Great news the one regarding power consumption monitoring, really useful, you've made a great tool.

 

By being able to monitor power consumption with ThrottleStop, now you can do your own testing and see what works best for your CPU in your laptop. There are 101 variables so rather than me trying to tell you what is best for your laptop, you're better off doing your own testing.

The default multiplier for a P8600 is 9.0. This CPU also supports Intel Dynamic Acceleration (IDA) which means when one core is asleep, the other core will automatically use the 9.5 multiplier for a small increase in performance. That's only a 5% performance boost at max so some users use a ThrottleStop Set Multiplier setting of 9.0 which disables this feature. The only reason to do this is if this multiplier requires significantly more voltage to be stable. Some of these CPUs are like that. If you are concerned about maximizing battery life then using IDA is not going to be worth it if you have to significantly increase the VID voltage to remain stable.

Run a stability test like Prime95 and see how low you can get your voltage while still maintaining full stability. Take the VID voltage down one step at a time and test for stability before lowering it another step. Everyone cheats and tries to go as low as possible as fast as possible. You might get a BSOD if you do this and you can also corrupt Windows so be careful. No cheating. I think most of these CPUs should run reliably at 1.100 volts but I haven't tested any to confirm that. Some are quite happy at a lower voltage than that but the only way to know for sure is to do some testing.

I think the Power Saver feature is more useful for the older 65nm Core 2 CPUs.

Try enabling the C States option at idle and when partially loaded to see if that decreases power consumption any. A 1024M and single thread of the TS Bench test should create a nice consistent load for testing purposes. At full load, the C States option shouldn't make any difference.

I mostly use my laptop plugged in so maximum performance is more important to me than maximum battery life. Do some testing and if you learn anything new, post it here to help others.

 

Thank you once again unclewebb
after the first testing that i've done i've been able to set the multiplier to 9 @ 1.0750V and it seems stable.
The Cpu also runs a bit cooler, nice.
Regarding the battery life, from what i've understood, it's hard to increase it because on a daily lightweight usage the voltage that is more used is the SFLM one, that cannot be changed (am i right?).