The ThrottleStop Guide

@Unclewebb: can you update the original post with your latest beta build? I just went through the pages looking for latest build links.

 

ThrottleStop 3.20 beta 2
http://www.mediafire.com/?4ug155h06073w5v

You can post this link wherever you want but I prefer to keep the beta builds fairly quiet for limited distribution while I iron out the bugs. beta 2 is very close to the next official version.

 

ThrottleStopDocs.html is a simple html file that opens fine for me in IE, Firefox and Chrome so I'm not sure what the problem is. You can try downloading it again and extracting all of the files from the zip archive. With ThrottleStop running and opened up you should be able to push the F1 key on the keyboard and the docs should open up in your default browser. Here's what the docs say about the SplitQuad option I was talking about.

ThrottleStop does not continuously calculate the bus speed value like most monitoring software does. On Core 2 CPUs, software must fully load a CPU core for a small amount of time over and over again. If you are not using SetFSB then the bus speed tends to be extremely consistent and it's pointless to calculate this value over and over again, especially on a laptop.

Each time you click on the ThrottleStop FSB button, it will recalculate your bus speed. The only other times it calculates your bus speed is when the program first starts up, when you resume from sleep and when you unlock your multiplier and increase it beyond the default.

When TS is reporting the wrong value, does clicking on the FSB button correct it? At any time, if you click on this button 20 times, does it report a correct and consistent bus speed? Does this problem only happen after you resume from sleep or hibernate? On my desktop and laptop and on the majority of screen shots I've seen, TS tends to be very accurate but I've also seen a few screen shots where it is out to lunch.

TS depends on a high performance timer within the CPU. There might be a problem with this timer in some of the ES CPUs. I wrote a program called WinTimerTester that compares the two main timers within a CPU to see if they are running in sync. After about 60 seconds, your two timers should be very close to a perfect 1.000:1 ratio.

WinTimerTester 1.1
http://www.mediafire.com/?xzo9n84d8lze9nb

If your timers are OK then think hard about when this problem happens so I can try and debug it.

I have the same problem where RM Clock exits correctly on my Core 2 Duo but hangs on my Core 2 Quad forcing me to use the Task Manager to kill it. That's a minor inconvenience compared to the performance hit I get when running RM Clock.









The problem is that when lightly loaded, RM Clock does not allow the CPU to get up to full speed. The Super PI benchmark is fully loading a single core but Windows is designed to rapidly move tasks from core to core to balance the load. Even when there is a significant load on an individual core, RM Clock still doesn't let that core run at anywhere close to its rated speed. That's a bug. I don't know if the Core 2 Quad mobile CPUs have this same bug so if you're interested, try doing some testing similar to what I did.

SuperPI
http://www.techpowerup.com/downloads/366/Super_PI_Mod_v1.5.html

 

Whoa, this is a very useful feature! Is there any way to set two cores to run at maximum AND the other two at a lower (manually set) multiplier? Basically, set different multipliers for each physical core?

That would allow me to find the highest multiplier for each die at the lowest voltage possible - get the most performance at the lowest heat and power consumption!

Just tried clicking the FSB button when it shows the wrong value - it still shows it. However, it seems to me that it's related to the voltage - if I increase it over 1.2V even at the stock 9x multiplier, the FSB reading goes haywire. Must be something with the CPU, it's ES after all, but it works fine in any application.

Edit: RMClock definitely does not throttle down the CPU multiplier in my case - it works at the full multiplier, although half multipliers are not supported, of course.

 

Are all the guides for throttle stop posted on the first page? There's a lot of pages to read through, and was just wondering if there's any guides not linked to the first page or not?

 

The Core 2 Quads internally consist of 2 separate Core 2 Duos. You can run each Core 2 Duo at a different multiplier but I'm not sure if you can run each Core 2 Duo at a different VID. When you're doing undocumented stuff with Intel CPUs, there isn't a book I can read that will tell me what's possible and monitoring software might not exist that can report this correctly. The entire CPU package might end up running at the highest VID. I'll think about adding more SplitQuad features in the future.

If you can make your MHz go haywire at the default multiplier by raising the voltage, can you try running my WinTimerTester program that I posted above. Just click on the Start button and see if the two timers are running in sync. After 60 seconds these two timers should be running very close to a 1.0000:1 ratio. I like using this timer because it is extremely accurate when it is working correctly.

I posted a registry fix a while ago so RM Clock can work correctly on the newer 45nm CPUs that use half multipliers.

http://forum.notebookreview.com/6130270-post4855.html

Can you post a screen shot of ThrottleStop in monitoring mode while using RM Clock about half way through a SuperPI mod 1M test like I posted above. I'm curious to see what multipliers your CPU reports.

kobe_24: ThrottleStop is a multi-purpose tool and can do different things depending on whether you have a Core 2 or Core i or one of the Extreme versions of these. The first post contains 3 guides and there is the built in documentation that covers most of the main features. There should be enough info there to get a user heading in the right direction. I'm sure you figured out how to get your X9100 running like a beast with the help of ThrottleStop. It's so simple that I don't think a guide is really needed but here's an old guide just in case.

http://forum.notebookreview.com/har...unlock-core-2-extreme-multiplier-windows.html

 

Well, I know the VID is fixed, but changing the multiplier separately is enough: some T9400 chips run 9.5x at 1.10V, while some need only 1.025V - more than likely, the latter can run at 10x or even 10.5x at 1.10V, which means you can get a performance boost at lower power consumption and heat compared to a C2Q overclocked across both dies to 10-10.5x at 1.15V. I think it'd make a great feature!

 

Unclewebb, you were right about the performance hit in Power Saver/Performance on Demand modes, however I think that the reason is the transition time - the default in RMClock is 200ms, which adds 4 seconds to the SuperPI calculation time. I changed it to 100ms, and the difference is only 1 second, Throttlestop being faster of course. Throttlestop switching time is super fast, going from 6 to 8x in a split of a second.
When setting a fixed multiplier in either utility, the result is pretty much the same.

But that doesn't really show in most apps - only real time audio/video processing might benefit, but for that purpose, I always say it's best to disable any power saving features (including on the video card and all ports) to achieve the best latency.

http://imageshack.us/photo/my-images/410/captureqm.png/

http://imageshack.us/photo/my-images/828/captur2h.png/

http://imageshack.us/photo/my-images/571/captur3.png/

http://imageshack.us/photo/my-images/542/captur.png/

 

Your ThrottleStop managed Power Saver test is way too slow because you have RM Clock running at the same time. If RM Clock is completely turned off, ThrottleStop Power Saver mode should be very competitive with the low 18 second Super PI times you were getting when using a fixed multiplier. It will be a lot better than the 27.1 second result that you posted when the two programs were probably fighting against each other for control of the multiplier.

A 27 second result when using RM Clock confirms that it has the same issue on your Core 2 Quad that it has on my Core 2 Quad where the CPU is being prevented from running at its designed speed.

I'm going to try and reboot and use a 266 MHz bus speed and a 10.0 multiplier for a better comparison.

 

Well I'll be damned! You're right, with RMClock not running (after I killed the process, again), Throttlestop's Power Saver results have improved A LOT. What the hell, I thought RMClock was supposed to be disabled on No Management.

The PowerSaver C0% setting was 95% initially, giving a 22sec result in SuperPI (still better than RMClock!). Lower settings give better results - I suppose that at 80-90%, the CPU doesn't switch fast enough to compensate for the increased load or it's just cycling through multipliers up and down too quickly. I found 70% to be the best setting.

Yep, I think I'll be using Throttlestop from now on .

 

I adjusted my Core 2 Quad so it is running at 10.0 x 266.7 so our SuperPi benchmark numbers will be a lot more comparable. My desktop motherboard is using a DDR2-889 memory speed so I might have a slight performance advantage but our results should be close.

The first comparison is ThrottleStop vs RM Clock when using a fixed multiplier. Both programs are very close with a slight advantage going to RM Clock.





The second set of tests compares the power saving modes of each program. Both programs use a 6.0 multiplier when the CPU is idle. The difference is that ThrottleStop lets the CPU quickly get up to the maximum speed and stay there for the duration of the benchmark. When RM Clock is in charge of the multiplier, the CPU rarely gets above the minimum 6.0 multiplier and that kills single threaded performance.

Here is ThrottleStop showing slightly slower performance compared to a fixed multiplier.



The next picture shows the performance hit when RM Clock is in charge of the multiplier.



When the benchmark time goes from 18.1 seconds to 25.7 seconds, that confirms that RM Clock is preventing this CPU from running at anywhere near its designed speed. The reason the CPU runs cooler is because it is killing performance. Your 27 second Super PI benchmark shows that RM Clock has the exact same bug on your QX9200.

The next argument I usually hear is OK, ThrottleStop is clearly faster but running my CPU slow will save me power. Peak power consumption is definitely less when you run a CPU slow but total power consumption in the above comparison is going to be very similar and sometimes ThrottleStop will consume less power.

The choice is 25.7 seconds at a moderate speed or 18.1 seconds at full speed and the other 7.6 seconds with the CPU at idle in one of the low power sleep states. The testing I did showed that having a fast computer can reduce power consumption because it lets the CPU spend a bigger percentage of time in one of the idle C States.

Here's an interesting paper that was written by some smart guys at Berkeley that challenges the old myth that a slow CPU saves power.

Power Optimization – a Reality Check
http://www.cs.berkeley.edu/~krioukov/realityCheck.pdf

It's really hard to convince people that a low multiplier does not save power. When a CPU has work to do, you want the CPU to get up to full speed where it can run more efficiently. Maybe lowering the PowerSaver C0% number in ThrottleStop back to under 50% might be a good thing for performance and it might reduce total power consumption.

The SplitQuad options are very limited at the moment. Feedback about this feature has been virtually zero so I haven't been too motivated. Maybe someday.

I'm not sure why you disable EIST when booting up. Any CPU should be able to boot up fine without a BSOD when EIST is enabled. I think you would be more likely to get a BSOD when EIST is disabled and the CPU is stuck running the lowest VID voltage as it boots up and is under load. ThrottleStop doesn't increase the maximum multiplier until after you boot up into your account so it shouldn't be causing any issues. Can you explain the problem you are having in a little more detail and I'll see what I can come up with.

 

I'm more worried about heat, and my CPU is always loaded at 6x, so it kind of never sleeps (I also run my computer without shutdowns or resets for months) - it's a dilemma...

 

Maybe.

What register are you talking about? It might be a secret. Send me a PM message so I can say more.

When do you get BSODs? After the CPU has warmed up and you reboot? When ThrottleStop unlocks the CPU, I think some bios versions on some laptops are not expecting this so they don't bother resetting the maximum multiplier. You could try rebooting and see what maximum multiplier you get without running ThrottleStop to find out how your bios handles this.

Your bios might also be blocking some of the deeper sleep states but I'm not sure how to monitor for this on a Core 2 based CPU. Try using the ThrottleStop C1E feature. It might help reduce power consumption when partially loaded without killing performance.

 

unclewebb have you had any luck on finding a way to mod bios's to allow overclocking in windows on SB systems? Just ordered an SB system with an 2920xm ES (less than 1/3 the price of a retail one) which has an unlocked multiplier. I didn't get it specifically to overclock, but overclocking would be a nice plus.

 

unclewebb, how can I use throttlestop to maximize battery life on a SB cpu? is it as simple as the power saver button or can more be done?

 

Khenglish: I wish I could help you but I have zero bios mod experience. I also don't have access to a laptop with a SB XM CPU so it will be a while before I learn how to do any overclocking.

technos: Same problem. I don't have access to any SB hardware so users will have to figure out what ThrottleStop settings work best for performance or battery power consumption.

 

Check what maximum turbo's are shown under TRL.

 

Will check when I actually get the laptop. Also EIST can be disabled in its bios (Dell Latitude E6520 btw). Maybe the core 2 dual-IDA trick could work on SB unlocked CPUs?

 

Nope, there are no Dual IDA tricks for Sandy Bridge.

 

After playing around with Throttlestop I've got issues with the minimum processor state and the processor downclocking. Even after restoring the power plans defaults (5% minimum state) it seems that the processor stays on maximum regardless. The only plan that remains working is "notebook max saving". but high performance and balanced no longer downclock
when starting TS and enabling power management downclocking works again. So it seems that some registry setting has been changed somewehere?

So how do i restore Win7 powermanagement without using TS, if restoring powerplans to default no longer works?

 

If you reboot your computer and don't run ThrottleStop, the Windows power plans will be in full control of your CPU. ThrottleStop does not write anything to your registry.

When you exit ThrottleStop, if you want Windows to immediately take control of your CPU then make sure that this box in the Options window is NOT checked.

"Do Not Reset FID / VID on Exit"

ThrottleStop is designed to work correctly when using the Windows high performance profile with the Minimum and Maximum processor state set to 100%. After you exit ThrottleStop you will have to adjust your Windows profiles to Balanced or whatever you prefer.

What software are you using to determine this? CPU-Z has been adjusted for consistent validation purposes when a CPU is lightly loaded so what it shows may or may not be a good indication of what speed the CPU is running at internally when idle.

If you want a more accurate look at what your CPU is doing when idle then try this.

RealTemp 3.69.1
http://www.mediafire.com/?4uixpjtezznuzkd

 

It's been a week since I first used this program and I'm just now noticing that it's been throttling my 2630qm at 2ghz the whole time unless I have ThrottleStop running, even though I last set it to Turbo multiplier and unchecked Disable Turbo... Unchecking "Do Not Reset FID / VID on Exit" seemed to have fixed the issue, until I restarted the PC... Now we're throttled at 2ghz again unless I start up TS :/

 

Now I'm throttled at 800MHz unless I turn it on... I don't recommend anyone try this program on an XPS 15 or 2630qm

UPDATE: I think I found the root of my problem. For some reason disabling SpeedStep in the BIOS limits the CPU to 2GHz Enabling it gives me back the 2.6GHz. It's not the program after all!

 

On some motherboards, EIST has to be left on for Intel Turbo Boost to work correctly. If you disable EIST on Core 2 CPUs, they usually boot up with the minimum 6 multiplier and are stuck there and can't go higher. I am not sure why users hate EIST and always disable it. I leave EIST enabled in the bios and then I like to use the Windows High Performance power profile and I set the Minimum processor state to 100%.

On the non-Extreme Core i CPUs, the C3/C6 sleep states also have to be available for full Turbo Boost to work correctly.

Glad to hear you got this solved and ThrottleStop wasn't the problem.

 

Im just wondering what i can do to make my system work without throttlestop running from startup. If i have EIST enabled it will BSOD. If i have it disabled works fine, excpet less frequency. It would BSOD on boot.

I am assuming it is because of the multiplier and voltage going up and down. Which resulted in a low voltage and a high frequency.

My system is an M1330 with a X9000

 

Users hate EIST, because laptop makers like Lenovo use EIST to throttle the max speed at which the Laptop CPU can operate on battery power. For example, my W520 maxes out at 1.2 GHz on battery power (running Prime95 with four threads active), despite the fact that the power profile is High Performance and the Minimum processor state is 100%. If I turn off EIST, I lose turbo on AC power, but the laptop gains the ability to run at its maximum frequency, which for the i7-2620M is 2.7 GHz, even when on battery power. Of course, this means that the battery won't last as long before it discharges, but I can utilize the full performance capability of my laptop while on the go.

Finally, EIST is junk! If it wasn't for the Turbo disabled issue on the W520, I would leave it off all of the time. With modern C states, who needs slow OS controlled CPU performance increases / decreases when the CPU can do this much better in near real time and with a ladder of more and more power savings (all the way to turning a core off).

 

I have been following the Lenovo issues on a couple of their laptops. EIST is not the problem. The problem is Lenovo locking their laptops to the minimum multiplier when on battery power. Lenovo is selling a computer that can not run at its rated speed when on battery power but doesn't have the guts to admit it and then continues to release bios updates that don't address the problem.

Most Core i mobile processors that I know of need EIST to be enabled in the bios so Turbo Boost works properly. That's part of the Intel design. If you turn off EIST in the bios, you can't blame Lenovo or Intel when your CPU doesn't work properly.

With EIST enabled, can't you use ThrottleStop so you get full turbo on AC power as well as run your CPU however you like when on battery power? I can't understand why Lenovo users continue to wait for a Lenovo bios fix which doesn't seem to be coming when ThrottleStop is available to take care of this problem. No one should be forced to run a third party program so their laptop runs as advertised but if this is the only option then why not use it? Can you boot up with EIST enabled and post a screen shot of ThrottleStop while running on battery power? Run a simple benchmark program like Super Pi mod and show me a screen shot about half way through that benchmark. I am interested to see what sort of multipliers you are getting across all threads while on battery power with ThrottleStop running.

Super PI mod
Download Super PI Mod v1.5 | techPowerUp

TheCodeBreaker: If a laptop can't boot up with EIST enabled then your CPU is defective. If you can only boot up with EIST disabled then do that and use ThrottleStop to adjust your CPU once you get into Windows. What sort of speeds and VID settings are you trying to run your X9000 at? Is your laptop Linpack stable. It might have a bad memory module causing the BSOD. If it is an ES CPU then it might also be a buggy CPU that might not have been 100% stable the day it left Intel.

 

Kevin, I can overclock it pretty well actaully using throttlestop, i can actaully run wrpime1024m at 3999MHz, except for the bad cooling, and the PSU i used my M17X's PSU lol. it is an ES chip, but im pretty sure its okay. I just need to figure out what im missing, all i can see in the bios is IDA, and EIST for the CPU.

I am actaully trying to run it on stock, using the bios itself. When i was benching it, i did not realize EIST was disabled.

SuperPI
Hardware news, Overclocking Competitions, Reviews

SuperPI32m
Hardware news, Overclocking Competitions, Reviews

CPU-Z
Hardware news, Overclocking Competitions, Reviews

WPrime1024m
Hardware news, Overclocking Competitions, Reviews


Ive been using throttlestop for some time, just not sure what im missing, currently have IDA disabled, and seeing if i get a BSOD running stock, so good so far. Could the chip be really defective just becaue of EIST? It Overclocks pretty well, considering its in a 13inch laptop.

 

After seeing those screen shots, you better not throw that chip out unless of course you want to send it to me.

Do you know if your board enables SLFM by default? You would need to delete the ThrottleStop.ini file and then boot up and see if that option is checked.

One problem I've seen before is when lightly loaded, these CPUs can enter some of the deeper sleep states where the VID voltage drops down significantly. When a 45nm Core 2 CPU is brand new, this isn't a problem. As they age a little and after they get pushed hard, the voltage requirement to remain stable goes up a little but the CPU isn't able to request more voltage when it is in C3/C4/C6. I don't know of any way to control these voltages and RM Clock doesn't let you control them either.

Here are the specs for an X9000 C0 stepping:

VID = 0.650 - 0.850 ~ C4
VID = 0.600 - 0.850 ~ DC4
VID = 0.350 - 0.700 ~ C6

If one of these CPUs enters the C6 sleep state, the voltage might not be enough to remain stable. With your very limited bios options, there isn't anything you can really do about this.

The Intel docs show that the X9000 does not use IDA mode so if your bios or ThrottleStop gives you an option to play with this, it probably doesn't do anything.

I don't know how to solve your BSOD problem.

 

It definitely is a good chip, the thing holding it back would be my laptop, lol. I thought i would bench it and sell it, but the performance gains are pretty sweet.

I think it does enable SLFM by default. I deleted the ini file, rebooted, checked the bios, IDA disabled, EIST enabled. got into windows, checked CPU-Z running at 2.8GHz. unplugged the wire, and replugged it in. got the following, not sure how it all comes together.



The voltage is low.
I appreciate your work Kevin, TS is awesome!

I did get a BSOD after some long use of windows before i rebooted, unpluggin and replugging, deleted ini file started TS, and deleted it once more, not sure what the settings were though. but i believe TS was off.



This was running TS after deleting the ini file and botting it up. SLFM checked.

 

Your screen shot with the 6.00 multiplier on both threads is showing a C0% of 36.4 so it is not using SLFM yet but when the CPU settles down after it is finished booting up, it might drop down.

The BSOD problem you have seems to be random. With the limited adjustments you have in the bios and with these CPUs automatically entering various C States when lightly loaded, I'm not sure if you will be able to control it or fix it.

If you can boot up reliably with EIST enabled, I would try running ThrottleStop at boot time and use it to disable SLFM mode and I would use it to disable the C States option like you are doing now.

If the BSOD is being caused by the CPU dropping down to C3/C4 without enough voltage, you probably won't be able to avoid this. Leave ThrottleStop running with those two items above unchecked and see how long it runs before a BSOD. Maybe try running MemTest86+ overnight to see if it finds a problem with your memory.

Memtest86+ - Advanced Memory Diagnostic Tool

I like using the Prime95 Blend test for checking memory.

 

FYI: The current SNB iMac i7-2600 start with C6r deepest package C-state and enabling C7 via TS has no effect. TS allows to change the TRL value numbers, but it has no effect. The rest seems to be working (unlocked), although I am not sure about C1 Auto Demotion yet (C3 works).

 

ThrottleStop 3.20 is now available.

Thanks for the feedback T1mur. From what I've read, there is no way to increase the multiplier with Sandy Bridge when you are in Windows. This can only be done on the unlocked K or XM CPUs and you can only change this before you boot up. I was going to block the TRL window for Sandy Bridge CPUs but I haven't done any testing to see if this feature could be used to lower the maximum amount of turbo boost. If that doesn't work either then I will block this feature for Sandy Bridge CPUs in the next release.

I also added a Priority= INI option so you can control the priority that ThrottleStop runs at. If you don't have a throttling problem then you should be able to lower the priority without any problems. Check out the included docs to find out more.

 

Just an FYI, I did try disabling EIST with turbo active, EIST just stayed on so you were right, the C2D IDA trick doesn't work on SB.

 

is up with having to register to download now?

 

Yes, can we have a link where we don't have to sign in blood?

 

Hello!

I have a Acer Timelinex 3830TG laptop with a i5 2410M CPU.
It has quite bad cooling so I get some throttle problems if I play games for a longer time.

I've downloaded throttlestop 3.00 but Im not sure what options I should have (I don't understand everything either).

if I want it around 2-2,2 GHz should I set multiplier to around 20 (20 x 99.8 MHz)? I've also enabled "Disable Turbo" if that feature works

 

I didn't realize that Tech Inferno was trying to sign up some new members. ThrottleStop has a bug on Sandy Bridge processors when the "Do Not Reset FID VID on Exit" option is not checked. I think that's been fixed now and I've added some more data to the log file for Sandy Bridge owners but I don't have any hardware to test on. As soon as I get some feedback from my testers, I'll release an updated version and post a direct link here if I have to. Hopefully this weekend.

Braide: I don't own an Acer 3830TG so try going to that forum to find out more information about setting up ThrottleStop for that laptop. Turn on ThrottleStop, check the Set Multiplier option and setting that to about 20 should help your Acer from getting too hot when gaming. I think for some games users are going down to about 18.

 

Just thought I'd post about something that seems odd to me. When playing around with prime95 and throttlestop, I noticed that setting a multiplier in throttlestop would make that multi the max multiplier, while the CPU can go lower if it wants. What's strange is if I set the multiplier to "turbo", it sets the maximum to x23, which is below my max non-turbo of x24. My max 4 core turbo is x32, and prime95 usually shifts between x25 and x26 with all cores active if I did not set a multiplier. Where did the x23 come from?

 

The processors being sold on EBay recently that are being called 2920XM are not the same as a genuine 2920XM CPU. Can you post a screen shot of the ThrottleStop TRL window so I can have a look? Reboot and delete the ThrottleStop.INI file first so ThrottleStop can read the details of your CPU in its default condition.

What Windows power plan are you using?

Can you run a single thread of Prime95 and take a screen shot of ThrottleStop while this benchmark is running so I can have a look.

 

The problem went away after I deleted the .ini file. I posted pics anyway since I took them and wrote the post before I realized the problem was solved.



I'm using balanced. I tried maximum performance some, but I didn't notice any difference in benchmarks so I just went back to balanced.

Here's a single thread in P95 without Throttlestop on:

 

ThrottleStop 3.30 should be available now. Let me know if you have any problems downloading it.

Khenglish: Your CPU seems to be running as designed. These Intel 0 ES CPUs are not exactly the same as a retail 2920XM but they are very similar. You are getting a 32.22 multiplier on the hardest working thread. When running a single threaded benchmark, there are lots of other background threads running on a Windows PC which keep waking up additional cores. Your multiplier shows that between 2 and 3 cores on average are active which isn't too unusual.

ThrottleStop is designed to work on the Performance profile. When you use ThrottleStop on the Balanced profile, you end up with ThrottleStop and Windows both writing data to the same register and fighting over control of the CPU so I don't recommend doing that.

Can you show me a screen shot of the problem you were having before? Make sure you are on the Performance profile, ThrottleStop is enabled, Set Multiplier is checked and you are running at least 1 thread of Prime95. If the CPU is not loaded, it can be entering and exiting various C States so is difficult to tell if it is working correctly or not.

 

I simply cannot recreate the problem. Like I said before, it went away after I deleted throttlestop.ini. I swear I've tried everything I did before to recreate it, but I can't.

 

On my mobile SNB i7 Quad it's blocked already anyway. On the 2600 of the iMac 27" lowering TRL values seems to work (minimum allowed by TS is 34 and it effectively caps the i7 to that value).

Having a "Reset Defaults" might be useful for the TRL dialog.

 

hey i've downloaded that software and made some changes like a noob and i think i messed it up, i wanted to ask if i exit the software does it restore the default settings of CPU or the changes i made becomes permanent if so PLEASE HELP ME to restore back to default settings
Thank you.

 

Just reboot the computer. Before you restart Throttlestop next time, delete the .ini file.

 

does throttlestop have logitech lcd support?

 

No. Maybe someday.

 

okay thank you

 

I decompressed my bios, but kizwan and I are having problems locating where MSR 0x610 bit 63 is written.

Unclewebb, I believe you once posted the windows max multiplier bit lock too. I looked in this thread for an hour and couldn't find it. Could you please repost it, or tell me if you never did actually post it and I am just crazy.