How to Supercharge the M11x with Core i5/i7 UM CPUs

Ok people, Dell just change my MB, Hearsink and fan... temp are a lor cooler but still need to udate the BIOS and OC my little beast to do some benchmarks.

 

Ok, so far so good... furmark does not send the temperatures out of control.
Runing TS Bench 1024 and Furmark 800x600 and my tems are 75C for CPU and 65 GPU

 

Ok, the TS Bench 1024 have finished with furmark (800x600) running in a window. My temperatures reach 76C and the CPU NEVER throtle down, rocksolid at 14x

Thanks for such great tool to bring out the performance of our laptops and to test them (Throtle Stop).

now, abouth TS Junior, how progress going?

 

Good glad to here that your m11x isn't fried with a side of Alien.

 

Hi guys !
I wanted to ask, what are the dangerous temperatures for the m11x r2?
'cause I downloaded ThrottleStop and everything runs really smooth, but the problem is that the temperature is about 80+ after 10 minutes of BFBC2.

So, is it normal ? or how can I make the temperature drop ??

OC= 156 Mhz
Turbo Boost= On
ThrottleStop= TDP= 30
TDC= 30

BFBC2 settings = All medium, Vsync Off, HBAO= off, AA= x2, AN= x2

 

That seems a little warm, not dangerous though. When I play with similar settings and OC, I can usually play for several hours without exceeding mid 70's. Are you using a notebook cooler? Is there proper air flow for the vent while you are gaming?

 

I don't have a cooler, but I was thinking of buying one. Which one do you recommend? and I 'm playing on my desk.\
Also I notice that the laptop itself doesn't feel that hot to be at 85*C

 

A peak core temperature of 85C is at a tiny spot on the CPU core. When a CPU is loaded, it's not unusual for the center of the core package to be up to 20C cooler than this peak core temperature number. Heat dissipates very rapidly over a short distance in a CPU. That's why they never feel as warm as these peak core temperature numbers.

The Core i5/i7 UM Dual Cores are rated by Intel to run 100% reliably up to a peak core temperature of 105C so you don't have to worry too much as long as your laptop is running reliably. These UM CPUs, even when overclocked, are not running that fast compared to other similar CPUs so they have lots of temperature head room.

 

I typically have a maximum temperature of 84°C after an hour gaming session of Left 4 Dead 2 with my i5 using ThrottleStop and the CPU overclocked to 166MHz. Nothing really to worry about.

Typical gaming temp is around 78 for me, but there are quick peaks over 80. That is using it on a wooden desk without a laptop cooler or raising the back end. Of course doing either of these things brings the temperatures down.

 

I was just wondering since I don't really get throttlestop.
I have 2.99 version of throttlestop.
1. Does throttlestop affect overclock ability? for example, I can only overclock to 146mhz, but I want to increase it. If I use throttlestop, does the mhz go up?
2. I just set TDP and TDC as 30 and clicked set multiplier as Turbo. Also, in option mode, I clicked AC on, Battery off, battery monitoring.
Then if I turn on the throttlestop, Do I restart and change the mhz to be higher?
3. if im using the battery mode, instead of ac, all i have to do is just turn off throttlestop?

Sorry for many questions, I just dont wannt screw up my laptop.

 

my laptop is i7

 

1) ThrottleStop can significantly increase the amount of Turbo Boost that your CPU uses. On a Core i7-640UM, the multiplier can increase from 9.0 to 14.0 when fully loaded. That can increase your total MHz by over 50%.

2) because of that potentially huge increase, you might have to reboot and lower your 146.0 MHz overclock. Do some stability testing and find out. Many users are overclocked to over 150.0 MHz in the bios and some lucky users with good CPUs and memory are running flat out at the maximum bios setting of 166.0 MHz which usually gets you an actual reading of 167.0 MHz. When that is combined with the 14.0 multiplier, it will make a noticeable difference to the performance of an M11x-R2 for a wide variety of apps that you run.

3) When you click on Turn Off or use the AC on / Battery off option; as soon as the system tray icon turns green, ThrottleStop will be disabled and will allow your CPU to run at its default Intel specs. Turbo boost overclocking will be replaced by normal turbo boost operation. On battery power you can also use a reduced multiplier which reduces performance below default but that might help to increase your battery run time.

Read the first post lots and lots and then do some testing. Ask specific questions and other users or myself will be happy to help you. Not a single M11x-R2 owner has reported nuking his laptop while using ThrottleStop so I don't believe it is that dangerous on this model. It is also easy to set up so there's no need to be worried if you follow the directions in the first post.

With the latest version, I've moved the Turbo options from the context menu onto the main ThrottleStop screen so they are easier to find for new users. The TRL button will let you control the Turbo Ratio Limits in the Extreme XM CPUs and the new TPL button lets you control the Turbo Power Limits.

 

So the fact that even though I used throttlestop, I can't increase my mhz? Because I tried higher ones and it gave me a blue screen.

 

I have read that some users in your situation (Low Overclock limits) have seen gains by re-seating their RAM. Not sure If it is concrete though. Never the less Throttlestop is an amazing tool, and should be utilized by all M11-R2 users.

 

ThrottleStop isn't meant to allow you to achieve a higher overclock. It's meant to stop the CPU from throttling when under a high/full load.

 

jrockroll: Overclocking is never guaranteed. If you are getting blue screens then either your memory or your CPU is not up to the task. Other users have been luckier and have been able to overclock far more than what you can overclock.

ThrottleStop can't create a golden CPU out of a dud and it can't make your memory run faster than it was designed to do. Some people have a CPU and memory that loves to overclock. Other users get memory that barely runs reliably at default speeds.

If I had a friend with some similar memory sticks, I would try those but I probably wouldn't bother buying some new memory to try to overclock higher. I find that CPUs are fairly consistent but the capability of your memory can vary a lot. Memory manufacturers can vary the individual chips they use on their memory modules and not bother to change the part number so what one user ends up with is not the same as what his neighbor ends up with. Dell likely also uses different memory manufacturers depending on where an M11x-R2 is built and what memory modules they have in stock that day. That's not unusual.

 

Just a slight query about the new version of throttlrstop. The top mhz readings im getting when i hover above the icon are slightly lower than what i was getting on the old version, wprime readings are coming in slightly lower though, have there been some changes to the accuracy? Also the readings seem to fluctuate less as well.

 

I wish I had an M11x-R2 so I could test my algorithms but unfortunately I don't. Improved performance in the wPrime benchmark and a multiplier that varies less sound like the changes I made were an improvement. ThrottleStop can not change the accuracy of wPrime so if your times are better or worse compared to using a previous version of ThrottleStop, I'd trust that.

If the reported multiplier has changed slightly, try adjusting the Force TDP/TDC value. Different versions of ThrottleStop might require slightly different settings for optimum performance. There shouldn't be any huge differences. I think a while ago users found that peak performance was usually obtained just before the multiplier reached its maximum. If the maximum is 14.00 when fully loaded then somewhere around 13.97 or 13.98 might give maximum performance. Setting the Force TDP/TDC value as low as possible might give you a higher multiplier but might not result in peak performance due to more overhead.

If anyone wants to do any benchmark testing for me while using different versions of ThrottleStop, that would be appreciated. Without hardware to test on and without feedback, it's difficult to make improvements.

 

Hi,

I have posted in here a while back that I have had phenomenal results with TS 2.89. I have a m11x R2 w/ i7 and 4 gigs ram overclocked at 166 and TS set to max. I play starcraft 2 which relies heavily on the processor and I am able to play on full ultra graphics settings with no lag whatsoever for full 1v1s 2v2s and even 3v3's for the most part, however if it gets to late game in a 3v3 or 4v4 with A LOT of units out then it starts to lag. But I cant complain because most systems out there can't handle SC2 on ultra and I love ThrottleStop for letting this happen. Anyway I just downloaded the latest version I'm gonna crank it up to max with the overclock set to 166 and play some SCII for a couple hours ill report back with my results!

Unclewebb, I'll do some benchmark testing for you, just tell me exactly how to go about doing that lol

 

Hmm I'm going to need some help here. I am getting lag straight from the menu screen for starcraft 2 with the latest version of TS..I messed around with it and still no luck just lag. I tried the default settings, changed TDP/TDC to a bunch of different values, tried different force TDP/TDC values, turbo on and off..any suggestions?

I can play with TS 2.89 no problems still..I even set the same exact values and settings that I have with 2.89 on the new version and the new version still lags. What else should I try?

 

psilocybin: This is simple. If you are getting your best results when using version 2.89 then continue using that version and be happy.

ThrottleStop is a multi-purpose kind of tool for a wide variety of laptops. It's possible that the M11x-R2 will perform best with 2.89 while maybe another laptop model might perform its best with version 2.99.

I can change and improve an algorithm so it improves wPrime times but for another user, that same change might result in worse game play for Starcraft 2 or whatever their favorite game is.

If anyone has some game time demos or wants to post some wPrime scores comparing 2.89 to 2.99, that would be interesting. Maybe even some 3DMark06 runs or something like that will show a difference.

ThrottleStop 2.89

Without hardware to test on, I have to rely on user feedback. If anyone else has a negative experience with v.2.99 like psilocybin has had then let me know or post some numbers or let me know that everything is OK in the games you play.

 

Is the core i5 multiplier supposed to lock up at 12? I thought for some reason it was 13. Any feed back would be helpful. I switched to 2.99 recently and got confused.

 

The maximum turbo multiplier when both cores of a Core i5-520UM processor are active is 12.00.

No wonder people think version 2.99 is making their CPU run slower.

Another good test is to run ThrottleStop and select the Log File option as well as the More Data option while you are gaming or doing whatever and then post the log file to www.mediafire.com so I can have a look at how your CPU is running. Every log file tells a story and I can never get enough data so don't be afraid to upload something big. That's what MediaFire is for.

 

Hey,

I started to use TS a few days ago (about time).
So far I have to say that I didn't understand everything from the first post (I have to read it again), but I made a quick test with the game Darksiders:

- with TS on profile TDP = 10, there are cuts in the game videos (not sure why or if any other experienced the same)
- with TS on profile TDP = 30, no problem with videos.

So it isa very small thing but it made me like TS already

So thanks a lot for this tool.
Now I have to learn more about it

 

The best results I have seen is when I set my TDP-100 and my TDC-100, Then clicking "Options" and setting my "Force TDP/TDC" at 4. You can test which settings work the best for you by clicking on the "TS Bench" and clicking on the 32M or 1024M.

The first page was set up for a previous version of Throttlestop. While it is similar, the later version(2.99) which I am assuming you have, is a lot more convenient. I suggest reading through some of the more recent posts as well to get more familiar with it.

Edit: This is where the benchmark tool was added --> http://forum.notebookreview.com/ali...w-supercharge-m11x-core-i5-i7-um-cpus-55.html

 

Thanks a lot Josh, you've just saved me the time I don't have to read every post
+1 for you!

Now I guess i understood what I needed, so I'm going to run some benches with TS, and post the results here.
Indeed I'm using the last 2.99 version.

 

Hello,

So here are my first bench results :

1) Situation with TS turned off
GPU Temp = 51
CPU Temp = 51
Mult FID average = 9

2) Situation with TS turned on

Bench both TDP / TDC at 100
GPU Temp = 61
CPU Temp = 81
Mult FID average = 13.70
So from what I understood, this is ok since I reach the max multiplier for my i7 620Q.

The Force TDP / TDC option is at 16 in TS Options, not sure what it means really (do you care to explain?).

Now I attach 2 screenies with the results, I'm not sure how to interpret the high score results unfortunately. But the point is to have the max multiplier anyway (if I understood ok).

Thanks a lot for the help guys!

 

If you decrease the force TDP / TDC option while a benchmark test is running, you should be able to see the average multiplier in the FID column increase to closer to 14.0.

A better benchmark program is wPrime if you can get it working for you. I just put the TS Bench together fairly quickly so you could get a general feel of how your CPU is operating. Seeing how fast you can calculate 32M in wPrime will probably fully load your CPU more than TS Bench does. I'm not sure what benchmark is a more realistic benchmark if you are mostly interested in gaming. Some 3DMark06 might be OK for those kind of tests.

Here's some more info from the first post in this thread.

 

Ok, I get the Force TDP / TDC part.
I'm going to run some more tests, maybe not with Bench tools (I'm not used to them) but with games directly (and the log function of TS).

Thanks again for your help

 

Bench tools are good so you can understand how different ThrottleStop settings change your average multiplier when fully loaded but I always like seeing real world testing. Benchmarks aren't too important if your game performance doesn't improve any. Any modern game that is CPU dependent should perform better when using ThrottleStop on an M11x-R2. Even if you don't have some hard numbers, you can still post subjectively how a game feels and plays with and without ThrottleStop. I'm sure some users are skeptical and probably think that a big boost in CPU performance for free is too good to be true.

The log function of TS should show lots of turbo boost when gaming.

 

I tried using it, it didn't work

On CPU-Z it reports my multipliers are still being throttled at 6 to 7x multiplier while on throttle stop it says 11x or 12x multiplier, what is happening?

 

Read through the instructions again in order to configure TS properly. If you still can't figure it out then turn on logging in TS, run the TS Bench 32M benchmark and then post the log data here. It'll be easy to tell what's going on from that.

 

Here's what the programmer of CPU-Z has to say about what his program reports.

You can believe and fully trust the multiplier that ThrottleStop is reporting. I can't say the same about CPU-Z. I tried to convince him to follow the method that Intel recommends and ThrottleStop uses but he has decided to create his own method for consistent CPU-Z validation purposes.

Read and follow the first post in this thread. The ThrottleStop menu items have been replaced so with version 2.99 just click on the TPL button to access your turbo power limits.

 

The thing about that was when I used throttlestop, black ops started off with like 91 fps then steadily dropped down to 10 fps on all high settings no shadows 4x aa. When it was off, the game in a consistent pattern throttles and then goes back to normal. Using throttlestop gave me a performance boost at first but later on, it lags the game so badly i cannot play at all and it doesn't recover.

Btw do you still have the link for throttlestop 2.89 if that helps? did 2.99 have a different system configuration or it's just a UI change?

 

I ran TS with limits at 20/60 with OC at 166 and wPrime at 32M gives me 23.9 seconds but when i run the TS test at 32M i get 38 seconds. Any clue whats going on?

 

You might want to turn on TS logging and GPU-Z logging while you're playing so that you can see what's throttling down. Having an external display attached so that you can see things in real-time is also helpful.

 

There's nothing wrong. The 32M TS Bench takes longer to run the wPrime's 32M bench routine.

 

ThrottleStop 2.89

zerogeneration: Try running a ThrottleStop log file and a GPU-Z log file while gaming. Without seeing any numbers, it's impossible to figure out what is causing your FPS numbers to drop. It could be your GPU is getting way too hot. This doesn't happen if the heatsink is attached correctly and your laptop has adequate air flow and isn't hiding under the blankets.

In theory, TS 2.99 should work harder to reduce any throttling but if this is causing your computer to get too hot, then it might be triggering heat related throttling so you end up running slower. When your GPU gets too hot, this can cause clock modulation throttling to start. You haven't posted a picture of how you have ThrottleStop set up or any log files so that makes it impossible for anyone to figure out what problem you are having and to help you out.

Warned: TS Bench and wPrime are two totally different benchmarks so the numbers are not comparable. Maybe I'll change 32M to "Short Benchmark" and 1024M to "Long Benchmark" so there is less confusion.

Edit: slickie88's testing seems to show a slight improvement when using TS 2.99 compared to TS 2.89 so if you guys are having problems, post some log files so I can have a look. You might be running into a GPU throttling problem due to heat but only if your heatsink is not installed correctly. Thanks for doing some tests for me slickie88.

 

Bump...for all the newcomers

 

Thanks so much for this software. I have a little preview of an article I am writing for my site, pnosker.com | Tech News and Reviews!, that will feature this technique.

I have an Alienware M11x R2 with the Core i5. Using ThrottleStop and some other techniques, I got the wPrime score down to 25.62 sec, consistently around 25-26 sec. This, is faster than a Core i7 at 166 mhz FSB and running ThrottleStop. The stock speed on the i5 is around 28 seconds with ThrottleStop.

 

How are you achieving this quickness? I'm getting 37 seconds. i5-520UM
13.25ish X 167.0 MHz. TPL changed to 30/30, force TDP/TDC changed from 16 to 6

 

tyuppp: If you have a Core i5-520UM, you should only be seeing the 12.0 multiplier while wPrime is running. Make sure that CPU-Z shows 2 Cores / 4 Threads and the Task Manager shows 4 threads and that wPrime is set to 4 threads. I think you might have a problem. Edit: I know you have a problem.

Run msconfig, click on the Boot tab and the Advanced Options button and make sure number of processors is not checked. I think wPrime not set to 4 threads is the problem.

pnosker: How about sharing your secrets rather than just teasing users and redirecting them to your site?

 

I bet it has to do with the fact that SetFSB now supports M11xR2, and he can go beyond the maximum FSB of 166MHz as set in the BIOS.
You can even see the name of his image with the 188MHz in there.
No need to give him free page views

 

Nice find. Many users have trouble using the maximum bios overclocking option and running reliably at 166 - 167 MHz. It's going to be a very rare M11x-R2 with a Core i5-520UM that is going to be able to run 100% reliably at a base clock of 188 MHz but congratulations to anyone that is able to do this.

 

I am surprised I haven't seen more people use SetFSB on the R2. But as said, many have trouble reaching 166 BCLK in BIOS so maybe it's not so strange.

 

Do I have to enable the overclocking in BIOS as well? (I'm a computer illiterate, btw.)

Obviously something must be wrong...

http://img171.imageshack.us/img171/8020/throttlestop1.png

EDIT: OK I overclocked to 166MHz through the BIOS and Throttlestop seems to be working correctly.

http://img196.imageshack.us/img196/8224/throttlestop2.png

Does that look about right?

 

Dup post, see the one below

 

Hello guys. I have been reading the posts and times people are posting an mine does not seem right.

OC in BIOS to 154 mhz
Throttlestop as per pics:



MSI afterburner on:



Uploaded with ImageShack.us


And as you will see, mine bench is like 48? TDP:30 TDC 70. When running the bench, all 4 hit 12% (FID)

I have also just run wprime: 32M: 35.513 with MSI AB running and throttlestop. I have run wprime without TS and MSI AB and got 32.25 sec. Am I running quicker without TS and MSI on?

Sorry, I hit Quote instead of Edit!

 

Hi Guys,

Just got my M11x R2 and was googling overclocking and stumbled upon this awesome thread.

I think I understand most of it but I just want some confirmation.

I have the i5-520UM

What settings should I put into ThrottleStop 2.99.6 if I have set the BIOS to 166Mhz?

I was testing it after reading the replies on here and tried

TDP = 100
TDC = 100

and was able to obtain a score of 29.697 sec on wPrime v2.04

Is that good? I havent run any games to try anything yet, about to try all of that now.

I also dont understand the whole profile thing with TS. Can someone explain it to me?


Many Thanks

 

Yes that is a good score.

The profiles are like, well, different profiles for different settings. For example, Profile 1, you can set it to say "TDP = 30 & TDC = 30", and on Profile 2 you could set it to "TDP = 50 & TDP = 50", for quick easy switching between different options.