Help undervolting i5 4200h Throttlestop

[Bloodsucker]

Hi everyone, I apologize for my English, since this is not my native language, I write through a translator.
i have asus N751JK laptop, Intel® Core™ i5 4200H , DDR3L 1600 МГц SDRAM 8gb, NVIDIA® GeForce® GTX850M 2Гб , 256GB SSD.
I play mainly in WOT on medium or high settings of SD graphics, sometimes in GTA 5 also medium or high settings of graphics.
In the past few months, I began to notice that the laptop is very hot, that I began to strain, FPS drawdowns and image jerks are also noticeable
I installed Intel® Extreme Tuning Utilities and I saw that the thermal throttling processor
Dust removal and replacement of thermal paste coolermaster - 1.85 W / m.k., but this was not enough, the processor same trolles and the temperature reaches 85-86 and remains at this level while I'm in the game.
I tried to do undervolting in Intel® Extreme Tuning, there were 2 options, 1) everything remained as it was before undervolting 2) or undervolting worked and everything became fine, but after rebooting Windows died, nothing opened or an error occurred. So I reinstalled the system 2 times and stopped experimenting with this utility.
so I came to Throttlestop, and no matter what values I set, I did not notice. The only thing if you lower it below 0.120 mV, WOT just freezes.
if you need any kind of logs or screenshots, I’m waiting for your advice.

 

[unclewebb]

the temperature reaches 85-86 and remains at this level while I'm in the game

Product Specifications

quick reference guide including specifications, features, pricing, compatibility, design documentation, ordering codes, spec codes and more.

ark.intel.com

Intel set the thermal throttling temperature of the 4200H to 100°C. Asus is using a minus 15°C offset. That means your laptop will thermal throttle at 85°C instead of at the Intel recommended 100°C. ThrottleStop shows this information on the main screen where it says, PROCHOT 85°C.

That was a bad decision by Asus. This setting is too low. Setting a low thermal throttling temperature will reduce maximum performance. Your CPU will constantly thermal throttle when you are trying to play games. The GPU might also start to throttle.

Your CPU is running too hot, even when lightly loaded. That means your heatsink is not making firm contact with the CPU or you did not do a very good job with the thermal paste. Try doing that again.

On the 4th Generation processors, the CPU core and CPU cache voltages do not have to be set equally. You can try adjusting these offset voltages to different values.

You can run and show me a ThrottleStop log file but the problem is your CPU temperature. The ThrottleStop under volt appears to be working OK. ThrottleStop will not fix the problem if your heatsink is not working properly.

Edit - Here is a comparison. It has 4 cores so I disabled turbo boost. When running the TS Bench test, power consumption is around 28 Watts. The cooling system in this Lenovo laptop keeps the CPU package temperature at 61°C. At 28 Watts, your CPU would probably be thermal throttling at 85°C. You need to improve cooling.

The NotebookCheck review of your laptop talks about severe throttling problems. It was poorly designed.

Asus N751JK-T4144H Notebook Review

In-depth review of the Asus N751JK-T4144H (Intel Core i7 4710HQ, Nvidia GeForce GTX 850M, 17.3", 3.4 kg) with numerous measurements, benchmarks and evaluations

www.notebookcheck.net

the CPU quickly reaches temperatures above 80 °C (~176 °F) and then throttles the Turbo clock to a minimum value of 2.4 GHz. The clock of the GeForce GTX 850M drops significantly quite quickly and runs with only 135 MHz, so gaming is hardly possible.

 

[Bloodsucker]

Thanks for the quick response,
I changed the thermal paste several times, but this does not help, can you advise a good thermal paste?

Your CPU is running too hot, even when lightly loaded.

I made the first screenshot of how I turned on the computer after 5-10 minutes, no load

On the 4th Generation processors, the CPU core and CPU cache voltages do not have to be set equally. You can try adjusting these offset voltages to different values.

Tell me, with what parameters can you safely start?
change only CPU Core, and CPU Cache, Intel GPU do not use?

 

[unclewebb]

This Asus laptop has bad cooling. Your log file shows that when your CPU is running at 13 or 14 Watts of power consumption, your CPU is already overheating and reaching the thermal throttling temperature of 85°C. That is terrible. ThrottleStop cannot fix a broken or poorly designed heatsink and fan.

You have to find a way to improve cooling. Is the heatsink tight to the CPU? Some people use tricks so the heatsink makes better contact with the CPU. Is the thermal paste covering the CPU when you check it?

Here is an example of no load. Look at the C0% number. When your computer is idle, there is lots of load. Something is running in the background. It is not the main problem but it is not helpiing the problem.

 

[Bloodsucker]

This Asus laptop has bad cooling. Your log file shows that when your CPU is running at 13 or 14 Watts of power consumption, your CPU is already overheating and reaching the thermal throttling temperature of 85°C. That is terrible. ThrottleStop cannot fix a broken or poorly designed heatsink and fan.

what are the reference values? what do i understand! The cooler is working, the radiator is pressed on all 4 bolts, if this laptop was originally designed poorly, then I vryatli will solve my problem.
what can you advise? good thermo paste? the heat dissipation of 1.85 W / m.k is apparently small for this computer?
made a screenshot when idle, the programs are not open. 4% is a lot?

 

[unclewebb]

The cooler is working

Your log file shows that your cooler is not OK. During this 1 minute, your CPU is constantly thermal throttling. Power consumption is only 13.5 Watts. The 4200H has a 47 Watt TDP rating. That means the heatsink should be able to handle 47 Watts of power consumption without the CPU overheating. My Lenovo laptop has a 47 Watt processor and it can run 24 hours a day at 47 Watts without overheating. That is how a properly designed heatsink and fan should work.

It is impossible for me to know what the problem is. I do not know how well your heatsink worked when your laptop was brand new. Different thermal paste is not going to fix this problem. I cannot imagine that Asus would use a heatsink that is so completely inadequate. The Notebook Review mentioned some problems but your problems are worse than that. Perhaps the copper tubes in your heatsink are damaged or leaking.

You are going to have trouble unless you can find a way to significantly improve your cooling.

   DATE       TIME    MULTI   C0%   CKMOD  CHIPM   BAT_mW  TEMP  GPU     VID   POWER
2020-03-27  21:14:00  28.00   72.7  100.0  100.0        0   83    79   0.7333   13.7
2020-03-27  21:14:01  28.00   70.9  100.0  100.0        0   84    79   0.7333   13.6
2020-03-27  21:14:02  28.00   74.2  100.0  100.0        0   83    79   0.7339   13.6
2020-03-27  21:14:03  28.00   66.0  100.0  100.0        0   83    79   0.7339   13.3
2020-03-27  21:14:04  28.00   73.9  100.0  100.0        0   83    79   0.7339   13.7
2020-03-27  21:14:05  28.00   73.8  100.0  100.0        0   84    79   0.7333   13.8
2020-03-27  21:14:06  28.00   67.0  100.0  100.0        0   82    79   0.7339   13.3
2020-03-27  21:14:07  28.00   62.7  100.0  100.0        0   83    79   0.7333   13.1
2020-03-27  21:14:08  28.00   62.3  100.0  100.0        0   83    79   0.7339   13.0
2020-03-27  21:14:09  28.00   63.1  100.0  100.0        0   82    79   0.7339   13.2
2020-03-27  21:14:10  28.00   64.0  100.0  100.0        0   82    79   0.7339   13.2
2020-03-27  21:14:11  28.00   63.0  100.0  100.0        0   84    79   0.7339   13.0
2020-03-27  21:14:12  28.00   61.8  100.0  100.0        0   82    79   0.7333   12.9
2020-03-27  21:14:13  28.00   64.6  100.0  100.0        0   83    79   0.7339   13.1
2020-03-27  21:14:14  28.00   65.3  100.0  100.0        0   82    78   0.7333   13.1
2020-03-27  21:14:15  28.00   69.3  100.0  100.0        0   82    78   0.7333   13.2
2020-03-27  21:14:16  28.00   61.5  100.0  100.0        0   81    78   0.7339   12.7
2020-03-27  21:14:17  28.00   61.8  100.0  100.0        0   82    79   0.7339   12.8
2020-03-27  21:14:18  28.00   61.7  100.0  100.0        0   82    79   0.7333   12.8
2020-03-27  21:14:19  28.00   60.5  100.0  100.0        0   82    79   0.7333   12.7
2020-03-27  21:14:20  28.00   60.6  100.0  100.0        0   82    79   0.7333   12.8
2020-03-27  21:14:21  28.00   63.2  100.0  100.0        0   82    79   0.7333   12.9
2020-03-27  21:14:22  28.00   63.0  100.0  100.0        0   82    79   0.7333   13.0
2020-03-27  21:14:23  28.00   68.3  100.0  100.0        0   83    79   0.7339   13.2
2020-03-27  21:14:24  28.00   70.8  100.0  100.0        0   83    79   0.7333   13.3
2020-03-27  21:14:25  28.00   66.8  100.0  100.0        0   82    79   0.7333   13.2
2020-03-27  21:14:26  28.00   61.7  100.0  100.0        0   83    79   0.7333   12.9
2020-03-27  21:14:27  28.00   59.7  100.0  100.0        0   83    79   0.7333   12.9
2020-03-27  21:14:28  28.00   63.7  100.0  100.0        0   82    79   0.7333   13.0
2020-03-27  21:14:29  28.00   64.1  100.0  100.0        0   83    79   0.7333   13.0
2020-03-27  21:14:30  28.00   66.7  100.0  100.0        0   83    79   0.7339   13.2
2020-03-27  21:14:31  28.00   62.0  100.0  100.0        0   83    79   0.7339   13.0
2020-03-27  21:14:32  28.00   63.0  100.0  100.0        0   82    79   0.7333   13.0
2020-03-27  21:14:33  28.00   72.6  100.0  100.0        0   83    79   0.7339   13.4
2020-03-27  21:14:34  28.00   76.4  100.0  100.0        0   82    79   0.7339   13.5
2020-03-27  21:14:35  28.00   65.5  100.0  100.0        0   82    79   0.7333   13.1
2020-03-27  21:14:36  28.00   64.1  100.0  100.0        0   82    79   0.7339   13.1
2020-03-27  21:14:37  28.00   62.9  100.0  100.0        0   83    79   0.7339   13.0
2020-03-27  21:14:38  28.00   73.4  100.0  100.0        0   84    79   0.7333   13.5
2020-03-27  21:14:39  28.00   73.2  100.0  100.0        0   83    79   0.7339   13.5
2020-03-27  21:14:40  28.00   70.9  100.0  100.0        0   83    79   0.7333   13.4
2020-03-27  21:14:41  28.00   70.8  100.0  100.0        0   83    79   0.7333   13.4
2020-03-27  21:14:42  28.00   72.9  100.0  100.0        0   83    79   0.7339   13.5
2020-03-27  21:14:43  27.92   79.0  100.0  100.0        0   83    79   0.7333   13.7
2020-03-27  21:14:44  28.00   71.3  100.0  100.0        0   82    78   0.7333   13.4
2020-03-27  21:14:45  28.00   72.6  100.0  100.0        0   83    78   0.7339   13.4
2020-03-27  21:14:46  28.00   73.0  100.0  100.0        0   83    78   0.7333   13.5
2020-03-27  21:14:47  28.00   68.6  100.0  100.0        0   82    79   0.7339   13.5
2020-03-27  21:14:48  28.00   75.0  100.0  100.0        0   83    79   0.7333   13.6
2020-03-27  21:14:49  28.00   73.5  100.0  100.0        0   83    79   0.7339   13.6
2020-03-27  21:14:50  28.00   73.5  100.0  100.0        0   83    79   0.7333   13.5
2020-03-27  21:14:51  28.00   71.5  100.0  100.0        0   83    79   0.7339   13.6
2020-03-27  21:14:52  28.00   69.1  100.0  100.0        0   83    79   0.7333   13.5
2020-03-27  21:14:53  28.00   70.2  100.0  100.0        0   84    79   0.7339   13.5
2020-03-27  21:14:54  28.00   72.6  100.0  100.0        0   83    79   0.7333   13.7
2020-03-27  21:14:55  28.00   73.3  100.0  100.0        0   83    79   0.7339   13.6
2020-03-27  21:14:56  28.00   70.2  100.0  100.0        0   83    79   0.7333   13.6
2020-03-27  21:14:57  28.00   69.9  100.0  100.0        0   82    79   0.7333   13.5
2020-03-27  21:14:58  28.00   69.7  100.0  100.0        0   84    79   0.7339   13.4
2020-03-27  21:14:59  28.00   68.9  100.0  100.0        0   84    79   0.7333   13.4

 

[Bloodsucker]

Sorry,
I bought it not new, in
replacement the MSI FX600. I have been using asus for 3 years.

I correctly understood that my 85 degrees = 100 according to the specification. and should be achieved at TDP 47 or higher?
if the processor has already overheated, if the temperature is still at 85? in relation to power consumption and current temperature?

I will take apart to see what is wrong with him. Undervolting, as I understand it, will not help me yet.

 

[unclewebb]

Under volting will help but you need more than just under volting to solve this problem.

The design of your heatsink does not look good. When you are gaming and the Nvidia GPU is putting out heat, the combined heat causes both the CPU and GPU to run too hot. The heatsink and fan cannot keep up to the amount of heat being created.

When Asus installs a 47 Watt CPU into a laptop, this CPU should be able to run at its full rated speed including turbo boost without the CPU overheating or thermal throttling. I think Asus realized that there was a big problem with their design. Instead of fixing the problem, they lowered the thermal throttling temperature from 100°C down to 85°C. That did not solve the problem. It just covered up the problem so less people would call them to complain about their overheating Asus laptop.

 

[Bloodsucker]

Under volting will help but you need more than just under volting to solve this problem.

What values should I start with? and what proportions should be observed in the cache?

What numbers can I count on at least Lenovo?
My asus is ideally 47 Watt CPU = 85°C мах ?

 

[unclewebb]

For the 4th Gen 4200H, you might not be able to undervolt much more than what you are doing right now. If you decide to go beyond -110 mV, do lots of testing to make sure your undervolt is stable. Test the core and cache offset voltages individually. See how far you can go with each one before losing stability.

Intel recommends that a heatsink and fan should be able to keep the CPU temperature under 100°C when it is running at its rated TDP. That is 47 Watts for the 4200H. Even when your laptop was brand new, I do not think that it was anywhere near that target, especially with an Nvidia GPU also creating heat in the system.

 

[Bloodsucker]

Hi, unclewebb.
I thought a lot about what you said about the malfunctions of the cooling system, and decided to conduct an experiment, I completely removed the bottom cover and started playing.
The temperature of my processor was around 76-77 degrees, and this is an achievement of 10 degrees.
I also found an interesting program NoteBook FanControl, and with it I also lowered the temperature with the bottom cover removed, to a relatively comfortable 71-72 degrees in games.
I wrote Throttlestop logs, though some time ago.

 

[unclewebb]

@Bloodsucker - Your log files show a problem. the two columns CKMOD and CHIPM are two different types of clock modulation throttling. Basically the CPU is throttling internally. These two columns in the log file should always show 100.0. Both columns are at 87.5. Did you do this deliberately? I cannot remember seeing a laptop that used both these types of throttling at the same time. Your previous log file did not show this.

To try and fix this problem, on the main screen of ThrottleStop, put a check mark in the Clock Modulation box and put a check mark in the Chipset Modulation box and make sure both of these are set to 100.0%. Click on the Turn On button if you need to. Make sure ThrottleStop does not show Monitoring in the title bar. You want ThrottleStop to be actively working on this throttling problem.

Some of the improvement in temperatures that you observed might be because your CPU is throttling inside.

 

[Bloodsucker]

@unclewebb - I read various tips on how to deal with temperature, one of them was lowering these parameters. returned both parameters to 100%.
are 71-72 degrees in games not good, for my laptop on this chipset?
Also, I did not quite understand how correctly undervolting i5 4200h through Throttlestop, so my parameters are default.

 

[unclewebb]

are 71-72 degrees in games not good

Product Specifications

quick reference guide including specifications, features, pricing, compatibility, design documentation, ordering codes, spec codes and more.

ark.intel.com

The Intel specified thermal throttling temperature for the 4200H is 100°C. According to Intel, any temperature lower than that is a "safe operating temperature". Asus decided to ignore the Intel specifications. They locked the thermal throttling temperature to 85°C when there was no reason to do this.

When your CPU reaches 85°C, it is not overheating. It is thermal throttling because of a bad design decision by Asus.

I would not use the Disable Turbo option. Let your CPU run at full speed. The reduced thermal throttling temperature has your CPU well protected. It is over protected.

The worst thing that can happen is your CPU will reach 85°C. Will it blow up at this temperature? Of course not. It will thermal throttle which means it will slow down just enough so it does not exceed this temperature. There is no reason to be concerned if a CPU starts thermal throttling at this low of a temperature.

I did not quite understand how correctly undervolting i5 4200h through Throttlestop, so my parameters are default.

Your original screenshots show that you were under volting the CPU core and CPU cache by -110 mV. What do you not understand? Reducing the voltage reduces heat. As long as your CPU is stable, you should undervolt it. Asus is forcing you to run your CPU at temperatures far under specification so you need all the help you can get.

 

[Bloodsucker]

Your original screenshots show that you were under volting the CPU core and CPU cache by -110 mV. What do you not understand? Reducing the voltage reduces heat. As long as your CPU is stable, you should undervolt it. Asus is forcing you to run your CPU at temperatures far under specification so you need all the help you can get.

yes, but it didn’t help, he warmed up to 86 degrees and left and trolling, and as a result I defaulted everywhere, i.e. 0
if set CPU core and CPU cache by -110 mV. Do you think the temperature will be lower than 71-72 degrees?

 

[Zielek]

The problem is the GPU temperature. Above 80 degrees, the GPU slows down. Lowering the CPU temperature will positively affect GPU performance.
The CPU with Turbo boost always tends to the maximum allowable TDP by increasing the frequency. If you reduced its voltage, you also reduced the TDP. The CPU frequency will be automatically increased until the maximum TDP or maximum turbo boost frequency is reached. To reduce the temperature with ThrottleStop, set Long power max to 10W and Short Power Max to 11W . That's what comes from your logs. The processor does not exceed 15W and together with the GPU causes a temperature rise to 85 degrees. 47W with the current heat sink is not available.

85 degrees was set for the CPU to preserve the graphics card temperature. Undervolting caused a slight increase in processor performance without raising the target temperature (85 degrees).

If you reduce the processor's TDP (Long power max and Short Power Max), its performance will be lower only in certain situations, but thanks to this you will give the graphics card a greater heat reserve. This will allow the graphics card to work without temporarily reducing the frequency. This should solve the problem of a sudden drop in FPS. Uncheck the Turbo boost option in ThrottleStop.
TurboBoost will automatically ensure full processor performance so that it does not exceed the TDP you set. Theoretically, this way the temperature will also reach 85 degrees, but this time the graphics card will cause it. CPU and GPU cooling are common in this laptop. GPU performance is more important in games.
There is nothing to worry about until the temperature exceeds 95 degrees. Thermal throttling just means that the processor must reduce the frequency to avoid overheating. When overheating, the computer will turn itself off.

 

[unclewebb]

Do you think the temperature will be lower than 71-72 degrees?

Do some testing and find out. There is no reason to run a laptop CPU at only 71°C or 72°C. You will likely have to run your CPU slower than its rated speed and sacrifice performance to get your CPU to run that cool.

 

[Bloodsucker]

HI, Zielek.
if you remove the daw from the turbo disable, then the processor again heats up to 85, and rests on trolling. and this is with the bottom cover removed. if close the lid, will be very hot the case, as originally, it is very bad for the components, the keyboard is not pleasant to heat up for hands. I do not feel discomfort in the game with the turbo turned off, the heat is more annoying.
also I can not set set Long power max to 10W and Short Power Max to 11W, these parameters are locked.
take a look at the logs today, I played with the turbo jackdaw removed.

 

[Zielek]

With the cover removed, the parameters are fine, unfortunately this is a quick hand for accidental mechanical damage to the fan. The performance of the cooling used in this laptop is very small, so the only thing you can do is reduce the power generated by the components.

Sorry, I didn't notice that the parameters are locked. In this situation, you can select and change the PP0 power limit value to 7 and be sure to select the clamp option. See screenshot 1
This should limit the power (TDP) to around 11.5W - the frequency will be set automatically to maintain power at this level. See screenshot 2 to screenshot 4. At different loads, the processor frequency changes but the power dissipated remains constant.
The solution is not perfect, but nothing better cannot be done.

The CORE POWER sensor should be marked in red in the limit reasons window. This means that the limit set using PP0 power limit works.
Also pay attention to the BD PROCHOT sensor. If it is red, it means that the graphics card has sent a "false" overheating signal to the processor to force it to decrease frequency. If you uncheck the bd prochot option in the main program window, then the processor's performance will take priority and the graphics card will slow down first. Better leave this option selected.

In the case of negligible effects, you can reduce the value of the PP0 parameter to 4. If the value of 7 works, you can gradually increase it to get the highest CPU performance at the temperature you want to get. Remember that you must check the CLAMP option.

On screenshoot 5 you can see how it looks like with my laptop heating up at full CPU load.
Screenshoot 6 shows the temperatures at full load of the processor and graphics card. Such high loads in games are very rare.
In my laptop, the thermal limit was set high at 98 degrees, thanks to which the processor can operate at a higher frequency, but the graphics card slows down as a result of overheating. I think all laptops have a similar problem.

The laptop from which the screenshots are Asus R510jk i5-4200H, GTX 850M 2GB, 12 GB RAM, SSD 480GB. It also has one common heat sink for the processor and the graphics card.
I used it for over 8200 Hours (Power On count parameter for hard disk in S.M.A.R.T). Of this, over 4000 hours worked under full load of the processor and graphics card (Games and video compilation using GPU). It is still functional today - so there is no reason to worry about high temperatures.
The hot keyboard problem can be solved by connecting an external keyboard to USB.

I think Unclewebb is right in saying that it's not worth sacrificing performance to lower the temperature.

If you find the right value for PP0 then also use undervolting - it will help a little to increase processor performance. At a value of 7, processor performance will drop to about 60%.
In my laptop, the processor never exceeds the TDP 25W value whatever I would do. Possible that Asus lowered the TDP value to solve the problem of poor quality of the cooling system.
I suspect your laptop's TDP limit is 15W.
It's easy to check. Turn on prome95 for a few minutes with 4 threads and observe the Power parameter. Probably will not exceed the value 15. Of course, all throttleStop settings should be restored to factory defaults before the test.
Screenshot 8 show tempeartures with limit the power using PP0 value of 7. Temperature significantly dropped, with the same graphics card performance. However, CPU performance has dropped significantly (Screenshot7 versus Screenshot6).

To sum up, your laptop is fine, just the manufacturer has used too poor cooling to simultaneously load the CPU and GPU. There is no risk of damage, but there will be performance drops no matter what you do.

 

[Bloodsucker]

The removed bottom cover is a temporary solution, I will buy the same cover for experiments, I will try to supply more air to the fan through the new holes that I want to drill. The cooler is not damaged, the laptop is on rubber legs. Screenshot1
I set a limit of PP0 value of 7 as you said, but nothing has changed, I even set a limit of PP0 value of 6, but also did not notice the difference.
In the game I play, it consumes up to TDP 15 W, with the turbo turned off, it accelerates to TDP 22 W.
Once again I want to say, I have enough performance, even with a consumption of up to 15, but when it heated to 86 it greatly upset me and fry my hands.
I’m rather experiencing a limit on the video card in my game, although I play on medium settings, she’s 100% all the time during the game Screenshot2
It bothers me, a very old driver for my processor 10.0.18362.387 dated 04/21/2009, do you have the same or fresh?

I wanted to push it to the maximum; he doesn’t want to, what am I doing wrong? Screenshot2020-05-19

 

[Zielek]

OCCT shows that the IA Cores parameter does not exceed 7. It seems that the processor cores are already running at low power (7W), while the other processor components consume the rest of the energy.
Try setting PP0 Power Limit to 2. Then the IA Cores parameter in OCCT should not exceed 2W. If it still has a value of 7W, it means that in your laptop model the manufacturer has set the PP0 register in hardware. If the manual setting of this parameter will be automatically overwritten by the factory , nothing more can be done.
You don't have to worry about the CPU driver version. My version is 10.0.18362.693 dated 04/21/2009
Attached is a photo of my version of the CPU driver. The operating system is Windows 10 Home.
It is also possible that some Asus driver / program, e.g. ATK Package or Power4Gear, will block the PP0 CPU register. I never install these applications.
The PP0 register only limits the power consumption of the processor cores. The PKG Power parameter in ThrottleStop shows the total power consumption of the entire processor (Package power), i.e. cores, bridge and integrated graphics card (iGPU).

In the case of Screenshot2020-05-19, you forgot to run the Prime95 program - it loads the processor 100%. Furmark only charges the graphics card and minimally the processor.
Instead of Prime95 you can run the CPU burner application (CPU Burner button in the Furmark window).

Maybe you should consider using vacuum cooler instead of experimenting with the bottom cover.
Here is a test of such a device (Watch from 1:41):