• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Help with overheating CPU (notebook)

tobythecat

New Member
Joined
Nov 18, 2021
Messages
22 (0.02/day)
Hey, I have an acer laptop meant for gaming and I am noticing some high temperatures even while idle or only opening a browser. It reaches around 67C as of writing this at the moment, and I only have a tab open and steam. I know laptops have really bad cooling but I'm convinced that I can at least maybe do something to at least help my temperatures. I feel it is important to note that I am using the software for Acer Predator to select the "Extreme" option which I'm assuming keeps the clock speed pretty high at all times so that might be why my idle temps are pretty high.

My specifications are as follows:

Predator PH315-53 (model)
Rtx 3060 (notebook)
I7-11800H
32GB 3200MHZ
three drives, 2 m2. ssds and a sata.

I'll post a log of throttlestop of just normal stress of playing a game with an uncapped frame rate. Also, I will include a screenshot of the "limits" windows as there is a concerning amount of yellow squares showing up. (just an fyi it is blinking red on PL2, and i just happened to screenshot it was it turned red for literally 10ms or so.
 

Attachments

  • 2022-12-11.txt
    2022-12-11.txt
    111.7 KB · Views: 90
  • Screenshot 2022-12-11 024007.png
    Screenshot 2022-12-11 024007.png
    57.3 KB · Views: 185
is the laptop still under warranty? if it is send it back to acer
 
is it that bad? note that the laptop is running on the max clock since im using the overclocking features built into the laptop.
Laptop have limited cooling capabilities, my msi gaming laptop fans spin up during charging and watching YouTube lol, maybe don’t overclock it
 
For how long do you have it? Some repasting (if you get this temperature from beginning - the paste was probably applied incorectly or low quality) and cleaning the fans can help - "a bit (or a lot - if you use it in a dusty environment - and they're clogged) ". But don't expect miracles - "for a laptop". The cooling design of your model - usually makes the biggest difference "among laptops (no room for comparison with a desktop - even the thought of that is laughable - to say the least)". That being said... if what i mentioned above is all checked - then (and only then) you can move to software solutions - like Your windows Power Plan or/and a 3rd party tool like ThrottleStop.

Based on the screens you attached - your laptop was running dangerously HOT - seeing the PROCHOT got triggered at 100*C. Instead of OC-ing might help going the opposite direction - like lowering your Max Turbo Boost. But first, you can try some undervolting - even limiting the power wattage - to see how much can that lower the temps.
 
is it that bad?
It is not that bad at all. Modern laptops run hot.

dangerously HOT
Most Intel CPUs set the shut down temperature to 125°C. Thermal throttling starts at 100°C to protect against any long term damage. The log file shows that the CPU only reached 100°C two different times for 1 second. That is not dangerous at all. It is completely normal. Most recent laptops spend more time than this bouncing off the thermal throttling limit.

@tobythecat - If you have a question or want any recommendations, post some screenshots of the entire main ThrottleStop window, the FIVR window and the TPL window. From what I can see, I would clear the BD PROCHOT box on the main ThrottleStop screen, I would check the MMIO Lock box near the top right of the TPL window and I would set Power Limit 4 in this window to a value of 0. Without seeing some more screenshots, I am not sure what other adjustments are unlocked and available to you.

The maximum core temperature appears to vary quite a bit from core to core. Replacing the thermal paste and having a good look to see if the heatsink is sitting square to the CPU and is snug would be a good idea. If performance is good, you do not have to do anything today. Go play a game and enjoy your hot running gaming laptop. I am sure that Acer already knows that their Predator laptops run hot.
 
It is not that bad at all. Modern laptops run hot.


Most Intel CPUs set the shut down temperature to 125°C. Thermal throttling starts at 100°C to protect against any long term damage. The log file shows that the CPU only reached 100°C two different times for 1 second. That is not dangerous at all. It is completely normal. Most recent laptops spend more time than this bouncing off the thermal throttling limit.

@tobythecat - If you have a question or want any recommendations, post some screenshots of the entire main ThrottleStop window, the FIVR window and the TPL window. From what I can see, I would clear the BD PROCHOT box on the main ThrottleStop screen, I would check the MMIO Lock box near the top right of the TPL window and I would set Power Limit 4 in this window to a value of 0. Without seeing some more screenshots, I am not sure what other adjustments are unlocked and available to you.

The maximum core temperature appears to vary quite a bit from core to core. Replacing the thermal paste and having a good look to see if the heatsink is sitting square to the CPU and is snug would be a good idea. If performance is good, you do not have to do anything today. Go play a game and enjoy your hot running gaming laptop. I am sure that Acer already knows that their Predator laptops run hot.
Alright, I’m resetting the pc atm (long overdue) so when I reinstall throttlestop it’ll be a clean install with default settings
 
It is not that bad at all. Modern laptops run hot.

Most Intel CPUs set the shut down temperature to 125°C. Thermal throttling starts at 100°C to protect against any long term damage. The log file shows that the CPU only reached 100°C two different times for 1 second. That is not dangerous at all. It is completely normal. Most recent laptops spend more time than this bouncing off the thermal throttling limit.

I get your point about the CPU (unlikely to get burned - with all the modern fail-safe features) - but It's still a "laptop". And when it comes to gaming laptops - over-heating is quite a common suspect for motherboard failure (or at least more common than dust - where usually it takes quite a lot of negligence, followed by physical damage or spills and rarely an electrical surge/spike - even tho - every sudden occurrence seems to point at that). Quite ironic for the CPU to survive its own heat - while damaging the surrounding components (if running at 67*C in idle and climbing to such high temps under load - tho, i can agree - that it depends how long it hovers close to that level of heat - if it's just a random spike or a common occurrence/overheating). On the other hand - that's far far far less likely to happen on a Desktop, as in... bigger (or humongous) Heat-sink on CPU, more room/space aka - incomparably better airflow and last but not the least "more solid motherboards (that's actually "a shinny component" in a desktop - not like laptops where good enough will do - mainly focused on a shinny exterior & the specs for most recent released components/technology)".
 
I get your point about the CPU (unlikely to get burned - with all the modern fail-safe features) - but It's still a "laptop". And when it comes to gaming laptops - over-heating is quite a common suspect for motherboard failure (or at least more common than dust - where usually it takes quite a lot of negligence, followed by physical damage or spills and rarely an electrical surge/spike - even tho - every sudden occurrence seems to point at that). Quite ironic for the CPU to survive its own heat - while damaging the surrounding components (if running at 67*C in idle and climbing to such high temps under load - tho, i can agree - that it depends how long it hovers close to that level of heat - if it's just a random spike or a common occurrence/overheating). On the other hand - that's far far far less likely to happen on a Desktop, as in... bigger (or humongous) Heat-sink on CPU, more room/space aka - incomparably better airflow and last but not the least "more solid motherboards (that's actually "a shinny component" in a desktop - not like laptops where good enough will do - mainly focused on a shinny exterior & the specs for most recent released components/technology)".
Yeah, I carefully use my laptop, I use it quite frequently, but I make sure never to eat or drink while using it, I always wash my hands and dry my hands before I use it. I never dropped it banged it. I just wanted to see if I can get some more performance by using throttlestop to negative some of the heat
 
i don't see any problem tbh.
modern laptops are so downcosted and cheap they almost thermal throttle while reading emails.
 
It is not that bad at all. Modern laptops run hot.


Most Intel CPUs set the shut down temperature to 125°C. Thermal throttling starts at 100°C to protect against any long term damage. The log file shows that the CPU only reached 100°C two different times for 1 second. That is not dangerous at all. It is completely normal. Most recent laptops spend more time than this bouncing off the thermal throttling limit.

@tobythecat - If you have a question or want any recommendations, post some screenshots of the entire main ThrottleStop window, the FIVR window and the TPL window. From what I can see, I would clear the BD PROCHOT box on the main ThrottleStop screen, I would check the MMIO Lock box near the top right of the TPL window and I would set Power Limit 4 in this window to a value of 0. Without seeing some more screenshots, I am not sure what other adjustments are unlocked and available to you.

The maximum core temperature appears to vary quite a bit from core to core. Replacing the thermal paste and having a good look to see if the heatsink is sitting square to the CPU and is snug would be a good idea. If performance is good, you do not have to do anything today. Go play a game and enjoy your hot running gaming laptop. I am sure that Acer already knows that their Predator laptops run hot.
 

Attachments

  • FIVR.png
    FIVR.png
    432.6 KB · Views: 142
  • TPL.png
    TPL.png
    341.9 KB · Views: 139
@tobythecat

Your FIVR screenshot shows that you have Windows 11 Virtualization Based Security enabled. This prevents ThrottleStop from working correctly. ThrottleStop needs direct access to the CPU voltage control register.

If you need to use WSL2, you will need to leave VBS enabled. If you do not require this enhanced security then consider disabling VBS including Core Isolation Memory Integrity. This can improve gaming performance. You have to decide what is important to you.


After disabling VBS, reboot and delete the ThrottleStop.INI configuration file before running ThrottleStop. Post an update screenshot of the FIVR window after you do that.
 
Yamwwac.

Yet another mobile with weak ass cooling

Under warranty, return it, otherwise find a how to vid to replace all thermal pads and pastes, increase fan speed to max if you can and get a cooling pad, otherwise look at desktop replacement eurocom or sager/sagem.

Or build a Desktop and keep the mobile for light use only.

Look up the Dell Inspiron 9100/XPS Gen 1, those are robust laptops that never overheated with the Prescott or Gallatin P4
 
Alright, atm I’m re applying paste and cleaning it, and look what I found haha.
goodness, how long have you had your pc? the specs are identical to mine cept the gpu
 
goodness, how long have you had your pc? the specs are identical to mine cept the gpu
I've had mine for like around half a year, i got it from amazon since it has an outdated CPU and it was pretty cheap for the specs in it. It was brand new too. It's strange, I kept opening it up once a month when i first got it and there was no dust, and just now I opened it up to replace the paste since the factory paste was pretty bad. And there was a giant clump of debris which is not surprising and surprising at the same time. Like i have two pets that shed a lot, a cat and dog, but I never had a dirty fan until now when I moved into my dorm.

@tobythecat

Your FIVR screenshot shows that you have Windows 11 Virtualization Based Security enabled. This prevents ThrottleStop from working correctly. ThrottleStop needs direct access to the CPU voltage control register.

If you need to use WSL2, you will need to leave VBS enabled. If you do not require this enhanced security then consider disabling VBS including Core Isolation Memory Integrity. This can improve gaming performance. You have to decide what is important to you.


After disabling VBS, reboot and delete the ThrottleStop.INI configuration file before running ThrottleStop. Post an update screenshot of the FIVR window after you do that.
had to go on a goose chase to find how to disable VBS, because for some reason it says it wasn't supported on my computer. I had to find a setting in the bios to disable so I could disable it.

goodness, how long have you had your pc? the specs are identical to mine cept the gpu
here’s a picture of the new friend we got, she hasn’t grown to full size yet, but she’s already shedding a bit, lol
 

Attachments

  • updated fivr.png
    updated fivr.png
    417.5 KB · Views: 121
  • updated tpl.png
    updated tpl.png
    491.2 KB · Views: 106
  • VBS.png
    VBS.png
    2 KB · Views: 121
  • image.jpg
    image.jpg
    1.8 MB · Views: 115
@tobythecat

ThrottleStop looks better. You are lucky that your computer supports CPU voltage control and very lucky that it is not locked. Avoid updating the BIOS, ever.

Try setting the core and cache to a negative offset of -50 mV. Do some testing after that to see if your computer is stable. The TS Bench is a quick and handy test. A quick 960M run is a good place to start testing. If you see any errors reported during this test then your CPU needs more voltage. Play some games too and run Cinebench. Check for improved performance or reduced temperatures. Not sure about 11th Gen but many 10th Gen CPUs were usually stable with an undervolt of -70 mV or -75 mV.

Have a look in the Options window for the PROCHOT Offset setting on the right hand side. This is set to 8 which tells your CPU to start thermal throttling a little early at 92°C instead of the full Intel recommended 100°C. If you do not see a lock icon near this setting, it is OK to lower the offset value a few degrees. This will increase your thermal throttling temperature. If this is locked by the BIOS then leave it alone because it is locked and cannot be adjusted.

Did getting the hair balls out improve your temperatures? Keep a close eye on things. Many thermal pastes can quickly degrade if you operate your CPU at a high temperature on a consistent and regular basis.

Edit - I would set IccMax for the core and the cache to the maximum, 255.75 to help avoid any EDP throttling.

Your CPU supports limited overclocking which means you can bump each of the turbo ratios up by +4. Most users run into cooling issues so are not able to take advantage of this feature. You need to check the Overclock box in the FIVR window if you want to see what she can do.
 
@tobythecat

ThrottleStop looks better. You are lucky that your computer supports CPU voltage control and very lucky that it is not locked. Avoid updating the BIOS, ever.

Try setting the core and cache to a negative offset of -50 mV. Do some testing after that to see if your computer is stable. The TS Bench is a quick and handy test. A quick 960M run is a good place to start testing. If you see any errors reported during this test then your CPU needs more voltage. Play some games too and run Cinebench. Check for improved performance or reduced temperatures. Not sure about 11th Gen but many 10th Gen CPUs were usually stable with an undervolt of -70 mV or -75 mV.

Have a look in the Options window for the PROCHOT Offset setting on the right hand side. This is set to 8 which tells your CPU to start thermal throttling a little early at 92°C instead of the full Intel recommended 100°C. If you do not see a lock icon near this setting, it is OK to lower the offset value a few degrees. This will increase your thermal throttling temperature. If this is locked by the BIOS then leave it alone because it is locked and cannot be adjusted.

Did getting the hair balls out improve your temperatures? Keep a close eye on things. Many thermal pastes can quickly degrade if you operate your CPU at a high temperature on a consistent and regular basis.

Edit - I would set IccMax for the core and the cache to the maximum, 255.75 to help avoid any EDP throttling.

Your CPU supports limited overclocking which means you can bump each of the turbo ratios up by +4. Most users run into cooling issues so are not able to take advantage of this feature. You need to check the Overclock box in the FIVR window if you want to see what she can do.
Just a few questions, I checked the overclock button, how do I change the turbo ratios, they are grayed out still, does this mean I can’t edit them?

also in the limits tab I’m hitting PL2 on my gpu and cpu, my gpu being yellow and cpu being mainly yellow and sometimes blinks red for a second. I didn’t edit any thing in the TPL menu so it’s still the same as the screenshot.
 
I haven't read all of this yet, but going to summarise the best things you can do

1. Throttlestop - work with unclewebb there, he's the genius for that
2. Undervolt the GPU. My 3090 drops from 375W to 220W depending on the game, and that should say it all.
3. Use an FPS cap. Even people on desktops should do that, you get better render and latency times from it anyway - unlimited FPS is for PC's that are stuck under 60FPS and for benchmarking, and are terrible otherwise.

You can cap to 90FPS and draw a theoretical 200W, but lower some fancy graphics setting (or enable DLSS) and that 90FPS suddenly uses 150W - but with an uncapped FPS you'd just get a higher frame rate and go right back to that 200W max


That acer software on extreme is likely disabling any and all power saving and keeping all cores active, which can actually hurt performance - cores need to power down and turn off, so their wattage can be used where it's needed (Be it other cores boosting higher, or even to the GPU on a power limited laptop)



Oh and uhhh... was that a wad of hair, or a hairtie or something? That didn't look too friendly.
 
I haven't read all of this yet, but going to summarise the best things you can do

1. Throttlestop - work with unclewebb there, he's the genius for that
2. Undervolt the GPU. My 3090 drops from 375W to 220W depending on the game, and that should say it all.
3. Use an FPS cap. Even people on desktops should do that, you get better render and latency times from it anyway - unlimited FPS is for PC's that are stuck under 60FPS and for benchmarking, and are terrible otherwise.
Yeah, I make sure to put on Adaptive V-sync, as well an FPS cap at 144. I don't even get that high on demanding games on higher graphics settings, but still I always do it.
You can cap to 90FPS and draw a theoretical 200W, but lower some fancy graphics setting (or enable DLSS) and that 90FPS suddenly uses 150W - but with an uncapped FPS you'd just get a higher frame rate and go right back to that 200W max
I'm going to try to see if I can undervolt my GPU next, since my GPU's max temperature is like 87, and it might hit that under extended periods of time.
That acer software on extreme is likely disabling any and all power saving and keeping all cores active, which can actually hurt performance - cores need to power down and turn off, so their wattage can be used where it's needed (Be it other cores boosting higher, or even to the GPU on a power limited lapto
So essentially there is 4 settings, quiet (as if it was running on battery), normal (nothing special), extreme ( the one I use ), and turbo (which is the maximum and keeps the clocks like maximum and always on)

I never use Turbo, like ever. Extreme is fine for me now since my idle Temperatures have imporved. Normal would probably be best for keeping my pc healthy for the long term, but for whatever reasons acer made normal cap your fan speed at a certain RPM so I never use it while gaming.
Oh and uhhh... was that a wad of hair, or a hairtie or something? That didn't look too friendly.
Looks like a hairtie, right? It's a big collection of random debris that took the shape of the fan and made a circle lol. I threw it away already but if I had to assume, it was a combination of pet hair and my college dorm being really dirty. Not because of me, just from the last people to use it. The dorm is very dusty and filled with weird debris that my roommates and I found in each of our rooms. I never had an issue with a dirty fan until I moved into the dorm So i'm forced to believe that my fans somehow sucked it up and its been just sitting there ever since waiting to be removed.
 
Replace SATA HDD with SATA SSD

Open laptop undercover for better airflow

Get a laptop fan
 
Edited with better info and less words

Undervolting the GPU can and will lead to massive drops in temperature and power consumption
Use geforce experiences Alt-R overlay and set it to show all

Underclock w/ FPS limit vs FPS limit stock vs unlocked stock
1670846454181.png
1670846511566.png

Notice how I got ~100Mhz more, at the same voltage/wattage?

And then unlocked stock with no FPS limit (What you're doing)
1670846586339.png

just ouch. 344W - notice how the GPU clocked up, but then wasnt even being used fully? Wasting power for no real gain.

What you need to see is the maximum clock your GPU sits at when its been at full load for long enough to heat up and hit those limits - mine peaks to 2.1GHz, but then would hit TDP limits and settle down to around 1.8GHz - especially when it was on air cooling


Youtube guides will show this better, but basically you pick a clock speed at or just under that value it sits at, and see how low a voltage is stable.
Done right, you have a flat curve like this
1670845581781.png


It can do 6mv steps, so when you do your testing and find a happy number add at least 12mv extra on top just to be sure it's stable in more demanding games in the future, or as the laptop ages
 
Last edited:
how do I change the turbo ratios
Did you try clicking on the turbo ratio adjusters on the right hand side of the FIVR - Turbo Ratio Limits section?
They are not locked. The values listed on the right hand side are the default maximum values.

1670870914864.png


Try setting all of them to 46. The CPU will run at this speed no matter how many cores are active.
Users call this an all core overclock.
If your new and improved cooling cannot handle the extra heat, there is no need to increase any of them.
 
oh okay, i see what you mean then, also for the limits section flashing showing PL2, is this something I should fix or does that just mean my cpu is using the max power it can use.
 
Run something consistent like Cinebench when testing. What PKG Power consumption does ThrottleStop report on its main screen when PL2 is red or flashing red?

Your screenshot shows that you have the MSR PL2 power limit set to 115W. That should be enough. Are you at this power level or much lower when PL2 throttling is happening? Some laptops set a lower power limit internally which ThrottleStop cannot adjust higher. Turn the Log File option on when testing. It makes it a lot easier to see how your computer is performing over time or while in a game for 15 minutes. It will help show how much, if any, thermal headroom you have.

When finished testing, exit ThrottleStop so it can finalize the log file. Attach a log to your next post. The default log file location is your ThrottleStop / Logs folder.
 
Back
Top