How can I limit my CPU temps?

as for game performence, i still have yet to test it out properly which i probbaly will do over the weekend when i game, ill see how the performence differs.

zqchri said:

i dont think i uavw that much time on my hands and im not allowdd to leave my laptop running overnight, so do you think i should just leave it be and when i get errors i see then?

Click to expand...

You could always decide to divide the longer test sessions up into multiple, shorter test runs. The only reason why you might want to undo the undervolting again each time before you are going to use the laptop for getting important work done (i.e. as opposed to letting the test continue to run) is because else the laptop might crash on you before you stand a chance to save your work. Let's for a moment suppose that it's been running OCCT perfectly stable over the course of, say, 10 hours in total and that you have verified this, not only by checking for BSODs, but also by checking for WHEA recoverable errors. Then what you have is a pretty strong indication that, at that specific voltage offset, at worst it should remain more or less stable in the long run. So if, for example, this voltage offset is 100mV below zero volts, then changing it to 90mV below zero volts is going to be a lot more stable, but might still rarely cause a crash. To avoid this small risk of still causing a crash, you could then decide to change it to 80mV below zero.

Whereas, if it turns out that setting the voltage offset to 100mV below zero is what causes BSODs to occur somewhat more frequently (like, e.g., at least once every few hours on average), that's when you'll know that continuing to run the tests at 100mV below zero is going to be pointless, as this would imply that it just isn't stable enough to still be able to justify running any additional tests at 100mV below zero, and that, next, you would be much better off running your tests at 90mV below zero instead.

hdmi said:

You could always decide to divide the longer test sessions up into multiple, shorter test runs. The only reason why you might want to undo the undervolting again each time before you are going to use the laptop for getting important work done (i.e. as opposed to letting the test continue to run) is because else the laptop might crash on you before you stand a chance to save your work. Let's for a moment suppose that it's been running OCCT perfectly stable over the course of, say, 10 hours in total and that you have verified this, not only by checking for BSODs, but also by checking for WHEA recoverable errors. Then what you have is a pretty strong indication that, at that specific voltage offset, at worst it should remain more or less stable in the long run. So if, for example, this voltage offset is 100mV below zero volts, then changing it to 90mV below zero volts is going to be a lot more stable, but might still rarely cause a crash. To avoid this small risk of still causing a crash, you could then decide to change it to 80mV below zero.

Whereas, if it turns out that setting the voltage offset to 100mV below zero is what causes BSODs to occur somewhat more frequently (like, e.g., at least once every few hours on average), that's when you'll know that continuing to run the tests at 100mV below zero is going to be pointless, as this would imply that it just isn't stable enough to still be able to justify running any additional tests at 100mV below zero, and that, next, you would be much better off running your tests at 90mV below zero instead.

Click to expand...

fair points, how would you think i should schedule my tests then? i dont want to do long occt tests as i think its a bit risky keeping it at such high temps for so long, no? but thank you for the heads up, ill definately report back here with any updates.

one question, i have been getting hard crashes once in a while before all of this undervolting, and if it happens again, how will i rule out what the cause is? i have had a discussion with zbooks and all he needs from me now is a memtester result, which i will need to run when i have the time. but if it goes well, can i assume its because of this undervolting?

also, how would i get to these WHEA recoverable errors?

im currently trying a occt test over 30 minutes, does doing it for long periods of time spoil or harm thr cpu in anyway?

zqchri said:

fair points, how would you think i should schedule my tests then? i dont want to do long occt tests as i think its a bit risky keeping it at such high temps for so long, no? but thank you for the heads up, ill definately report back here with any updates.

one question, i have been getting hard crashes once in a while before all of this undervolting, and if it happens again, how will i rule out what the cause is? i have had a discussion with zbooks and all he needs from me now is a memtester result, which i will need to run when i have the time. but if it goes well, can i assume its because of this undervolting?

Click to expand...

Good question... Personally, I, would go on MSI tech support and open a ticket to get the problems resolved, professionally ASAP. It's the shortest pain. Just be sure to backup all your important data before you send it in for servicing. Also ask them to fix the damaged cooling fan TBH.

zqchri said:

also, how would i get to these WHEA recoverable errors?

im currently trying a occt test over 30 minutes, does doing it for long periods of time spoil or harm thr cpu in anyway?

Click to expand...

See this reddit post by a user named Tetedeiench:

On the subject of gaming laptops running hot. They are specially designed to run hot for prolonged periods of time. My Asus TUF Gaming F16 (2024) laptop actually even uses dust filters on its cooling fans by design. In a modern PC/laptop, both the CPU and GPU will be slowed down dynamically on-the-fly before they can start to overheat. Especially in laptops these days the system manufacturers also make use of the BD PROCHOT flag (which will typically also be locked by default) to protect the CPU and various other sensitive hardware components against overheat damage.

hdmi said:

Good question... Personally, I, would go on MSI tech support and open a ticket to get the problems resolved, professionally ASAP. It's the shortest pain. Just be sure to backup all your important data before you send it in for servicing. Also ask them to fix the damaged cooling fan TBH.

See this reddit post by a user named Tetedeiench:

https://www.reddit.com/r/overclocking/comments/13g4p7b/comment/jjzh6uo

Click to expand...

ill try that, but thus is inly like 8 months old so idk if thats needed. also, thrre arent any damaged cooling fans i believe

zqchri said:

my gpu temps usually seem fine, are you sure this is needed?

Click to expand...

If they're fine, No - it's not necessary.

Tho, it can still help. Lowering temps for free (nothing the loose) - is always a good thing - both for the GPU and the system (lower heat dissipation).

neves said:

If they're fine, No - it's not necessary.

Tho, it can still help. Lowering temps for free (nothing the loose) - is always a good thing - both for the GPU and the system (lower heat dissipation).

Click to expand...

i dint think ive seen my gpu pass 50 like ever, unless im running stress tests for for a split second. its generally in the 40s

zqchri said:

i dint think ive seen my gpu pass 50 like ever, unless im running stress tests for for a split second. its generally in the 40s

Click to expand...

Only a demanding hardware accelerated app can do that - like a demanding game. For most part - even in less demanding titles - a GPU's computational need are rather low (for a modern GPU). So hey, if you get max 50 even in the games you play - that temperature is quite optimal (no point in bothering with extra tolls - i guess).

zqchri said:

ill try that, but thus is inly like 8 months old so idk if thats needed. also, thrre arent any damaged cooling fans i believe

Click to expand...

If it also crashes while doing everyday normal stuff outside of gaming, e.g. while browsing on the internet or watching a movie, running the kind of programs/apps that don't push the hardware close to its limits even by far. Then if we can normally assume that you did nothing wrong with Windows, the device drivers that are installed on Windows, the important BIOS settings, etc.. Then certainly, yes, if I were you I would send it back to MSI for servicing under warranty (problem #1), and also I would tell them to fix the cooling fan that you said is now half-broken as a result of holding a can of compressed air at too close a distance away from it (problem #2). It's not only the shortest pain. It's also the only way to be sure that the undervolting will have no negative impact on system stability (problem #3).

As for why some games don't seem to use the GPU that much. These are commonly referred to as CPU limited games. The performance of a CPU limited game is limited by CPU performance more heavily than it is limited by GPU performance. The opposite are GPU limited games. (Baldur's Gate 3 and Alan Wake 2 are only two examples of a GPU limited game.) To some certain degree, changing the graphics quality related settings of the game in question can alter the balance between being more CPU limited and being more GPU limited, for better or worse. Even so, if the game's performance is seriously held back by the CPU with choppiness during those times when it matters the most, upping the game's graphics quality settings isn't usually going to make it run smoother in pure terms of fps (frames per second). You'd have to find out what are the game's graphics settings that are a lot more demanding of the CPU, less demanding of the GPU, then try lowering these specific settings to be able to learn the results for that game.

zqchri said:

also, how would i get to these WHEA recoverable errors?

im currently trying a occt test over 30 minutes, does doing it for long periods of time spoil or harm thr cpu in anyway?

Click to expand...

OCCT will show any that occur as you run it. To see a history, open "event viewer" and navigate to

Applications and Services Log -> Microsoft -> Windows -> Kernel WHEA -> Errors

recovered ones are categorized as informational

This list contains all WHEA (Windows Hardware Error Architecture) errors, not just CPU but PCI-E etc. But others are rare.

hdmi said:

If it also crashes while doing everyday normal stuff outside of gaming, e.g. while browsing on the internet or watching a movie, running the kind of programs/apps that don't push the hardware close to its limits even by far. Then if we can normally assume that you did nothing wrong with Windows, the device drivers that are installed on Windows, the important BIOS settings, etc.. Then certainly, yes, if I were you I would send it back to MSI for servicing under warranty (problem #1), and also I would tell them to fix the cooling fan that you said is now half-broken as a result of holding a can of compressed air at too close a distance away from it (problem #2). It's not only the shortest pain. It's also the only way to be sure that the undervolting will have no negative impact on system stability (problem #3).

Click to expand...

the crashes so far only occur while i play valorant, that too for prolonged periods of time, but this was akl before the undervolt and i have yet to test it. however, there mightve been some miscommunication, i didnt use a can of compressed air to clean out my fans, it was just a small blow dry made for dust and in no way had it spoilt the cooling fans, it actually helped quite a lot with the cooling! although yeah if problems persist i will definately look into it, i think i should be fine as again its relatively new

PerpetualCycle said:

OCCT will show any that occur as you run it. To see a history, open "event viewer" and navigate to

Applications and Services Log -> Microsoft -> Windows -> Kernel WHEA -> Errors

recovered ones are categorized as informational

This list contains all WHEA (Windows Hardware Error Architecture) errors, not just CPU but PCI-E etc. But others are rare.

Click to expand...

gotcha, thanks!

zqchri said:

the crashes so far only occur while i play valorant, that too for prolonged periods of time, but this was akl before the undervolt and i have yet to test it.

Click to expand...

Well if it's only one game and only after long periods or almost, then the laptop is probably not to blame. It could (most likely) be just a bug in the game itself or (not very likely, but still possible nevertheless) a bug in a device driver or in the BIOS that gets triggered under some certain set of circumstances that only occurs during that game after a long period.

zqchri said:

however, there mightve been some miscommunication, i didnt use a can of compressed air to clean out my fans, it was just a small blow dry made for dust and in no way had it spoilt the cooling fans, it actually helped quite a lot with the cooling!

Click to expand...

Sorry I remembered it wrong. In post #50 you did say that you used an air blower, but in that same post you also mentioned that one of the fans appears to run a little bit slower.

zqchri said:

although yeah if problems persist i will definately look into it, i think i should be fine as again its relatively new

Click to expand...

You can use HWiNFO64 to monitor the fan speed. MSI Raider GE68HX laptops fan speed monitoring and temperature.
I could be wrong, but I think you should be able use it to check if one of the fans runs slower.

hdmi said:

Sorry I remembered it wrong. In post #50 you did say that you used an air blower, but in that same post you also mentioned that one of the fans appears to run a little bit slower.

Click to expand...

i thought that was the case, but after comparing it from the previous fan speeds, it was about the same. the only difference was one fan going down by a hundred or so rpm upon starting up my cooler boost, but they both now sustain similar rpm and now seem to have better airflow out of all vents

hdmi said:

Well if it's only one game and only after long periods or almost, then the laptop is probably not to blame. It could (most likely) be just a bug in the game itself or (not very likely, but still possible nevertheless) a bug in a device driver or in the BIOS that gets triggered under some certain set of circumstances that only occurs during that game after a long period.

Click to expand...

i personally thought it was just the temps, and after doing some tests with zbook i believe the problem is probably due to temps OR my ram, which i have yet to test using memtester86+. i dont think its the game itself, i reckon youre right about the certain set of circumstances because that's the most high performing game i play, used to bring my cpu to 80 before the undervolting and processor editing.

zqchri said:

i personally thought it was just the temps, and after doing some tests with zbook i believe the problem is probably due to temps OR my ram, which i have yet to test using memtester86+. i dont think its the game itself, i reckon youre right about the certain set of circumstances because that's the most high performing game i play, used to bring my cpu to 80 before the undervolting and processor editing.

Click to expand...

If you play a game like, e.g., Cyberpunk 2077 on it you might actually even see it run much hotter still. Reaching far above 80°C on the CPU while playing recent, demanding, games titles on a gaming laptop with an Intel i9-13950HX is just normal, but if it also crashes during many of these specific types of games, then you've got something to really start worrying about. Also remember that these CPUs were still relatively very new at the time when you bought your laptop. BIOS updates can help to improve compatibility and stability of the RAM. Technically speaking, running DDR5 above 4800MHz is called "overclocking" with the CPU you have. What's certain is that the XMP related configurations/settings data often tends to be among various important things that get updated when you update your BIOS. Not saying that the outdated BIOS is the culprit, but it could be.

hdmi said:

If you play a game like, e.g., Cyberpunk 2077 on it you might actually even see it run much hotter still. Reaching far above 80°C on the CPU while playing recent, demanding, games titles on a gaming laptop with an Intel i9-13950HX is just normal, but if it also crashes during many of these specific types of games, then you've got something to really start worrying about. Also remember that these CPUs were still relatively very new at the time when you bought your laptop. BIOS updates can help to improve compatibility and stability of the RAM. Technically speaking, running DDR5 above 4800MHz is called "overclocking" with the CPU you have. What's certain is that the XMP related configurations/settings data often tends to be among various important things that get updated when you update your BIOS. Not saying that the outdated BIOS is the culprit, but it could be.

Click to expand...

what would the bios update exactly do? would it remove any data whatsoever from my computer or anything? i still do not know the cause of the crashes but now its either just valorant itself or my ram, which i cant get memtester working due to USB errors but windows memory diagnosis said i was fine. then again, WMD was said to be unreliable so im unsure.

zqchri said:

what would the bios update exactly do? would it remove any data whatsoever from my computer or anything? i still do not know the cause of the crashes but now its either just valorant itself or my ram, which i cant get memtester working due to USB errors but windows memory diagnosis said i was fine. then again, WMD was said to be unreliable so im unsure.

Click to expand...

Installing the latest BIOS update can help to fix various problems and also can help to avoid bumping into various problems in the future. RAM stability problems are a good example of problems that can sometimes be solved by updating the BIOS, and these problems also are a good example of why the importance of updating the BIOS should not be underestimated. RAM stability problems that only start to occur above a certain temperature are not that extremely uncommon, and, contrary to popular belief, watching these problems disappear after updating the BIOS is not that extremely uncommon either. If it turns out that updating the BIOS doesn't help, at least you'll know that the problem needs to be sought elsewhere, which brings you one step closer to the solution. Whereas if it turns out that it does make the problem go away, then all the better.

"Loading BIOS default settings before executing update is mandatory. Refer to steps in the quick guide for more information."
Here is the link again: https://www.msi.com/Laptop/Raider-GE68-HX-13VX/support?sub_product=Raider-GE68HX-13VF

Also, could you please describe the USB errors that you are experiencing?