Event 41 Kernel-Power crashes when gaming on newly built pc

Toate discuțiile > Forumuri Steam > Hardware and Operating Systems > Detaliile subiectului

luunar cu 23 ore în urmă

Finished first pc build around two weeks ago and since day one been experiencing sudden shutdowns exclusively when running games. I will be playing something and all of a sudden, screen will turn black, I hear a click noise from my pc, and the lights shut off, all at once. I am completely lost, and have no idea how to fix this. There are no other critical events listed in Event Viewer aside from the Kernel-Power 41 error. I know this particular one can be a real pain to track down, but I was hoping someone here might have some advice for me :]

I’m unable to reliably reproduce the issue— It's occurred within 20m of launching a game like GTA V, but other times I'm able to run something like Fortnite for hours on end with all the ray tracing and whatnot maxed on highest settings. A common trigger for the sudden crashes appears to be when trying to apply any sort of graphic settings changes inside a game, but other times I can change them just fine. The shutdowns happen in the middle of gameplay as well, at around the same frequency.

The crashing doesn't occur in every game I play. My pc did not crash a single time when playing through an older game like portal 2. It's crashed the most times while playing Fortnite, likely because that's the game I've played the most so far on this pc. Because the crashing is so random it's really difficult to test across different games/circumstances. For instance, I have gone five hours without a crash in Cyberpunk with ray and path tracing enabled, but crashed within the first 20 minutes of launching GTA V Enhanced.

I have so far tried:

reseating all cables in my psu & motherboard
removing GPU drivers with DDU and reinstalling directly from NVIDIA while in safe mode
starting over with fresh windows 11 installation
updating BIOS to latest version
disabling integrated graphics in BIOS
making sure EXPO was disabled in BIOS
running sfc /scannow (returned fine)
plugging pc into a different surge protector and outlet
unplugging riser cable and slotting GPU directly into the motherboard
collecting HWiNFO logs- saw nothing that immediately stood out to me. The img attached displays some of the final values prior to a crash across some (what I think to be) important categories.
leaving synthetic benchmarks (OCCT [ran the 3D Adaptive, Power, CPU, and Memory test], Furmark, Unigine Superposition, Cinebench, 3Dmark free version) to run for extended periods of time, none of which ever resulted in a shutdown.
Setting GPU power limit to 83% (lowest it goes for me in MSI afterburner)
Checking "C:\Windows\LiveKernelReports" (the files/folders inside that directory are dated back more than a week ago yet I have crashed numerous time since then. I didn't really know what I was doing but I tried my best to analyze them in WinDbg and saw some error labled "VIDEO_TDR_TIMEOUT_DETECTED (0x117)." Regardless, I think the fact that the file date doesn't correlate with crashes rules them out.)

Specs are as follows, I purchased everything brand new:
CPU: AMD 7800x3d
CPU cooler: Lian Li Hydroshift II 360 TL
Motherboard: GIGABYTE B850 AORUS ELITE WIFI7 ICE rev. 1.x (currently on BIOS version F6)
Ram: Corsair Vengeance DDR5 RAM 32GB (2x16GB) 6400MHz
SSD/HDD: Samsung 990 EVO PLUS SSD 2TB
GPU: MSI RTX 5060 Ti 16G Gaming Trio OC
PSU: Corsair RM750e ATX 3.1 PCIe 5.1 Fully Modular 750W (2025)
Chassis: HAVN HS420 VGPU
OS: Windows 11 LTSC IoT 24H2 (Initially was running normal Windows 11, a new OS was my first step of troubleshooting. On second installation, I used a different USB, recreated the installer, and installed everything in offline mode.)

If anybody has any suggestions I would be truly grateful to hear them. I’m getting desperate enough to just try and go through the hassle of taking my pc apart and returning every single component while my window is still open.

Editat ultima dată de luunar; cu 21 ore în urmă

< >

Se afișează 1-13 din 13 comentarii

luunar cu 23 ore în urmă

Forgot to include the aforementioned hwinfo pic, here is some of my log just before a crash: https://imgur.com/ugT0Kia

Illusion of Progress cu 23 ore în urmă

Event ID 41 is a symptom, not a cause or clue. The way Windows works is that it "leaves a note to itself" so to speak when shutting down saying "yes, this shutdown is expected". Upon start, it checks for this note. If it is found, nothing happens. If it is not found, which it won't be if the shutdown was unexpected and thus it wouldn't have created it, then it creates an event entry which is ID 41 so that you can see a sudden shutdown happened in case you didn't know (such as the PC shutdown unexpectedly while you were away).

If this is happening while gaming and if the shutdowns are not restarts, then I'd be looking at the PSU. Neither that CPU nor GPU is very power hungry, and while the PSU has more than enough wattage for them, OCP/OVP/whatever could be getting tripped or otherwise the PSU might not be coping with some spikes.

Alternatively, something might be unstable at a certain point in the frequency/voltage curve, but... I'd expect a restart or crash instead of a sudden power off in that case.

VIDEO_TDR_TIMEOUT_DETECTED (0x117)

This is indeed the video card having a problem (and trying to adjust the TDR timeout is usually a bandage because you "shouldn't" be tripping the default value to begin with in most cases). However, this might simply be a cascading result of, say, a power issues. If these do not coincide with the time of the shutdowns then they are probably not related, but I would keep an eye on these as their own thing.

Normally I'd say one of the first things to try is to disable any RAM profiles and ensure no overclocks/undervolts/curve optimizers are in place, but... it sounds like you already correctly tried those steps.

That RAM speed might be a bit ambitious and isn't always as guaranteed as 6,000 MHz is, but if you tried with EXPO disabled then the RAM probably isn't related.

Do you have a spare GPU or PSU to test?

Editat ultima dată de Illusion of Progress; cu 23 ore în urmă

emoticorpse cu 22 ore în urmă

When it crashes, does it stay off or instantly reboot and go back to powered on state?

luunar cu 22 ore în urmă

Postat inițial de emoticorpse:
When it crashes, does it stay off or instantly reboot and go back to powered on state?

The PC remains off after a crash. After manually hitting the power button it turns back on immediately.

luunar cu 22 ore în urmă

Thank you for taking time to leave such a detailed reply.

Postat inițial de Illusion of Progress:
Do you have a spare GPU or PSU to test?

Unfortunately I have no spare parts as this is my very first pc.

Postat inițial de Illusion of Progress:
OCP/OVP/whatever could be getting tripped or otherwise the PSU might not be coping with some spikes.

I will go ahead and buy another PSU as it's one of the easier items for me to return.
Would you recommend buying the same PSU, or maybe something with higher rating/Wattage? I read online 750W would be more than enough for my setup, but maybe I'm mistaken?

Editat ultima dată de luunar; cu 22 ore în urmă

Numlock687 cu 22 ore în urmă

its a log. dont worry about it. its been implemented since win10 and 11

hawkeye cu 22 ore în urmă

I used to get this but fixed it a few weeks ago.

I reduced the Windows power setting to "normal".
I replaced the power board as it was getting old.
I ran DDU and removed Nvidia drivers and installed the Creators version (DCH?) from the nvidia website.

One of the 3 fixed it and my bet is on the Nvidia drivers being the problem as it seemed to start when I got a windows 11 / nvidia driver update. rtx4060ti.

I also turned off the e-cores but that didn't do anything.
And it could have been the psu but I monitored the power draw and none of the components ever used more than 50% of their max.
Happened under Win 11. Never happened under Win 10.

Editat ultima dată de hawkeye; cu 22 ore în urmă

luunar cu 21 ore în urmă

Postat inițial de hawkeye:
I used to get this but fixed it a few weeks ago.
I reduced the Windows power setting to "normal".

Earlier on in my troubleshooting I tried switching my power plan from balanced to both high performance and ultimate performance. None of them seemed to fix my crashes :(

Postat inițial de hawkeye:
I replaced the power board as it was getting old.

By power board do you mean surge protector? If so, do you happen to know what kind of stats it has? I purchased one (1875W - 15A at 125V) along with my pc and tried another I had laying around in my house, but maybe both are insufficient.

Postat inițial de hawkeye:
I ran DDU and removed Nvidia drivers and installed the Creators version (DCH?) from the nvidia website.

Did you download the newest version? I have tried three driver versions (the latest, as they have been releasing) since I built my PC and the issue has persisted across all of them.

Illusion of Progress cu 20 ore în urmă

Postat inițial de luunar:
I will go ahead and buy another PSU as it's one of the easier items for me to return.
Would you recommend buying the same PSU, or maybe something with higher rating/Wattage? I read online 750W would be more than enough for my setup, but maybe I'm mistaken?

Unfortunately, I'm not up to date on PSUs so I'll let someone else answer that. On wattage alone, that is certainly more than enough for your hardware, but certain models may have considerations beyond the total wattage alone.

The PSU is generally a good first step because you need to ensure you have stable power. It's far less likely you'll have two bad PSUs in a row, so if the problem remains after trying a new PSU, then the issue is likely elsewhere. I'd probably be looking at the graphics card after that point.

If you do change the PSU, be sure you change the cables as well. Don't take the shortcut of reusing your existing cables. They are not always interchangeable between different PSUs (the same PSU model should be fine though) and can sometimes cause immediate failure if you use the wrong cables.

Set-115689 cu 20 ore în urmă

Decrease the gpus clock and memory speeds in msi afterburner then test. It might be when you reach x wattage for the entire system it crashes. Gimping the gpu should decrease the power draw of the entire system.

Remove/disable any overclocks/overclocking software (task manager-start up)

Set power plan to a normal setting vs high performance, etc... It's not worth it.

Case is weird. Psu get ok airflow/cooling? Vents blocked? Psu overheating?

Editat ultima dată de Set-115689; cu 19 ore în urmă

#10