Question "Kernel-Power 41" restarting my PC under high load games ?

[Burger1003]

-PC has been restarting whilst playing games with heavy load in the last month after i have replaced motherboard and CPU.
-In event viewer it will say Kernal power 41

CPU - AMD Ryzen 5 5600X 6-Core Processor 3.70 GHz
GPU - NVIDIA GeForce RTX 3070 (Driver version 516.59)
PSU - Cooler Master 850v2 full modular
MB - B450 Tomahawk MAX II (MS-7C02)
Memory - 2x16 DDR4 Corsair 3200mhz

-Games it has happened in (Escape from Tarkov, Hell let loose, Apex Legends)
-The exact scenario is, that i am playing the said game and i can play around in the menu's and all that for hours. But as soon as i decide to load into a game i play with normal fps for roughly 5mins and then it'll just randomly restart my PC with no warning, just black screen and restarting. But sometimes i wont have an issue at all but the next day it will start happening again.
- i have tried playing Tarkov as soon as I reinstalled my parts and that's when i first started to notice the issues. i had read on Tarkov forums and people saying that its just in recent updates and stuff similar so i gave up on fixing the issue. until days ago i tried playing Hell Let Loose and it did exactly the same as when playing Tarkov.

-I have installed my new CPU, Motherboard and SSD at the same time.
-i thought it could be the SSD (970 EVOPlus) but i have moved games from across different drives and the error still occurs.


-I have read many other forums and they also have mentioned it being a PSU, i did replace mine and nothing had changed/error still occurs.
-I have also seen people saying it could be GPU related, so i used a known good GPU and issues still occur.
-I have changed out my new ram with old ram and the issue still occurs.
-Have also enabled XMP profiles 1 & 2 and errors still occur.
-My GPU drivers are all up to date
-Have performed several stress tests on GPU, CPU, PSU, did a memtest64 and it was all fine.
-Have performed this test "DISM /Online /Cleanup-image /RestoreHealth" in CMD and it came up with issues then followed up with " sfc /scannow" after that i did an "in-place repair" which downloaded an ISO and then i did a "repair install" (I have just basically followed this forum and i did have corrupted files and i updated windows but the issue still occurs - https://answers.microsoft.com/en-us...rruption/32c68c1b-d153-4338-83ec-17bcf130fe6d)
-I have Malware-bytes and performed multiple scans and nothing is detected.
-I've checked my 24pin plug to MB and its secure, ram is secure, GPU is secure, temps are fine.
-Have also done a CMOS repair or something where i short 2 pins on the MB for 10sec and that didn't fix anything
-Have disabled all programs from running on start up.


DXDIAG

My last option if nothing else is achieved is to run this script my friend sent me - https://old.reddit.com/r/TronScript/wiki/index

 

[Aeacus]

-I have read many other forums and they also have mentioned it being a PSU, i did replace mine and nothing had changed/error still occurs.

1. What PSU you had earlier? Make and model?
2. Replacing PSU with questionable quality PSU doesn't mean the issue, which 90% of times is PSU issue, is fixed.

I have also seen people saying it could be GPU related, so i used a known good GPU and issues still occur.

This is called transient power spikes.

GamersNexus made a nice vid about it:

View: https://www.youtube.com/watch?v=wnRyyCsuHFQ


Some PSUs, namely questionable quality ones, are incapable of handling those transient power spikes.

--

Here, i'd look towards better quality PSU, and to be safe, 1kW unit as well. E.g Corsair RM/ RMi/ RMx series. <- Once you have that, then you can rule out PSU issue (and i think, you won't get the issue anymore either).

 

[Cyberat_88]

"Temps are fine" Is not a Number, what you define as fine is not the standard you should use.
Anything above 75C on CPU or GPU and you are cooking components.

 

[Burger1003]

1. What PSU you had earlier? Make and model?
2. Replacing PSU with questionable quality PSU doesn't mean the issue, which 90% of times is PSU issue, is fixed.

My psu earlier was a Be Quiet 750W, and this psu is brand new that i have put in

This is called transient power spikes.

GamersNexus made a nice vid about it:

View: https://www.youtube.com/watch?v=wnRyyCsuHFQ


Some PSUs, namely questionable quality ones, are incapable of handling those transient power spikes.

--

Here, i'd look towards better quality PSU, and to be safe, 1kW unit as well. E.g Corsair RM/ RMi/ RMx series. <- Once you have that, then you can rule out PSU issue (and i think, you won't get the issue anymore either).

 

[Cyberat_88]

Sure the video is very general and doesn't address that 850W PSU is in general ok, albeit minimal for today's intel G-12, Ryzen 9, Nvidia 3000s & AMD 6000s combinations.
The next gen will set the bar much higher no doubt at 1000W or 1200W depending on testing, but THIS is not the case.
I still think it's overheating issues and I'm not getting a response in temp. numbers.

 

[Burger1003]

Sure the video is very general and doesn't address that 850W PSU is in general ok, albeit minimal for today's intel G-12, Ryzen 9, Nvidia 3000s & AMD 6000s combinations.
The next gen will set the bar much higher no doubt at 1000W or 1200W depending on testing, but THIS is not the case.
I still think it's overheating issues and I'm not getting a response in temp. numbers.

Im performing a stress test for the last 10mins

 

[Cyberat_88]

Where is the usage % ? 77C is just a bit over 75C. There is a program called 'Heavy Load' see if it runs above 77C even.
To lower temps on CPU, set your CPU cooler/fan to 100% at say 55C, usually done from Bios to be sure.

 

[Burger1003]

Where is the usage % ? 77C is just a bit over 75C. There is a program called 'Heavy Load' see if it runs above 77C even.
To lower temps on CPU, set your CPU cooler/fan to 100% at say 55C, usually done from Bios to be sure.

its at 100% and stayed below 78 for the remainder 20minutes. i really dont think its a temp related issue

 

[Cyberat_88]

At what temp do your CPU fans go 100% ?

 

[Burger1003]

At what temp do your CPU fans go 100% ?

i wouldnt know, still dont see how this is relevant.

 

[Cyberat_88]

Looking at your log, something strikes me odd, did you install motherboard drivers at all ?

 

[Cyberat_88]

I don't think you understand Degrees Celsius to begin with, at 100C water boils, at 80C your PC components start to melt, soldering loosens, etc...
Most CPU manufacturers set it to disregard YOUR Opinion and shutdown or reboot at 75C.

 

[logainofhades]

I don't think you understand Degrees Celsius to begin with, at 100C water boils, at 80C your PC components start to melt, soldering loosens, etc...
Most CPU manufacturers set it to disregard YOUR Opinion and shutdown or reboot at 75C.

I don't think you understand what modern processors are capable of. Max temp for Ryzen is 95c. 75c is still plenty fine with these chips. 77c, with an extreme load of a stress test is nothing to worry about. A modern chip will thermal throttle, before it just shuts down, and they are well below that point.

OP: I would look at getting a better PSU. CM is quite hit and miss on quality, and the transient spikes, as mentioned earlier, it probably cannot handle.

 

[Cyberat_88]

I don't think you understand physics, run your cpu at 95C and you'll buy a new one soon. If I was a CPU manufacturer, what max. temp. should I put in the manuals ?

 

[Cyberat_88]

Setting fans to 100% at 55C in Bios does not cost as much as buying a new PSU and finding out it's not the problem.

 

[rgd1101]

https://www.pcbdirectory.com/commun...oint of solder? ,221°C – 226°C 11 more rows

https://www.amd.com/en/products/cpu/amd-ryzen-5-5600x
Max. Operating Temperature (Tjmax)
95°C

 

[Burger1003]

I don't think you understand what modern processors are capable of. Max temp for Ryzen is 95c. 75c is still plenty fine with these chips. 77c, with an extreme load of a stress test is nothing to worry about. A modern chip will thermal throttle, before it just shuts down, and they are well below that point.

OP: I would look at getting a better PSU. CM is quite hit and miss on quality, and the transient spikes, as mentioned earlier, it probably cannot handle.

yeah thats fair enough, ill keep that in consideration but i feel that getting a 1kw is a bit overkill. what is a good psu brand?

 

[Colif]

You shouldn't need 1kw PSU for that, maybe an 850 like you have now.

I would get a Corsair RM850x
Seasonic make good PSU as well. But they were known for having transient problems, not sure if thats still case.

 

[logainofhades]

Evga G6 are another good choice. I know Tech Deals had lots of issues with his Seasonics, and the transient spikes. He changed to Evga G6's, and the problems went away.

 

[shADy81]

Its not actually a bad PSU....

https://www.tomshardware.com/reviews/cooler-master-v850-gold-v2-power-supply-review

Also a Ryzen 5600X is maybe around 120 Watts with PBO enabled and a 3070 maybe hits transient spikes of around 300 W max. That shouldn't be the problem on an 850watt unit. You'd probably be ok with a good quality 650 watt.

I'll have a stab and say try running with only one stick of RAM at stock (no XMP) and see how that goes before you order a new PSU. 16GB sticks on a B450 board which maxes out at 3200 might be tickling its boundaries, and it doesnt cost anything to try.

 

[logainofhades]

The B450 Tomahawk max II can handle over 3200, with a Zen 3 chip. Also 2x16gb is perfectly fine. If it was 4x16gb then maybe I could see it being something the board was having trouble handling.

https://www.msi.com/Motherboard/B450-TOMAHAWK-MAX-II/Specification

1866/ 2133/ 2400/ 2667Mhz (by JEDEC)
For AMD Ryzen Gen3 (R5/R7/R9)
2667/2800/2933/3000/3066/3200/3466/3600/3733/3866/4000/4133 by A-XMP OC mode
For AMD Other CPU
2667/2800/2933/3000/3066/3200/3466 by A-XMP OC mode

OP: There is something else, that I just thought of. Is your bios up to date? It looks like MSI finally released an Agesa 1.2.0.7 update. If you are on 1.2.0.6, it could be the problem.

https://www.msi.com/Motherboard/B450-TOMAHAWK-MAX-II/support

 

[Aeacus]

I don't think you understand physics, run your cpu at 95C and you'll buy a new one soon.

Yeah, No.

I had high temps issue in my old AMD build (full specs with pics in my sig), and in there, i have AMD Athlon II X2 250 CPU. It's quite weak CPU actually and way back then, there were several instances where i saw up to 102C out of my CPU, until PC automatically shut down, to keep CPU alive. Did my CPU burn up? No, it did not. Once i pinpointed the issue, i overhauled the PC (new PC case, better fans etc) and my CPU is still going strong, to this day (it being 11 years old now). Despite it reached 102C several times.

CPUs are the most durable component inside the PC and it would take a lot frying before it's toast. Modern CPUs thermal throttle, to keep the temps ever reaching over 90C, when max what they can do is 95/100/105C.

but i feel that getting a 1kw is a bit overkill. what is a good psu brand?

Yeah, 1kW unit is a tad bit overkill, but when looking towards upcoming GPUs and their power draw, that is going to skyrocket, 1kW or even 1.2kW units could become a norm, in high-end systems with very power hungry GPU.

850W good/great quality PSU will do fine for now, 1kW is more for a future proofing, if you plan to use PSU with next gen high-end GPU.

Good PSU brands are Seasonic and Corsair. Now, as said above, some older Seasonic units did have issues with GPU power spikes, namely:

Older Seasonic Focus revisions (made before 2018) had transient sensitivity issues. Some new SKUs are not affected because they were released later (Seasonic Focus GM/GX/PX refreshes, Phanteks AMP, NZXT C/E etc.).

and by JonnyGuru:

Seasonic PRIME based units experience shutdowns with RTX3080/3090 (and possibly RX6900 XT) GPUs, especially ones with unlocked power limit like FE and ASUS Strix. It is recommended, if going with such units, to overprovision wattage, 1kW for RTX3080 and 1.2kW for RTX3090, or power-limit the GPU.

If you don't want to risk it, by getting pre-2018 Seasonic unit from the store, good choice is going with Corsair RM/ RMi /RMx unit. HX/ HXi/ AX/ AXi is the high-end series from Corsair and if you want better than RM-series, look towards those.

Its not actually a bad PSU....

https://www.tomshardware.com/reviews/cooler-master-v850-gold-v2-power-supply-review

Also a Ryzen 5600X is maybe around 120 Watts with PBO enabled and a 3070 maybe hits transient spikes of around 300 W max. That shouldn't be the problem on an 850watt unit.

OP might have a lemon. Since even among the best, there are lemons.

I avoid Cooler Master PSUs since their quality is so sporadic, that it's best to avoid them completely.

--

Trying with 1 stick of RAM doesn't hurt, but i don't think the issue would be with it.

And BIOS update could be viable, but i, personally, doubt that it fixes the issue.
Oh, it is important to know, that if, for whatever reason, BIOS update should be interrupted (e.g power loss), MoBo will be bricked. If very lucky, BIOS rollback can be done (some MoBos have that feature), or instead using 2nd, working on-board BIOS, if MoBo has that dual BIOS feature as well.

 

[shADy81]

@Aeacus

As I understand OP Its a brand new PSU after the first one was suspected, what are the chances both models were the same issue?

-I have read many other forums and they also have mentioned it being a PSU, i did replace mine and nothing had changed/error still occurs.

 

[Aeacus]

As I understand OP Its a brand new PSU after the first one was suspected, what are the chances both models were the same issue?

OP haven't disclosed to us, what his previous PSU was. Could've been completely different make and model. IF it was the same CM unit, then i wouldn't be surprised of the poor quality of CM units.

In the same sense, Thermaltake PSUs are also best avoided. There are few good units Tt produces but bulk of them is sub-par. Same essentially goes to EVGA as well. While SuperNova series should be good, there are variants that are bad, e.g G5.

All-in-all, end consumer should not thread themselves through the PSU lineups to look what is good and what is not. Sadly, in PSU world, you can't buy an unit based on it's price alone. You actually have to dive deep into reviews and read if the PSU is any good. And even then, you may end up on a lemon, but for that, there is RMA and warranty.

 

[Roland Of Gilead]

Sure the video is very general and doesn't address that 850W PSU is in general ok, albeit minimal for today's intel G-12, Ryzen 9, Nvidia 3000s & AMD 6000s combinations.
The next gen will set the bar much higher no doubt at 1000W or 1200W depending on testing, but THIS is not the case.
I still think it's overheating issues and I'm not getting a response in temp. numbers.

I don't think you watched the video through, otherwise you would have picked up on the transients and roughly how they are calculated.

for RTX3xxx the transients are typically 2.5x the TGP
for RTX2xxx the transients are typically 2.0 x the TGP
and even for GTX1xxx they also have spikes, but only to the order of 1.5 x.

So with that in mind a 3070 with a TGP of 220w would require approx 550w for the GPU alone, with another lets say 200w for the system (depending on CPU) and another 100w for clearance.

So with that said, a very good quality 850w would suffice just fine.

Of course you to have AIB partners with higher power options, so the spikes can increase when removing power restrictions. Another thing to factor into transient spikes.

A RTX3090 on the other hand, would need 875w for just the GPU alone, with an additional 300w for other components and some clearance.

The good news is that when RTX4xxx comes around (with 3 x TGP spikes), with reported TGP being massive, we will soon need to have a 1.2kw PSU just to get by!!