Kernal 41 (task 63) power errors

[Darthain]

I am having a problem, and it is recuring Kernel 41 (task 63) black screen crashes.  Windows has not been helpful.  It is a 0x0, all bug checks are 0x0…   It started about once every 8 days or so.  Not a blue screen hang, black screen…  No minidumps available (folder does not exist).  This can happen anytime, playing a game, editing images, it just happened at windows key in.  I found it to be more repeatable under heavy load, and could be accelerated.  Can be had 'reliably' with prime 95, but timeline varies from instant, to within 8 hrs...  The problem is intermittent and very hard to repeat.  

 

Fast boot is off.

 

Computer specs:

Windows 10 x64 (only OS on system)

Desktop System built April 2021, OS installed then

Ryzen 5950x (Noctua d15s cooler)

64 gb (16x4) G. skill Ripjaw 3200mhz ddr4 (running in docp standard)

Gigabyte Aorus Master 570s board (was a asus hero viii crosshair wifi…)

MSI Mecha x2 GT 6700 xt Radeon card

SSDs (680 and 670 nvme)

HX750 corsair PSU (750w platinum cert

Phanteks 400A case

 

Diagnostic steps:

-Did general burn in day 1 (prime95m gpu stress, memtest86 (16 passes)) no issues

-When first crash occurred, uninstalled and reinstalled all drivers.  Reseated ram, switch configuration. Reseated all drives and cards

- Replaced the PSU (and all cables)

-Removed CPU cooler and reseated CPU and cooler

-Ran with ½ ram loadout, switched when problem continued.  Memtested each ½ load out 12 passes, no errors detected.  Was running a lot of prime 95.  Both blend, small FFT, and Large FFT all resulted in crashes.

-replaced CPU – problem continues, had been testing with small FFT prime95

- replaced motherboard, problem continues, had been testing with small FFT prime95

- Currently running further ram diagnostics 1 stick at a time, 4 passes in each DIMM on new motherboard.

 

Let me know if you have any suggestions.  I’ve been trying to solve this for 3 months now and am beyond frustrated.  The only existing parts are the RAM and GPU, and I can't get any diagnostics of the ram showing errors to date.  

[Guest]

Is this one of those that shuts off while idle?

 

Try disable power down mode in memory timings menu. Do your testing at idle, not load.

 

If you are running xmp, try a lower frequency like 2933mhz as well.

 

GL

[Darthain]

25 minutes ago, ShrimpBrime said:

Is this one of those that shuts off while idle?

 

Try disable power down mode in memory timings menu. Do your testing at idle, not load.

 

If you are running xmp, try a lower frequency like 2933mhz as well.

 

GL

It has done it while idle yes.  Several times.  The last bit have been under load as I've been trying very hard to pin something down.

 

I'll try reducing the memory profile after this test round.  I just had a crash at windows keyin... With only 1 stick in, but it was on the docp 3200 profile.

 

When you say test idle not under load, what specifically do you mean?  Any memory testing has been done idle/ in uefi, etc.  Not within windows doing other tasks,

 

[Darthain]

To amend.

 

Memtest86 each stick single in each dimm, at 3200 clock. All pass.

Ran HCI memtest in windows, was able to crash while opening enough threads to use the ram twice.

While using pro, it ran 600% coverage then did what I can only call a spinning wheels hang.  Windows where there, HCI stopped updating, hardware monitors stopped updating.  Windows where moveable. Nothing was usable, task manager blank, etc. Log stopped 2 hrs before, no errors...

 

Removed all memory OC.  Prime 95 with all ram went 17 hrs.  Same spinning wheel crash.  When I tried to log out of my profile it crashed kernal 41. But before that my clock read 5am... At 6:02.

 

Removed 1 pair of memory, other pair still installed.  Currently moving  fine 15.5 hrs in.  If successful, seems to hint none of my ram can hold advertised clock, and that Infact one pair can't hold 2133...

 

 

 

 

 

 

 

 

 

[Darthain]

Had a stable 24 hr prime (blend) with 1/2 the ram at 2133.

 

Moved on to running one pair (not the really poor one) at 3133, it was looking fine until today, it crashed.  First one in 14 days though.  Aug 4 through Jul 21 was fine.  Moving in the right direction.  Once my one pair of ram returns going to also return this one as it isn't stable at 3200 anyway, and it should be given that's its spec.

 

Still open to any and all suggestions, but overall feel this is moving in the right direction for a change at least.

[Okutida]

Hi @Darthain

Did you solve your Kernal - Power -41(63) issue?

 

I have thesame/similar problem. 

 

Could you let me know how it goes? 

 

Kind regards, 

Okutida

[SM Phoo]

22 hours ago, Okutida said:

Hi @Darthain

Did you solve your Kernal - Power -41(63) issue?

 

I have thesame/similar problem. 

 

Could you let me know how it goes? 

 

Kind regards, 

Okutida

 

I've been plagued with this error for a good few months now, and also at my witts end!

 

I formatted my machine around 3 weeks ago, and everything seemed good, up until 4 days ago and now it's back, which screams software, but I just don't think it is.

I've not done any hardware changes because my system is watercooled, so it's hard to just swap out parts.

I mostly get the error when I wake from sleep or when the PC has just turned on, although it can happy at other points.

 

I'm starting to troubleshoot the hardware side of things soon, so I'll see if that comes up with anything.

I did a full memtest which came back clear.

Ran stress tests on the CPU, GPU and PSU and had no crashes, so I feel I've tested them well enough.

 

As of last night, I turned off quick startup, and so far so good, but I'm not filled with confidence.

When I look into the crash reports, even though it's a kernal issue, it seems to do with Nvidia, so I think it's the GPU, so my next 3 options are:

 

I use a riser cable so it could be that.

I use custom extension cables for cablemod, so could be that.

It could be a faulty GPU

 

Today I am lucky enough to borrow another GPU, so I will connect that up, which will negate the rise cable, extension cable and GPU, and if I get no errors, I know it's one of them

[Okutida]

Hi 

@SM Phoo

@Darthain

 

If you have AMD CPU please read this out. 

 

Kind regards, 

Okutida

[SM Phoo]

3 hours ago, Okutida said:

Hi 

@SM Phoo

@Darthain

 

If you have AMD CPU please read this out. 

 

Kind regards, 

Okutida

I could try that again.

I did actually lower my speeds to 3200 and from what I remember, I do think it worked (I cant fully remember though!)
When I did my clean wipe of windows, I did actually take it back to 3600 and did a memtest and it cleared, so I just left it.

 

Worth trying again though!

[Okutida]

Hi

@SM Phoo let me know how did it work out. 

 

 I done this yesterday - please see photo. 

 

Last evening 27.11.2021. gaming - Hell let loose, aprox - 60 min  - no blacks creen. 

I know - it is not much  - will do more. 

 

Lets keep in touch,

Okutida

 

[SM Phoo]

8 minutes ago, Okutida said:

Hi

@SM Phoo let me know how did it work out. 

 

 I done this yesterday - please see photo. 

 

Last evening 27.11.2021. gaming - Hell let loose, aprox - 60 min  - no blacks creen. 

I know - it is not much  - will do more. 

 

Lets keep in touch,

Okutida

 

 

Hey

Not done it yet.

So far so good for me turning fast boot off.

Had my PC on for around 3 hours, sometiems restarting to see if I could force the problem, but nothing.

Went back on later and played 3 hours of Apex fine.

 

I don't think this is the solution I am looking for, but I only want to test one thing at a time, so, for now, turning fast boot off is working. Changing ram frequencies will be next on the list though

[Okutida]

Hi @SM Phoo

 

Bad news. 

I got black screen today - playing Hard west. Not demanding video game. 

But most interesting thing - under Control panel - Administrative tools - Event viewer - it does not count as Critical - Kernel Power. 

 

Error did not count at all. 

 

Kind regards, 

Okutida

[Okutida]

Hi @SM Phoo

Still got black screen.

 

Now I will thy these recomendations: https://windowsreport.com/kernel-power-41-error-windows-10/

 

thats includes fast boot as well. 

Kind regards, 

Okutida

 

[Okutida]

Hi @SM Phoo

Did you manage to sort it out? 

 

Kind regards

Okutida

[SM Phoo]

On 12/3/2021 at 8:11 PM, Okutida said:

Hi @SM Phoo

Did you manage to sort it out? 

 

Kind regards

Okutida

Hey, sorry for the late replay, I didn't get a notification.

 

So, no. (hence logging in)

My PC has worked perfectly for over a week, so I thought turning on quick startup worked.

For the last week or so I've been playing Skyrim, but today, after an hour or so on it, I decided to load up Apex and I got the crash.

Restarted the PC, loaded up again, crash. Did this about 8 times in total, all resulting in crashes...

Then it decides to work, so I messed around in the shooting range for a while, tabbed out.... crash

[Okutida]

Hi @SM Phoo

 

For me did not work out as wel. 

 

Thingd what I have done:

1. Add extra voltage to SoC and DRAM  - via BIOS. 

2. Disable fast startup

3. Hardwere never sleeps

4. Disabled motherboard audio in BIOS. Run everthing on NVIDIA Display Port audio. 

5. Diasbled NVIDIA audio draver. 

 

Done everthing step by step. If Black screen appears, moved to next step.

 

!!!!

Only what works is - Undercloacking GPU via MSI Aterburner. 

 

Generally launch MSI-Afterburner in default settings and run OC scan. 

 

After aprox 22 min it comes back with results. Mine MSI Afterburner said that it is not stable -power issues. 

 

To much Core clock - recommendation is -100Mhz Core clock

-100Mhz - Ram clock.

 

I done it, and no more black screens. ATM  - 3days. 

 

my kid plays quite a lot COD  - modern warfere, he said no more black screens. 

 

Next week I will try to instal new Win10 - Enterprice os1903. We will see how it goes. 

 

kind regards, 

Okutida

 

[SM Phoo]

@Okutida

The most frustrating thing for me is everytime I think I narrow it down to a cause, it surprises me.

 

Today, I was able to borrow a GPU for the weekend, so hardware testing was the plan.

 

I first loaded up Apex as yesterday it was being consistent with the crashing, and it crashed almost instantly, so that was a good thing as it gave me some consistency to work with.

My first test was checking the cablemods cable.

So I ran Apex again, it crashed after 2minutes, great, consistency.

So I unhooked the cablemod extension cable, plugged in the standard, ran apex and it was running fine for around 15minutes. Fantastic, I think I've found the cause, but I want to make sure, so I hook up the extension again, ran the game, crashed quickly, so finally I think I've narrowed it down. I'll try again just to really confirm what I think is the problem... 

 

This time, the game ran fine for 30mins. I do a restart, test again, runs fine, so now I am back to not knowing. 

The crash is so inconsistent it's hard to even test.

 

I only have the GPU for the weekend, but if I cant replicate the crashing consistently enough, it's hard to pin down a cause

[SM Phoo]

@Okutida

Hey, I dunno if you're still encountering the same problem, but I think I've finally got to the bottom of mine... I hope.

 

I had to update my audio driver from the motherboard manufacturer.

Whenever I did searches, people did say a bad audio driver, but I just assumed windows always found the newest drivers. Apparently not always, especially when it comes to audio.

 

I've not had a single crash since, and I've really tried to push the system to have one, so I am somewhat confident I may have sorted mine

[Okutida]

Hi. 

@SM Phoo

Thanks I will try. 

 

Last thing what I have done is - instaled new Win10 Enterprise -1903. 

But must addmit - black screens appaers more often. 

Even underclocking via MSI afterburner  - does not help. For example - Win10 Pro when using MSI -Aer, if I underclocked -200Mhz, no more black screens.

Unfortunatly it does not work for Win10-enterprice.

 

Please keep me informed how it goes with - sound - audio driver thing. 

 

Kind regards, 

Okutida

[SM Phoo]

@Okutida

I am now away until after Christmas, so I wont be able to test anymore until then sorry, but it's been perfect for over a week, so I am hopeful.

Definitely give it a go and let me know if it solves your issue also

[Okutida]

Hi. 

@SM Phoo

 

Yes, I already done it (today instaled audio driver from motherboard - Asus web site). Now is just to test it out. 

 

Have a great Xmas!

 

Lets keep in touch.

 

Okutida