Question PC crashing/freezing with no BSOD, getting kernel power errors, usually under moderate load ?

[RSPC]

Hey all, first time posting in a forum like this, apologies if any of this is out of order, i usually am able to solve my issue by googling it, but I've been stuck for over a week now. I feel like there may be enough info to kinda pinpoint whats wrong, but I'm not tech savvy enough to figure it out

recently I have been having some issues with my PC restarting/crashing/freezing with no BSOD;
- When its under heavy-ish load like streaming a game it will randomly just reset, no BSOD, just black screen then start again like nothing happened. Its more likely to happen during a game that uses a fair amount of resources, but I've had it happen while trying to play a browser game. here is a log from my last stream attempt from OBS
- Sometimes instead of resetting it will freeze and i have to manually reset it. in both cases i see some kernel power errors in the event viewer, but nothing else out of the ordinary
- Left the PC on overnight idling once and it reset, i don't think it was a windows update

Other issues I've been noticing, uncertain if its 100% related to the crashing:
- I've noticed when playing Overwatch that the game will randomly freeze for a couple seconds, this happens around maybe once an hour. i will still hear the audio fine, then it will almost fast-forward to catch up. I noticed during this time that the GPU usage seemed to significantly drop, though i don't know if that is a cause or an affect. but i did notice last time it happened i got windows error reporting 1001 in the event viewer, which pointed to overwatch for a radar_pre_leak_64
- Sometimes website will not load, but its not an internet issue, it just... stops trying to load the page if that makes sense?
- Moving my mouse quickly over a chat like on twitch causes the animated emotes to get real hitchy

the noteworthy events in event viewer i was getting are:
- Kernel power 41 (63)
- WHEA logger 18 APIC 1, 7, 13, 11, 15
- kernel power 172 (203)

Specs:
CPU: AMD RYZEN 7 3800XT
MOBO: MSI MPG B550 Gaming Carbon Wifi (MS-7C90)
GPU: GIGABYTE AORUS GeForce RTX 3070 Master 8G
RAM: 2x16 Corsair Vengeance cmk32gx4m2b3200c16
Storage: Samsung SSD 850 evo 500GB
Samsung SSD 870 evo 4TB
PSU: Seasonic PRIME TX-1000
Monitors: 2x ASROCK PG32QF2B Phantom 32"
Windows 10 home
(Most of these parts are 3 years old or newer, except the samsung 850 iirc)

other than these issues it appears to run perfectly fine. Maybe a bit overkill on the PSU, but with the streaming equipment i have as well as my drawing tablet, i wanted to make sure I have enough power for everything.

- Been closely looking at temps and nothing seems to be overheating, or running at an insane %. CPU mhz got up there from what i saw on OCCT, but apparently that's pretty common for my CPU. tried running stress tests on OCCT, everything was fine, furmark was fine too.
- This was a fresh install of windows as i had about a month ago had bsod loop which the only solution was to do a fresh install of windows.
- Reinstalled graphics driver using DDU
- All drivers are up to date
- re-seated parts and made sure the connections were good, didn't see any signs of fraying cords or burn marks
- reset CMOS, which changed some bios setting like the mhz on the ram, system seems to run the same but now I'm seeing a Ntfs warning and error in event viewer (55 + 98) which says that my C drive has some corruption and requires a chkdsk scan, after doing so and restarting the pc i get the error and warning again
- have not updated bios yet, that is the next thing i plan on doing once i get my hands on a flash drive, though hoping i can avoid having to deal with more bios stuff.
- not overclocking anything as far as im aware

From what i've scoured online it seems like it may be a hardware issue, but unclear what part the issue may be, feel like i'm about at the point where i got to swap parts out. I was hoping you might be able to help me narrow this down and save me from this despair. thanks!

 

[triplex1]

Hey all, first time posting in a forum like this, apologies if any of this is out of order, i usually am able to solve my issue by googling it, but I've been stuck for over a week now. I feel like there may be enough info to kinda pinpoint whats wrong, but I'm not tech savvy enough to figure it out

recently I have been having some issues with my PC restarting/crashing/freezing with no BSOD;
- When its under heavy-ish load like streaming a game it will randomly just reset, no BSOD, just black screen then start again like nothing happened. Its more likely to happen during a game that uses a fair amount of resources, but I've had it happen while trying to play a browser game. here is a log from my last stream attempt from OBS
- Sometimes instead of resetting it will freeze and i have to manually reset it. in both cases i see some kernel power errors in the event viewer, but nothing else out of the ordinary
- Left the PC on overnight idling once and it reset, i don't think it was a windows update

Other issues I've been noticing, uncertain if its 100% related to the crashing:
- I've noticed when playing Overwatch that the game will randomly freeze for a couple seconds, this happens around maybe once an hour. i will still hear the audio fine, then it will almost fast-forward to catch up. I noticed during this time that the GPU usage seemed to significantly drop, though i don't know if that is a cause or an affect. but i did notice last time it happened i got windows error reporting 1001 in the event viewer, which pointed to overwatch for a radar_pre_leak_64
- Sometimes website will not load, but its not an internet issue, it just... stops trying to load the page if that makes sense?
- Moving my mouse quickly over a chat like on twitch causes the animated emotes to get real hitchy

the noteworthy events in event viewer i was getting are:
- Kernel power 41 (63)
- WHEA logger 18 APIC 1, 7, 13, 11, 15
- kernel power 172 (203)

Specs:
CPU: AMD RYZEN 7 3800XT
MOBO: MSI MPG B550 Gaming Carbon Wifi (MS-7C90)
GPU: GIGABYTE AORUS GeForce RTX 3070 Master 8G
RAM: 2x16 Corsair Vengeance cmk32gx4m2b3200c16
Storage: Samsung SSD 850 evo 500GB
Samsung SSD 870 evo 4TB
PSU: Seasonic PRIME TX-1000
Monitors: 2x ASROCK PG32QF2B Phantom 32"
Windows 10 home
(Most of these parts are 3 years old or newer, except the samsung 850 iirc)

other than these issues it appears to run perfectly fine. Maybe a bit overkill on the PSU, but with the streaming equipment i have as well as my drawing tablet, i wanted to make sure I have enough power for everything.

- Been closely looking at temps and nothing seems to be overheating, or running at an insane %. CPU mhz got up there from what i saw on OCCT, but apparently that's pretty common for my CPU. tried running stress tests on OCCT, everything was fine, furmark was fine too.
- This was a fresh install of windows as i had about a month ago had bsod loop which the only solution was to do a fresh install of windows.
- Reinstalled graphics driver using DDU
- All drivers are up to date
- re-seated parts and made sure the connections were good, didn't see any signs of fraying cords or burn marks
- reset CMOS, which changed some bios setting like the mhz on the ram, system seems to run the same but now I'm seeing a Ntfs warning and error in event viewer (55 + 98) which says that my C drive has some corruption and requires a chkdsk scan, after doing so and restarting the pc i get the error and warning again
- have not updated bios yet, that is the next thing i plan on doing once i get my hands on a flash drive, though hoping i can avoid having to deal with more bios stuff.
- not overclocking anything as far as im aware

From what i've scoured online it seems like it may be a hardware issue, but unclear what part the issue may be, feel like i'm about at the point where i got to swap parts out. I was hoping you might be able to help me narrow this down and save me from this despair. thanks!

if you get errors in the processor with OCCT, check your memories, remove and try again one by one

 

[RSPC]

if you get errors in the processor with OCCT, check your memories, remove and try again one by one

didnt see any errors with any OCCT tests, do you think it is a ram issue?

 

[triplex1]

didnt see any errors with any OCCT tests, do you think it is a ram issue?

run a memory test, the first thing i think i would look at is the memory then i would go to the graphics card you have

 

[RSPC]

run a memory test, the first thing i think i would look at is the memory then i would go to the graphics card you have

ran memtest86+ all day, everything passed. im assuming ram is most likely not the culprit. i ran furmark and the gpu stress test in OCCT and had no issues, is that sufficient enough evidence to rule out the gpu?

 

[triplex1]

ran memtest86+ all day, everything passed. im assuming ram is most likely not the culprit. i ran furmark and the gpu stress test in OCCT and had no issues, is that sufficient enough evidence to rule out the gpu?

the next thing you look at if you find the first 2 ok , is your currents and basically the power supply, which I find a bit difficult to fault with the specific one you have.Let me ask you something else ..... with what cooling do you have the 3800XT?

 

[triplex1]

also try to isolate the hard drives you have and leave only the one with windows

 

[RSPC]

the next thing you look at if you find the first 2 ok , is your currents and basically the power supply, which I find a bit difficult to fault with the specific one you have.Let me ask you something else ..... with what cooling do you have the 3800XT?

i cant seem to find any documentation that tells me what model (even in my receipts), but it is a Deep Cool, and i remember the model being very highly rated, cpu temps have been good from what I've seen.

Bit of an update though:
Updated my bios, haven't done tested to see if the crash is still happening, plan on doing that tonight, however there were some things that i noticed that might be worth mentioning?
-first is something that was happening occasionally before the bios update but it happened on first load up after the update is that sometimes my taskbar doesn't fully load, as in most of it is just black. works fine, seems to be just visual, but may be indicative of something?
-looking through event viewer after the update i noticed some new things that caught my eye, first being

kernal-pnp warning 225 (223)​

The application \Device\HarddiskVolume4\Windows\System32\audiodg.exe with process id 3268 stopped the removal or ejection for the device HDAUDIO\FUNC_01&VEN_10EC&DEV_0B00&SUBSYS_1462EC90&REV_1000\5&4c28ee1&0&0001.​

as well as

Kernel-general info 16​

The access history in hive ??\C:\ProgramData\Microsoft\Provisioning\Microsoft-Desktop-Provisioning-Sequence.dat was cleared updating 0 keys and creating 0 modified pages.​

last thing i saw was

Error 86 Certificateservicesclient-certenrollSCEP​

Certificate enrollment initialization for WORKGROUP\DESKTOP-ANHOT0F$ via https://amd-keyid-907d65e9b562315997dd5ad086b2b7598957b92c.microsoftaik.azure.net/templates/Aik/scep failed:GetCACapsGetCACaps: Not Found{"Message":"The authority "amd-keyid-907d65e9b562315997dd5ad086b2b7598957b92c.microsoftaik.azure.net" does not exist."}HTTP/1.1 404 Not FoundDate: Sat, 29 Jun 2024 01:00:47 GMTContent-Length: 121Content-Type: application/json; charset=utf-8X-Content-Type-Options: nosniffStrict-Transport-Security: max-age=31536000;includeSubDomainsx-ms-request-id: 0b90392c-0bc0-4bfc-b594-0f2b1c4b4a83Method: GET(328ms)Stage: GetCACapsNot found (404). 0x80190194 (-2145844844 HTTP_E_STATUS_NOT_FOUND)​

​

i ran dism scan which showed "the component store is repairable", in which i ran the dism repair which seemed to fix it? ran sfc after and restarted and it looks like the first 2 things may be gone but the error 86 stayed.

pc seems to be running about the same all and all, though i will give it some testing rn... does this info help diagnose this in any way? i did notice some people in forums with similar issues to what i have been having when i googled the errors and warnings...

 

[triplex1]

you have passed the latest drivers for the chipset?
I would do a test to do a clean installation of windows on another disk ssd
to see what's really happening

 

[triplex1]

here amd chipset drivers

 

[triplex1]

after the drivers one last thing from me..........if it continues the same go to bios and disable the tpm

 

[triplex1]

I also insist on the cooling system you have, at least I would open it, clean it and apply new thermal paste (I've only been using mx-'19 for the last 5-6 years).if you can, I would like you to upload a photo of the cooler you have so I can see what it is...because the processor you have is quite hot

 

[RSPC]

I also insist on the cooling system you have, at least I would open it, clean it and apply new thermal paste (I've only been using mx-'19 for the last 5-6 years).if you can, I would like you to upload a photo of the cooler you have so I can see what it is...because the processor you have is quite hot

it looks like a DeepCool Gammax C40, thermal paste was recently applied ~4 months ago, cpu temp when idle ~45C, when gaming and doing stuff no more than 70C.

I also have the latest chipset drivers from the MSI website, but i noticed the ones you linked from AMD are different, should i prioritize updating from one site over the other? i was doing it from MSI because i assumed it was more tailored for my specific board I'll make sure to try the other suggestions after we get the chipset thing sorted

I'm starting to suspect it may be my 850 evo as it appears to only have a life of ~5 years and i've had it for maybe just under double that. (as well as all the errors regarding the C drive... could that be a plausible cause for all the issues i've been seeing?

 

[triplex1]

no msi , from amd and finallly remember if continued turn tpm disabled from your bios , something has been damaged in the past without you realizing it and that's why I referred you before and for a clean install on another disk.

 

[RSPC]

no msi , from amd and finallly remember if continued turn tpm disabled from your bios , something has been damaged in the past without you realizing it and that's why I referred you before and for a clean install on another disk.

so it seems like maybe the bios plus amd drivers may have stabilized it (along with maybe the fresh installs i did of graphics/audio/other drivers)? Too early for me to say for certain but i had 2 sessions without is crashing, i will update when im more confident this is solved.

when i updated my bios i noticed my default ram frequency was set lower than what it is advertised at, 2133 mhz vs 3200, is that something i should change to match or would that be inviting instability on my system?

Either way appreciate the help, fingers crossed this sticks

 

[triplex1]

so it seems like maybe the bios plus amd drivers may have stabilized it (along with maybe the fresh installs i did of graphics/audio/other drivers)? Too early for me to say for certain but i had 2 sessions without is crashing, i will update when im more confident this is solved.

when i updated my bios i noticed my default ram frequency was set lower than what it is advertised at, 2133 mhz vs 3200, is that something i should change to match or would that be inviting instability on my system?

Either way appreciate the help, fingers crossed this sticks

every time you update bios , everything goes back to default settings........you will have to disable xmp and especially AMDftpm , which after the update is automatically activated

 

[RSPC]

every time you update bios , everything goes back to default settings........you will have to disable xmp and especially AMDftpm , which after the update is automatically activated

xmp wasnt on when i reset, but what is amdftpm and what do i get from disabling it? will that bump the ram frequency up to the advertised number?

 

[triplex1]

xmp wasnt on when i reset, but what is amdftpm and what do i get from disabling it? will that bump the ram frequency up to the advertised number?

AMD CPU fTPM is a trusted module of AMD for its CPU hardware protection. This module uses inside the system firmware instead of installing a chip. You can say that AMD fTPM is an extended form of ordinary or standard TPM used in the form of microchips.
before you save anyone else in the bios and go out, you can put the xmp profile of the memories

 

[RSPC]

okay so update, things were looking alright for a little bit, games were still getting the occasional freeze for a couple seconds, but i wasnt getting any crashes, even with streaming. still have yet to get a crash under heavy load, but now my system seems to be crashing randomly, a lot of times when i leave it idle, though i just had a crash happen to me and there was no bsod, the screens turned black and it reset as if nothing happened, i had just finished watching a youtube video not 10 seconds earlier.

I was seeing:
-kernel power 41(63), bugcheck 1001 errors on 7/8 and 7/11

-7/8 bugcheck code was 0x0000003b (0x00000000c0000005, 0xfffff805795c65c6, 0xffffd30447a11b30, 0x0000000000000000)​

-7/11 bugcheck code was 0x00000050 (0xffffe203f00fd160, 0x0000000000000002, 0xfffff8047ce38a88, 0x0000000000000000​

-kernel power 41(63), WHEA-logger 18 (APIC ID 7) errors earlier today 7/12
-kernel power 41(63), WHEA-logger 18 (APIC ID 6) errors just now 7/12

additionally i did notice that on 7/9 i got an nfts 55 error that indicates that corruption was found on the C drive, running chkdsk didnt find anything wrong with it. the unfortunate thing is i also cant check the reliability of the drive using samsung magician, due to it being an older drive.

 

[triplex1]

okay so update, things were looking alright for a little bit, games were still getting the occasional freeze for a couple seconds, but i wasnt getting any crashes, even with streaming. still have yet to get a crash under heavy load, but now my system seems to be crashing randomly, a lot of times when i leave it idle, though i just had a crash happen to me and there was no bsod, the screens turned black and it reset as if nothing happened, i had just finished watching a youtube video not 10 seconds earlier.

I was seeing:
-kernel power 41(63), bugcheck 1001 errors on 7/8 and 7/11

-7/8 bugcheck code was 0x0000003b (0x00000000c0000005, 0xfffff805795c65c6, 0xffffd30447a11b30, 0x0000000000000000)​

-7/11 bugcheck code was 0x00000050 (0xffffe203f00fd160, 0x0000000000000002, 0xfffff8047ce38a88, 0x0000000000000000​

-kernel power 41(63), WHEA-logger 18 (APIC ID 7) errors earlier today 7/12
-kernel power 41(63), WHEA-logger 18 (APIC ID 6) errors just now 7/12

additionally i did notice that on 7/9 i got an nfts 55 error that indicates that corruption was found on the C drive, running chkdsk didnt find anything wrong with it. the unfortunate thing is i also cant check the reliability of the drive using samsung magician, due to it being an older drive.

as far as I know, everything you mention is due to memories, find a pair from someone you know. or your friend and take a test

 

[PiersPlowman]

In the BIOS, toggle the RAM back to the default, non-XMP/DOCP, settings to ensure stability.

 

[RSPC]

okay been a while since the last update, managed to try some new ram which is on the list of approved sticks for the mobo, and the perf and stability of my pc seems about the same. havent crashed on the new sticks, but i havent used them for super long, still seems like something is not right with the pc tho... it feels like its chugging more than it should, especially on the web browser, sometimes i gotta refresh the page 3 or 4 times just for the YT thumbnails to load correctly

Had a crash before the switch where there was no bsod while i was drawing. whocrashed tells me that was due to IRQL not less or equal, and "This is a typical software problem. Most likely this is caused by a bug in a driver."... could have been just my drawing tablet messing up tho because the pen stopped working right before it crashed. also saw bugcheck 1001 in the event viewer 0x0000000a (0x0000000000000020, 0x0000000000000002, 0x0000000000000000, 0xfffff80573accce0). im also noticing:

- a significant amount of distributedCOM warnings and errors, but i've been told those arent really an issue
- some Schannel 36881 errors have been popping up recently
- corruptions keep being found on the C: drive (ntfs error 55)

im thinking this is not an issue with the ram sticks themselves but something else... any of this info a help?

 

[triplex1]

''
especially on the web browser, sometimes i gotta refresh the page 3 or 4 times just for the YT thumbnails to load correctly''

this has more to do with the graphics than with the RAM