Question Plagued with random reboots ( No BSOD ).

[deantje.91]

Hi everyone,

Since a couple weeks i've been getting random restarts a lot without a BSOD. It does not create any dump files either.
My screen just goes black and reboots the system by itself. ( Kernel 41 shows in event viewer but besides that, nothing useful )
It can go hours without crashing but it can crash shortly after eachother aswell, it's completely random.

At this point I don't even know where to look anymore.

It started after swapping my GPU from 7900XT to a 7900XTX, with this I also switched from a 850W PSU to a 1000W PSU.
Since it started happening after the upgrade, the first thing I obviously replaced was the PSU. Still crashing. I tried running on the iGPU, still crashing. So that rules out the PSU and the GPU. Weird since the problems started after swapping these parts.

Ran multiple memtest86 instances and no problems there, so I suspect it's not the RAM either.

I've done a fresh installation of Win11 and installed only some testing apps, OCCT, Superposition and HWInfo. Trying to keep possibly bad software out of it.
Still crashing after this.
It also doesn't matter what I'm doing, it can crash during gaming but also when watching youtube videos or just browsing the web.
Gaming seems more stable for some reason tho, most crashes happen when on low load. Sometimes it reboots right before getting into windows aswell.
I can run Prime95 fine aswell, so I doubt it's the CPU.
Nothing is overheating, temps are well within safe ranges on every component.


Things i've tried so far:
Replace PSU
Rule out GPU by using iGPU
Fresh Win11 install on a brand new M.2 SSD and installing the newest chipset, gpu drivers etc.
Flashing BIOS to latest version.
Testing RAM through memtest and memory diagnostics

Disabling HAGS
Disabling any OC/PBO Settings I have and XMP on/off
Disabling C-states
Disabling fast startup and also turning off auto restart when running into issues ( for some reason it still restarts anyway ).
Changing powerplan to high performance and setting 5% min CPU state.
Ran SFC scan, 1 error that was fixed ( BT Driver )
Ran dskchk but also tested my 2 M.2's through Samsung Magician

Any ideas would be very welcome.

Spec list:
CPU: 7800x3D
GPU: 7900 XTX
RAM: 2x16GB Trident Z 6000Mhz cl30
PSU: Seasonic Vertex GX-1000
Mobo: B650 Tomahawk Wifi

 

[Ralston18]

When you swapped the GPU did you also remove RAM? Some motherboards require that the first physically installed RAM stick be placed in a specific slot. Check the motherboard's User Guide/Manual.

When you swapped the PSU did you use only the cables that came with the new PSU?

Look in Reliability History/Monitor and Event Viewer for any error codes, warnings, or even informational events being captured just before or at the time of the restarts.

Try Reliability History/Monitor first. Much more end user friendly and the timeline presentation can reveal patterns.

Event Viewer is more difficult to navigate and understand.

To help:

How To - How to use Windows 10 Event Viewer | Tom's Hardware Forum (tomshardware.com)

Pay attention to any error codes. You can click on any given entries for more technical details. The details may or may not be helpful.

 

[deantje.91]

When you swapped the GPU did you also remove RAM? Some motherboards require that the first physically installed RAM stick be placed in a specific slot. Check the motherboard's User Guide/Manual.

When you swapped the PSU did you use only the cables that came with the new PSU?

Look in Reliability History/Monitor and Event Viewer for any error codes, warnings, or even informational events being captured just before or at the time of the restarts.

Try Reliability History/Monitor first. Much more end user friendly and the timeline presentation can reveal patterns.

Event Viewer is more difficult to navigate and understand.

To help:

How To - How to use Windows 10 Event Viewer | Tom's Hardware Forum (tomshardware.com)

Pay attention to any error codes. You can click on any given entries for more technical details. The details may or may not be helpful.

Thanks for replying.

I did not remove the RAM when I swapped the GPU, I have reseated the RAM multiple times the last couple of days though, they are in A2 and B2, which according to the MOBO manual are the correct spots.

Yes, I only used the cables that came with the PSU.

I reinstalled windows yesterday so I only have 2 days worth of errors in the reliability monitor.
However i don't see anything out of the ordinary before or during the crashes.
I do get this warning a lot: " The application-specific permission settings do not grant Local Activation permission for the COM Server application with CLSID" ( Event 10016 DistributedCOM ) But as far as I can tell, this shouldn't be related to my issue

 

[Ralston18]

Try "dism" and then "sfc /scannow" again.

The Windows reinstall may still trying to do some updates.

Look in Update History for any failed or problem updates.

How did you reinstall Windows?

FYI:

https://forums.tomshardware.com/faq/windows-10-clean-install-tutorial.3170366

 

[deantje.91]

Try "dism" and then "sfc /scannow" again.

The Windows reinstall may still trying to do some updates.

Look in Update History for any failed or problem updates.

How did you reinstall Windows?

FYI:

https://forums.tomshardware.com/faq/windows-10-clean-install-tutorial.3170366

Ran them again but neither gave any errors/corruptions.

I did a windows install by using a USB drive created with the media creation tool.
I also had the same problem on Windows 10, which is why I went on and upgraded to Win11, hoping that would do something.

I'm gonna run Prime95 blend overnight ( if it doesn't crash ) just to be sure and will see from there.
I ordered another RAM kit to test aswell, just to be sure. But I doubt it's the RAM.

I'll be getting another GPU which is in an RMA process atm in the next couple days aswell, will test with that.
After that all that's left is the Motherboard i guess? Unless you have any other ideas.

I might try and rollback my bios version aswell, I'm currently on the latest 7D75v1E5(Beta version), might roll back to the latest non beta bios, which would be 7D75v1B and see if that makes any difference.

 

[deantje.91]

Try "dism" and then "sfc /scannow" again.

The Windows reinstall may still trying to do some updates.

Look in Update History for any failed or problem updates.

How did you reinstall Windows?

FYI:

https://forums.tomshardware.com/faq/windows-10-clean-install-tutorial.3170366

Update:

Ran prime95 for 9 hours straight, no errors, 1 hour of Furmark and 2 hours of Linpack. No problems on either.

After that I went to youtube and my PC crashed instantly. No BSOD. Just black screen and a restart so no dumps created.
I'm at a loss here. Pretty sure now that it's not hardware related. If you have any suggestions, i'd be happy to hear them.
I rolled back to a different BIOS version aswell of which i know I had no problems on.

 

[Ralston18]

No error codes, warnings, or informational events in Reliability History/Monitor and/or Event Viewer?

Does any YouTube related link cause the BSODs? No problems with other websites - correct?

Do the "YouTube BSODs" occur no matter what browser you use?

Try temporarily turning off ad blockers.

Look in Process Explorer (Microsoft, free) for any unknown or unidentifiable processes running in the background.

https://learn.microsoft.com/en-us/sysinternals/downloads/process-explorer

Also check Task Scheduler for any processes being triggered when a browser is opened.

Objective simply being to discover some cause and effect relationship: if this then that.

Right now the cause and effect is broad. If YouTube then BSOD.

Try to narrow down what is happening.

 

[deantje.91]

No error codes, warnings, or informational events in Reliability History/Monitor and/or Event Viewer?

Does any YouTube related link cause the BSODs? No problems with other websites - correct?

Do the "YouTube BSODs" occur no matter what browser you use?

Try temporarily turning off ad blockers.

Look in Process Explorer (Microsoft, free) for any unknown or unidentifiable processes running in the background.

https://learn.microsoft.com/en-us/sysinternals/downloads/process-explorer

Also check Task Scheduler for any processes being triggered when a browser is opened.

Objective simply being to discover some cause and effect relationship: if this then that.

Right now the cause and effect is broad. If YouTube then BSOD.

Try to narrow down what is happening.

Nothing useful no, all i get is the Kernel 41 critical. But nothing else useful at or before the crash time. I checked both event viewer and reliability monitor.
I don't get a BSOD, even after turning automatic reboot on errors off. Screen just goes black and reboots the system

I've had it crash this morning on this website aswell, happened once while playing cyberpunk aswell, however watching a video seems to trigger it more often. Especially when going fullscreen.
It does not matter which video it is though.

The frequency of the crashes has reduced though, not sure how but it's getting harder for me to replicate the crashes.

Checked the process explorer and don't see anything out of the ordinary.
Already checked task manager when browsing or opening youtube and nothing out of the ordinary there either.

I guess i'll just see how it goes now after reverting back to the previous BIOS.
Seen some issues here and there of people who had some problems with some BIOS' aswell.

I'm atleast happy it's nothing hardware related, atleast as far as I can tell after all the testing.

 

[Ralston18]

Power down, unplug, open the case.

Clean out dust and debris.

Verify by sight and feel that all connectors, cards, RAM, jumpers, and case connections are fully and firmly in place,

Use a bright flashlight to inspect for signs of damage: bare conductor showing, pinched or kinked wires, melted insulation, cracks, loose or missing screws, browned or blackened components. Anything at all.

Just as general matter of elimination.

 

[deantje.91]

Power down, unplug, open the case.

Clean out dust and debris.

Verify by sight and feel that all connectors, cards, RAM, jumpers, and case connections are fully and firmly in place,

Use a bright flashlight to inspect for signs of damage: bare conductor showing, pinched or kinked wires, melted insulation, cracks, loose or missing screws, browned or blackened components. Anything at all.

Just as general matter of elimination.

Thanks again for the advice.
Already did that, twice to be sure.
No noticable damages, as far as I can tell, I also cleaned everything a couple weeks ago + the components are fairly new so barely any dust.

I swapped the ram sticks with new ones just in case, back to testing it is since it crashed again even with the older BIOS.

 

[deantje.91]

Power down, unplug, open the case.

Clean out dust and debris.

Verify by sight and feel that all connectors, cards, RAM, jumpers, and case connections are fully and firmly in place,

Use a bright flashlight to inspect for signs of damage: bare conductor showing, pinched or kinked wires, melted insulation, cracks, loose or missing screws, browned or blackened components. Anything at all.

Just as general matter of elimination.

Crashed again with the new RAM, this time I uninstalled chrome from my PC just to rule out the program. Even though it couldn't just be that, but i'm trying to rule out everything just to be sure.
This time I ran youtube on MSEdge while simultaneously running a blend test on Prime95. ( max temp was 71 degrees so no temp issues. )

Again no BSOD or anything useful in either event viewer or reliability monitor.

I guess all there is left is the motherboard itself maybe?
Just seems weird to me because everything ran fine until I switched to the XTX + new PSU.
I really don't know what else it could be.

 

[deantje.91]

I think I found the cause.
Whenever I enable EXPO the crashes are way more frequent.

Does this mean a faulty Memory controller/CPU? Even though it worked fine the first couple of weeks?

Or is it just unstable EXPO settings? Any way to offset this by increasing Voltage somewhere?
Tried increasing the SOC voltage but still crashing.

 

[Ralston18]

May well be.

My thought would be to disable EXPO. Wait a few days to determine if the system becomes and stays stable.

Then, if still required, make little changes, wait a few days. If stable then take another little step. If stable - repeat.

Keep an eye on system performance and keeps notes.

 

[deantje.91]

May well be.

My thought would be to disable EXPO. Wait a few days to determine if the system becomes and stays stable.

Then, if still required, make little changes, wait a few days. If stable then take another little step. If stable - repeat.

Keep an eye on system performance and keeps notes.

Update:

Cannot for the life of me get this PC stable.
Tried upping SOC (up to 1.3v) and RAM voltages( up to 1.4v) which made the crashes less frequent( once a day and only on low load/idle).

My guess is that the memory controller on my CPU is complete garbage. I can barely run it at the recommended safe speed of 5200, let alone 6000. Even tried running it at CL32 instead.

Will probably order another CPU and send this one out for RMA. Or wait a bit for the 8000 series and just get that.

The weird thing is that on whatever RAM settings and speed I test, literally every stress test just passes. ( I ran P95 small FFTs, Large FFTs, blend for 12 hours each fine ). Also TM5 absolute gives me no errors.

But whenever I do something low load i.e. youtube or just sitting in the main menu of a game, it crashes within 10 minutes.

This makes me 2nd guess that it's a RAM issue in the first place.
However I have no idea what else it could be.

If anyone has any additional ideas, they are very welcome. Else i'll probably just give up for now, even though it's nagging me pretty badly knowing there is something wrong.

 

[mourvok]

I've been having the same issue and I also have a 7900xtx and 7800x3d, I've been seeing multiple posts about this with the same gpu and cpu I wonder if it's something to do with that.

 

[deantje.91]

I've been having the same issue and I also have a 7900xtx and 7800x3d, I've been seeing multiple posts about this with the same gpu and cpu I wonder if it's something to do with that.

Yea i cannot for the life of me figure out what it is.

I've been focusing on my RAM but i'm fairly certain it's not the RAM right now.
Pretty much every ram stress test passes without errors.
I also tried a different kit of RAM which gave me the same issue.

Everything seems fine for a while but then it ends up crashing randomly again.
It's totally random but seems to happen more on low load rather than high load. What about you?

I'm fairly certain it's not the MOBO, i ran a 7600X on the same mobo for about 4 months and I had no problems with that, together with a 7900XT.

My problems started after switching to a 7900XTX and a 7800x3D.
I'm leaning more towards a bad CPU right now so will do some testing towards that and maybe replace it if I have to.

Since it also happens during low loads/idle or even right after booting i'm kinda certain it's not the GPU aswell. It also runs fine under full load (Furmark, 3dmark etc)

What have you tried so far?

 

[mourvok]

Yea i cannot for the life of me figure out what it is.

I've been focusing on my RAM but i'm fairly certain it's not the RAM right now.
Pretty much every ram stress test passes without errors.
I also tried a different kit of RAM which gave me the same issue.

Everything seems fine for a while but then it ends up crashing randomly again.
It's totally random but seems to happen more on low load rather than high load. What about you?

I'm fairly certain it's not the MOBO, i ran a 7600X on the same mobo for about 4 months and I had no problems with that, together with a 7900XT.

My problems started after switching to a 7900XTX and a 7800x3D.
I'm leaning more towards a bad CPU right now so will do some testing towards that and maybe replace it if I have to.

Since it also happens during low loads/idle or even right after booting i'm kinda certain it's not the GPU aswell. It also runs fine under full load (Furmark, 3dmark etc)

What have you tried so far?

I saw another post saying that it was the new amd chipset drivers and that turning off freesync helped them. I'm gonna try that but if not I guess I'll have to save up for replacements

 

[mourvok]

Sadly that didn't help, now going to try a couple of things I saw on reddit that seem to be helping others with the 7800x3d
"I think I have fixed this issue for anyone that comes across this I disabled global c states in the BIOS and set the power plan in windows to high performance."

"You might need to uninstall AMD's 3D V-Cache optimizer and PPM Provisioning as this is strictly used for the 7900X3D and 7950X3D to help scheduling the 2 CCD's. Gigabyte bundles this with the chipset driver when it shouldn't. Start by installing [DriverStore Explorer](https://github.com/lostindark/DriverStoreExplorer/releases). Open the program up, scroll down to "AMD 3D V-Cache Optimizer". Force delete the driver and restart. Secondly, go into Device Manager -> System Devices. Disable "AMD PPM Provisioning File". Hopefully it helps with any scheduling issues. Cheers."

 

[Winter Roze]

I'm having the same issues with my PC now as well

Ryzen 5500
48gb trident Z ddr4 ram
RTX 3070

I've tried unplugging my fans (thought one might be shorting and tripping the current protection)

I've tried 2 other CPUs, different ram configs, 2 other GPUs (2070 super and 1070) and a different PSU and a different motherboard.

No matter what I do or what system I build with this hardware I keep having this exact same issue

 

[deantje.91]

Sadly that didn't help, now going to try a couple of things I saw on reddit that seem to be helping others with the 7800x3d
"I think I have fixed this issue for anyone that comes across this I disabled global c states in the BIOS and set the power plan in windows to high performance."

"You might need to uninstall AMD's 3D V-Cache optimizer and PPM Provisioning as this is strictly used for the 7900X3D and 7950X3D to help scheduling the 2 CCD's. Gigabyte bundles this with the chipset driver when it shouldn't. Start by installing [DriverStore Explorer](https://github.com/lostindark/DriverStoreExplorer/releases). Open the program up, scroll down to "AMD 3D V-Cache Optimizer". Force delete the driver and restart. Secondly, go into Device Manager -> System Devices. Disable "AMD PPM Provisioning File". Hopefully it helps with any scheduling issues. Cheers."

Already tried this aswell.
Yesterday I replaced my CPU altogether, this did nothing so it's not a CPU issue either.

Literally the only thing I haven't replaced is the GPU. So it's either a bad GPU or just bad drivers.
I see a lot of problems with the 7000 series in relation to drivers.
I however have tried a lot of different driver versions and none of them seem to work properly.
Surely they can't all be bad?

Mine is only crashing once in maybe 2 days atm so that's not too bad. But still annoying to know that something isn't right.

 

[deantje.91]

I'm having the same issues with my PC now as well

Ryzen 5500
48gb trident Z ddr4 ram
RTX 3070

I've tried unplugging my fans (thought one might be shorting and tripping the current protection)

I've tried 2 other CPUs, different ram configs, 2 other GPUs (2070 super and 1070) and a different PSU and a different motherboard.

No matter what I do or what system I build with this hardware I keep having this exact same issue

Sorry to hear that man, it's a pain to locate the problem.
I've been at it for about a month now and I still don't know what it is.

Have you tried reinstalling your windows?

Hardware wise, you pretty much swapped everything out. So it's not likely to be hardware related for you.

 

[mourvok]

You're gonna laugh. Well I fixed the random turning off on idle, did a completely clean windows and bios. And now the dang thing won't shut down, it just restarts every time I try to shut down

 

[mourvok]

Sorry to hear that man, it's a pain to locate the problem.
I've been at it for about a month now and I still don't know what it is.

Have you tried reinstalling your windows?

Hardware wise, you pretty much swapped everything out. So it's not likely to be hardware related for you.

You ever find a fix?