BSOD without the BS

Posting this here because it's the closest to the problem I'm having and I'm pretty sure it isn't the GPU (or at least not significantly related to it).

Occasionally, and increasingly more often, my computer has been having fatal errors, which also take down the graphics and I believe the sound system, as well as preventing the power button on the top of my case from working (I have to switch off or unplug the computer to restart it). The lights and fans remain on, but the monitor goes dark and does not register any signal from mouse or keyboard (unsure if it is actually being conveyed to the computer or not).

The first time this happened was in a discord call, the above happened. Sound cut out immediately, as with the display. I tried plugging my HDMI cable into the non-GPU slot and it didn't help anything. I restarted the computer once using the switch, and it came on properly but the GPU was not detected by the system (in task manager or by my fan controller). I restarted it normally, and the startup screen had a blue background, with a bar of black cycling through it (the startup screen normally has a black background). It then had the same issue as the initial crash. I restarted it one more time and it was fine afterwards.

Since then I've had several crashes, including a few as soon as I walk away, and I experienced two shortly before making this post, including one a few minutes after startup. I don't think I've had this happen while I've had a game open or watching a video, and aside from the first time it did not happen during the startup phase. I've only had the audio do that angry buzz once during the crashes I was there for, despite audio having been playing for most of them.

Windows Version: 22H2 (Build 19045.2965)
Logs: MANDELBROT-(2023-05-21_19-34-42).zip - Google Drive

Updated the drivers, hoping that was the issue. I used windbg to analyze the most recent dump file, not sure what parts are important so I put it all in this pastebin

The dumps you uploaded are identical, they are all VIDEO_TDR_FAILURE bugchecks. TDR is the graphics Timeout Detection and Recovery feature, it's designed to check graphics hangs and resart the driver (and card) function to recover from the hang. If recovery isn't possible then you get this VIDEO_TDR_FAILURE BSOD.

The call stack in each dump shows what was going on, as you can see from this extract....

Code:

fffff002`68a37a18  fffff800`1b8665be dxgkrnl!TdrBugcheckOnTimeout+0xfe
fffff002`68a37a20  00000000`00000116
fffff002`68a37a28  ffffba8f`bae1c010
fffff002`68a37a30  fffff800`33143ab0 nvlddmkm+0x1073ab0
fffff002`68a37a38  ffffffff`c000009a
fffff002`68a37a40  00000000`00000004
fffff002`68a37a48  00000000`00000004
fffff002`68a37a50  ffffba8f`bae1c010
fffff002`68a37a58  fffff800`1b816b64 dxgkrnl!ADAPTER_RENDER::Reset+0x174
fffff002`68a37a60  fffff800`33143ab0 nvlddmkm+0x1073ab0
fffff002`68a37a68  ffffba8f`ae9aa050
fffff002`68a37a70  00000000`00002000
fffff002`68a37a78  ffffba8f`ae9aa110
fffff002`68a37a80  00000000`00000001
fffff002`68a37a88  fffff800`1b80f69c dxgkrnl!DXGADAPTER::Reset+0x4dc
fffff002`68a37a90  ffffba8f`ae984000
fffff002`68a37a98  00000000`01000000
fffff002`68a37aa0  00000000`00000004
fffff002`68a37aa8  00000000`00000004
fffff002`68a37ab0  ffffba8f`ba205180
fffff002`68a37ab8  00000000`00000000
fffff002`68a37ac0  ffffba8f`ba205040
fffff002`68a37ac8  ffffba8f`ae984000
fffff002`68a37ad0  fffff800`0a525440 nt!ExNode0
fffff002`68a37ad8  ffffba8f`b1a2e790
fffff002`68a37ae0  00000000`00000000
fffff002`68a37ae8  ffffba8f`bae1c010
fffff002`68a37af0  ffffba8f`b1a2e790
fffff002`68a37af8  ffffba8f`ae984000
fffff002`68a37b00  ffffba8f`bae1c010
fffff002`68a37b08  fffff800`1b865ce5 dxgkrnl!TdrResetFromTimeout+0x15
fffff002`68a37b10  00000000`00000100
fffff002`68a37b18  ffffba8f`ae984a70
fffff002`68a37b20  00000000`a834ded0
fffff002`68a37b28  fffff800`09ad443c nt!PfSnNameQueryWorker+0xdc
fffff002`68a37b30  ffffba8f`ba205040
fffff002`68a37b38  fffff800`1b865e57 dxgkrnl!TdrResetFromTimeoutWorkItem+0x27
fffff002`68a37b40  fffff800`0a525440 nt!ExNode0

These are push-down stacks, so you read them from bottom up. The dxgkrnl.sys driver is the Windows DirectX kernel, it will cooperate with the Nvidia graphics driver (nvlddmkm.sys) to actually get the display on the graphics card. In the stack above you see the dxgkrnl!TdrResetFromTimeoutWorkItem TDR function detecting a hang (a timeout). Then we see the dxgkrnl!TdrResetFromTimeout TDR function reset the dxgkrnl.sys driver. Later we see the dxgkrnl!DXGADAPTER::Reset function resetting the graphics card (the adapter). Then the Nvidia nvlddmkm.sys graphics driver is called (probably to retry the operation) and the dxgkrnl!ADAPTER_RENDER::Reset function resets the render operation on the graphics card. We see the Nvidia graphics driver called again (probably to abandon the operation) and finally the dxgkrnl!TdrBugcheckOnTimeout is the TDR bugcheck being generated.

All thse dumps show a similar pattern, so the issue is either in the Nvidia graphics driver or in the card itself. If neither the current driver version, nor the two immediately prior versions of the driver fix the problem (using DDU in between each driver install to remove the older version) then it's either the graphics card, the motherboard, or possibly a PSU problem.

Alright, thank you!

Hearsepilot said:

Do you have the latest available Video Card drivers? No one is likely to download logs so you'l have to post the relevant bits back here.

the logs are from the v2 log collector tool here, so, yes, the logs most definitely will get opened....