Question Is the GPU potentially flawed, or port, or... ? (driver timeout)

[Bjarki22]

Hello everyone,

Maybe a week ago I bought a new PC with RX 6600. Only a few days later, lightning struck somewhere in the city and I know for a fact that ports on my switch are dead, my router was as well.

I took a look at my PC, didn't see anything wrong and unplugged it right away. Now I'm wondering, since electricity travelled through my house, could it have damaged my GPU?

I played very intensive PC Games all day (RDR2, Warzone etc.), and had no problems, FPS was just about the same. However, as I was watching a video on a certain website using Chrome, I got a black screen for a second and "AMD Driver timeout" message.

I immediately panicked and downloaded all the benchmark tools and ran them for some time, and it was all perfect. Temperatures were beautiful at 100% utilization. I then go back to Chrome, to the same website and it happened again.
A guy on discord told me that maybe 3d rendering part wasn't damaged, but the video part was, which didn't make sense to me, but again I'm no expert on how GPUs work.

Next morning, I decided to change the Display Port on my GPU, and took some time to stress the entire PC for longer.

Here's the picture:

The stress test ran for 01:38:12 with VIDEO playing the whole time, and it seems the temps were fine throughout the whole process, and no driver timeouts happened.

My question is, does this test actually prove that GPU and PC are healthy? I mean, that it wasn't a hardware issue? Or could it have been that the port that was plugged in was damaged?

Does it actually make sense that it can run for this long without problems and still be a gpu problem?

Help me ease my mind lol

 

[Newtonius]

What your friend on Discord meant is that there are sections of your GPU die that might be damaged, like the video decoder portion, but not your 3D renderer. A GPU die looks something like this (this may or may not represent what your particular GPU die looks like):

Notice the Video decode portion is not part of the main graphics core portion, the decode portion is responsible for playing back video files. Transistors all across your GPU die have different functions, and just like the brain, can still work even if some parts are malfunctioning.

Try playing some video files or watching videos from multiple different sites to see if the same issue pops up. 3D render stress tests won't touch video decode/encode portions since they don't need them. Though I think it's unlikely that is the real issue, when a surge happens usually the motherboard is one of the first affected components.
Try:

• Uninstalling and reinstalling your AMD grphics drivers
• Flashing your motherboard BIOS to the same version or a newer one.
• Checking your OS storage drive for any data corruption (through command prompt in admin mode, type "sfc /scannow")

 

[Bjarki22]

What your friend on Discord meant is that there are sections of your GPU die that might be damaged, like the video decoder portion, but not your 3D renderer. A GPU die looks something like this (this may or may not represent what your particular GPU die looks like):

Notice the Video decode portion is not part of the main graphics core portion, the decode portion is responsible for playing back video files. Transistors all across your GPU die have different functions, and just like the brain, can still work even if some parts are malfunctioning.

Try playing some video files or watching videos from multiple different sites to see if the same issue pops up. 3D render stress tests won't touch video decode/encode portions since they don't need them. Though I think it's unlikely that is the real issue, when a surge happens usually the motherboard is one of the first affected components.
Try:

• Uninstalling and reinstalling your AMD grphics drivers
• Flashing your motherboard BIOS to the same version or a newer one.
• Checking your OS storage drive for any data corruption (through command prompt in admin mode, type "sfc /scannow")

Thank you very much for your informative input!

I'm wondering, is it the same if during this stress I put that 10 hour video playing as well? Are all videos run by the same portion? And if so, how likely it is that just that part caught it? Thanks!

 

[RodroX]

Hello everyone,

Maybe a week ago I bought a new PC with RX 6600. Only a few days later, lightning struck somewhere in the city and I know for a fact that ports on my switch are dead, my router was as well.

I took a look at my PC, didn't see anything wrong and unplugged it right away. Now I'm wondering, since electricity travelled through my house, could it have damaged my GPU?

I played very intensive PC Games all day (RDR2, Warzone etc.), and had no problems, FPS was just about the same. However, as I was watching a video on a certain website using Chrome, I got a black screen for a second and "AMD Driver timeout" message.

I immediately panicked and downloaded all the benchmark tools and ran them for some time, and it was all perfect. Temperatures were beautiful at 100% utilization. I then go back to Chrome, to the same website and it happened again.
A guy on discord told me that maybe 3d rendering part wasn't damaged, but the video part was, which didn't make sense to me, but again I'm no expert on how GPUs work.

Next morning, I decided to change the Display Port on my GPU, and took some time to stress the entire PC for longer.

Here's the picture:

The stress test ran for 01:38:12 with VIDEO playing the whole time, and it seems the temps were fine throughout the whole process, and no driver timeouts happened.

My question is, does this test actually prove that GPU and PC are healthy? I mean, that it wasn't a hardware issue? Or could it have been that the port that was plugged in was damaged?

Does it actually make sense that it can run for this long without problems and still be a gpu problem?

Help me ease my mind lol

Probably a stupid question but, have you tried watching this "video" of yours using Firefox for example?

Another stupid question probably, have you tried to clean up the chrome temporary files and cache?

 

[Newtonius]

Thank you very much for your informative input!

I'm wondering, is it the same if during this stress I put that 10 hour video playing as well? Are all videos run by the same portion? And if so, how likely it is that just that part caught it? Thanks!

Adding on to what RodroX wrote above: try playing that video on chrome through your CPU's video port if your CPU has an integrated GPU (Google your CPU model's name to find this out), though only if your CPU does support this - the video port will be located on the I/O portion of your motherboard (the face with all the USB ports). Do this just to confirm whether or not if it's your GPU/drivers or just chrome/windows acting up.

 

[Bjarki22]

Probably a stupid question but, have you tried watching this "video" of yours using Firefox for example?

Another stupid question probably, have you tried to clean up the chrome temporary files and cache?

Not stupid at all, actually came to mind even though it makes 0 sense to me.. I gotta say, I watched that particular video again and nothing happened.. I did clear cookies though so maybe that's that? (cookies-gpu -> really does sound stupid)

 

[Bjarki22]

Adding on to what RodroX wrote above: try playing that video on chrome through your CPU's video port if your CPU has an integrated GPU (Google your CPU model's name to find this out), though only if your CPU does support this - the video port will be located on the I/O portion of your motherboard (the face with all the USB ports). Do this just to confirm whether or not if it's your GPU/drivers or just chrome/windows acting up.

It doesn't have integrated GPU, but, I haven't had it happen to me single time this morning. Could it potentially be a display port which I changed before I booted this morning?

 

[Newtonius]

It doesn't have integrated GPU, but, I haven't had it happen to me single time this morning. Could it potentially be a display port which I changed before I booted this morning?

I wouldn't think that would be the case, especially if you were getting error messages. Hopefully one of the numerous other things you've done fixed the issue since you're not getting it anymore, but if it does happen again definitely try troubleshooting your software drivers and apps before anything. Data corruption causes issues like these all the time.

 

[Bjarki22]

Thank you both for helping out. Till today, no timeouts happened, and I'm not quite sure what "fixed" it. So hopefully that's that!