Advice please - AMD Radeon RX 5700 XT - PC Restarts

[alexburn]

For the system to restart under any GPU load would indicate to me either power supply and/or graphics card related issues. I’d definitely clean and redo the thermal paste as mentioned by @eidairaman1

When you are in the BIOS next, please get a picture of the systems voltages if you can.

I'll look at reseating with new thermal paste next.
Photos of the bios page attached.

Run memtest86 on your system to find out if memory is unstable.

Can you provide us the complete system specs (cpu, ram, motherboard)

Memtest86 completed and passed (XMP disabled)

Have you disabled Automatically Restart option under Startup and Recovery / System Failure?

It wasn't disabled, but it is now. Next time it crashes, I'll post the blue screen.

 

[ShrimpBrime]

XMP was loaded in bios until I bricked Windows. One of the things I did when reinstalling everything was reset BIOS to default which included disabling XMP profile.

Besides... It was running stock without XMP perfectly before. If all tests pass, then I'll enable XMP and run again.

Sounds like a clear cmos and then fresh os install for testing. Maybe windows is fubar again.

 

[alexburn]

Sounds like a clear cmos and then fresh os install for testing. Maybe windows is fubar again.

I've literally just done that yesterday. I'm back to a fresh running state again, just with random reboots/crashes.

 

[eidairaman1]

I've literally just done that yesterday. I'm back to a fresh running state again, just with random reboots/crashes.

Blue screen view

 

[alexburn]

Blue screen view

Yep, at work for the next 10 hours so will look at this again when I'm back.

 

[eidairaman1]

Yep, at work for the next 10 hours so will look at this again when I'm back.

Ok i will be back then

 

[Ruru]

Sorry for being lazy and not reading the whole thread, but do you have an another PC to test the card with?

After 20+ years of tinkering with computers, incompatibility/software issues are always my number 1 suspects.

e: ah, looks like @eidairaman1 is already helping, dude you are pure gold.

 

[eidairaman1]

Sorry for being lazy and not reading the whole thread, but do you have an another PC to test the card with?

After 20+ years of tinkering with computers, incompatibility/software issues are always my number 1 suspects.

e: ah, looks like @eidairaman1 is already helping, dude you are pure gold.

Any ideas are good

 

[alexburn]

I've just got home to run a test. Ran Benchmark, no issues. Loaded up COD and it crashed before the game even started. BSOD states KERNEL MODE HEAP CORRUPTION. Restarted machine, and it crashed on boot up... BSOD states PAGE FAULT IN NONPAGED AREA.
Any ideas?!

 

[eidairaman1]

Test card in another system if possible

 

[alexburn]

Test card in another system if possible

Not easy to do unfortunately... Will have to wait until I see my mate at the weekend to see if I can test it on his system.

New problem... When I load windows up now, it takes ages. When it finally loads, I can only see the desktop, no task bar or start menu! Looks like more corruption!

Update: wouldn't even start after I had to turn off the power due to it going unresponsive. Ended up with the same "could not repair" blue screen. However, this time, I was able to go back to an old restore point. I'm now going through the process of removing the GPU entirely to try and eliminate that.

Update2: lots of different blue screens, even after removing the GPU. I'm now going through the process of reinstalling windows again!

 

[VuurVOS]

Not easy to do unfortunately... Will have to wait until I see my mate at the weekend to see if I can test it on his system.

New problem... When I load windows up now, it takes ages. When it finally loads, I can only see the desktop, no task bar or start menu! Looks like more corruption!

Update: wouldn't even start after I had to turn off the power due to it going unresponsive. Ended up with the same "could not repair" blue screen. However, this time, I was able to go back to an old restore point. I'm now going through the process of removing the GPU entirely to try and eliminate that.

Update2: lots of different blue screens, even after removing the GPU. I'm now going through the process of reinstalling windows again!

Did you run memtest86 with XMP enabled?

 

[OolongGoat]

I finally got the pictures to load. I would definitely service the video card. I would have your mate help if they can in redoing the paste and pads on the card. A 45C delta is quite high, even for an AMD card.

Also, many of those blue screens correspond to data corruption in memory or an error with memory. When you ran Memtest did you just do the one pass or did you let it run overnight??

I would start with removing all the RAM and testing one stick at a time, one slot at a time. I did a similar thing to my brother’s system a few months ago. The RAM proved faulty.

He still had some gremlins running about but the system was much more stable than previously.

By doing the RAM one stick at a time, one slot at a time, you are also testing the motherboard.

 

[dirtyferret]

I would start with removing all the RAM and testing one stick at a time, one slot at a time. I did a similar thing to my brother’s system a few months ago. The RAM proved faulty.

+1, although weird both sticks worked fine all this time (unless his friend didn't tell him something) as any bad RAM I've had tends to be bad from day one

 

[Veseleil]

Check your NVMe health with Crystal disk info or similar. Did you reinstalled your OS after formatting the system drive (C: )?

 

[ShrimpBrime]

Not easy to do unfortunately... Will have to wait until I see my mate at the weekend to see if I can test it on his system.

New problem... When I load windows up now, it takes ages. When it finally loads, I can only see the desktop, no task bar or start menu! Looks like more corruption!

Update: wouldn't even start after I had to turn off the power due to it going unresponsive. Ended up with the same "could not repair" blue screen. However, this time, I was able to go back to an old restore point. I'm now going through the process of removing the GPU entirely to try and eliminate that.

Update2: lots of different blue screens, even after removing the GPU. I'm now going through the process of reinstalling windows again!

From those blue screens.....

I would try a different cpu first. Then RAM, board last.

Typically see these types of blue screens when I OC memory and memory controller.

Need parts to test with unfortunately. Or just go ahead and RMA all 3 and wait.

GL!!

 

[eidairaman1]

I've just got home to run a test. Ran Benchmark, no issues. Loaded up COD and it crashed before the game even started. BSOD states KERNEL MODE HEAP CORRUPTION. Restarted machine, and it crashed on boot up... BSOD states PAGE FAULT IN NONPAGED AREA.
Any ideas?!

H

I finally got the pictures to load. I would definitely service the video card. I would have your mate help if they can in redoing the paste and pads on the card. A 45C delta is quite high, even for an AMD card.

Also, many of those blue screens correspond to data corruption in memory or an error with memory. When you ran Memtest did you just do the one pass or did you let it run overnight??

I would start with removing all the RAM and testing one stick at a time, one slot at a time. I did a similar thing to my brother’s system a few months ago. The RAM proved faulty.

He still had some gremlins running about but the system was much more stable than previously.

By doing the RAM one stick at a time, one slot at a time, you are also testing the motherboard.

Data os corruption, Ram, SSD or motherboard/cpu...

 

[alexburn]

Thanks for the suggestions guys... I just broke my only spare USB flash drive, so need to buy a new one tomorrow to do the XMP enabled memtest and also to restore an uncorrupt version of Windows.

I'm midway through a benchmark test with just 8gb ram and no GPU and it's passed without crashing (albeit obviously a really low score). I think my first step will be to test each of the 4 sticks of RAM individually and see if any of them cause an issue.

 

[ShrimpBrime]

Thanks for the suggestions guys... I just broke my only spare USB flash drive, so need to buy a new one tomorrow to do the XMP enabled memtest and also to restore an uncorrupt version of Windows.

I'm midway through a benchmark test with just 8gb ram and no GPU and it's passed without crashing (albeit obviously a really low score). I think my first step will be to test each of the 4 sticks of RAM individually and see if any of them cause an issue.

Yeah, probably be ok with 2 occupied in A2 and B2 slots. Memory controller doesn't like your memory lol.

Figured you already did all this stuff before came to ask help meh.

 

[alexburn]

Yeah, probably be ok with 2 occupied in A2 and B2 slots. Memory controller doesn't like your memory lol.

Figured you already did all this stuff before came to ask help meh.

I literally had no idea where to start! Still don't, but hoped someone here might have a better idea than me! ‍

 

[ratirt]

These artifacts on boot with the blue screen that is some serious crap. My bet is PSU problem since I have had same thing happening. I tried everything and it turned out to be PSU. I also needed to reinstall everything.
The other thing would be the thermal pads since this seems strictly GPU problem but like i said, it might be PSU. It is hard for me to believe it is actually thermal pads due to the severity of the issue. Win boot problems? That is unlikely the pads but worth a shot.
I would check the CPU as well since, someone mentioned, iGPU has problems as well?
If you havent done it yet. Clean your entire PC. Put it apart, clean and reassemble. Clean win install and try. If your problems wont go away, try swapping PSU if you've got one spare or a friend that can lend you one.
You can also try your GPU with a friend's computer to see if it works ok if that is an option.

 

[alexburn]

Update...

Completed 3 tests using memtest86 with XMP enabled... All 3 passed, however, the PC did reboot half way through the 2nd test and came up with the blue screen saying that automatic repair failed. But when I restarted and booted back to USB, the memory test continued where it left off and passed. It then passed 3rd time with no problem. Strange!

I've had the GPU removed all day now. Initially, I had no issues, no crashes, all fine. But as the day went on, it has crashed around 3 times now, usually when restarting the machine. I'm leaving the GPU out for now, but I do have a spare GPU available now to do some testing with.

I've also tried using the system with 1 of the 4 sticks of RAM at a time, swapping them in turn to test them all. No issues experienced with any one of them on their own.

Other than swapping the PSU... Any other thoughts for next step on testing? Presumably, all that will be left after that is the motherboard or NVMe drive.

These artifacts on boot with the blue screen that is some serious crap. My bet is PSU problem since I have had same thing happening. I tried everything and it turned out to be PSU. I also needed to reinstall everything.
The other thing would be the thermal pads since this seems strictly GPU problem but like i said, it might be PSU. It is hard for me to believe it is actually thermal pads due to the severity of the issue. Win boot problems? That is unlikely the pads but worth a shot.
I would check the CPU as well since, someone mentioned, iGPU has problems as well?
If you havent done it yet. Clean your entire PC. Put it apart, clean and reassemble. Clean win install and try. If your problems wont go away, try swapping PSU if you've got one spare or a friend that can lend you one.
You can also try your GPU with a friend's computer to see if it works ok if that is an option.

Just to confirm, the PC has been cleaned (was pretty clean anyway) and it has a fresh Win11 install. Although I'm considering refreshing the Win 11 again anyway, given the recent crashing and suspected corruption. I've got new thermal paste for the GPU, but as that's been out all day, I doubt that will do anything now.

Update...

Completed 3 tests using memtest86 with XMP enabled... All 3 passed, however, the PC did reboot half way through the 2nd test and came up with the blue screen saying that automatic repair failed. But when I restarted and booted back to USB, the memory test continued where it left off and passed. It then passed 3rd time with no problem. Strange!

I've had the GPU removed all day now. Initially, I had no issues, no crashes, all fine. But as the day went on, it has crashed around 3 times now, usually when restarting the machine. I'm leaving the GPU out for now, but I do have a spare GPU available now to do some testing with.

I've also tried using the system with 1 of the 4 sticks of RAM at a time, swapping them in turn to test them all. No issues experienced with any one of them on their own.

Other than swapping the PSU... Any other thoughts for next step on testing? Presumably, all that will be left after that is the motherboard or NVMe drive.


Just to confirm, the PC has been cleaned (was pretty clean anyway) and it has a fresh Win11 install. Although I'm considering refreshing the Win 11 again anyway, given the recent crashing and suspected corruption. I've got new thermal paste for the GPU, but as that's been out all day, I doubt that will do anything now.

Update 2:
Just as I posted the last messages, I got another BSOD.
Stop code 0xc000021a. A quick Google says this is yet another corruption!

 

[eidairaman1]

Try aUpdate...

Completed 3 tests using memtest86 with XMP enabled... All 3 passed, however, the PC did reboot half way through the 2nd test and came up with the blue screen saying that automatic repair failed. But when I restarted and booted back to USB, the memory test continued where it left off and passed. It then passed 3rd time with no problem. Strange!

I've had the GPU removed all day now. Initially, I had no issues, no crashes, all fine. But as the day went on, it has crashed around 3 times now, usually when restarting the machine. I'm leaving the GPU out for now, but I do have a spare GPU available now to do some testing with.

I've also tried using the system with 1 of the 4 sticks of RAM at a time, swapping them in turn to test them all. No issues experienced with any one of them on their own.

Other than swapping the PSU... Any other thoughts for next step on testing? Presumably, all that will be left after that is the motherboard or NVMe drive.


Just to confirm, the PC has been cleaned (was pretty clean anyway) and it has a fresh Win11 install. Although I'm considering refreshing the Win 11 again anyway, given the recent crashing and suspected corruption. I've got new thermal paste for the GPU, but as that's been out all day, I doubt that will do anything now.


Update 2:
Just as I posted the last messages, I got another BSOD.
Stop code 0xc000021a. A quick Google says this is yet another corruption!

Try a fresh storage drive, igp only, possible the cpu or motherboard are bunk

 

[ShrimpBrime]

Update...

Completed 3 tests using memtest86 with XMP enabled... All 3 passed, however, the PC did reboot half way through the 2nd test and came up with the blue screen saying that automatic repair failed. But when I restarted and booted back to USB, the memory test continued where it left off and passed. It then passed 3rd time with no problem. Strange!

I've had the GPU removed all day now. Initially, I had no issues, no crashes, all fine. But as the day went on, it has crashed around 3 times now, usually when restarting the machine. I'm leaving the GPU out for now, but I do have a spare GPU available now to do some testing with.

I've also tried using the system with 1 of the 4 sticks of RAM at a time, swapping them in turn to test them all. No issues experienced with any one of them on their own.

Other than swapping the PSU... Any other thoughts for next step on testing? Presumably, all that will be left after that is the motherboard or NVMe drive.


Just to confirm, the PC has been cleaned (was pretty clean anyway) and it has a fresh Win11 install. Although I'm considering refreshing the Win 11 again anyway, given the recent crashing and suspected corruption. I've got new thermal paste for the GPU, but as that's been out all day, I doubt that will do anything now.


Update 2:
Just as I posted the last messages, I got another BSOD.
Stop code 0xc000021a. A quick Google says this is yet another corruption!

Try a couple things please?

First, I would manually set System Agent to 1.1v do not go over 1.2v. This may help stabilize the memory controller. Default is 1v, I wonder if it's drooping.

Another, would be to manually check and set the timings. There should be a section in the bios that shows spd and xmp timings. Using either, be sure the timings are set manually instead of auto in the dram timings menu. Then, manually set DRAM voltage to it's spec + 0.050v.

You can put your GPU back in, we've determined it's not the issue.

 

[eidairaman1]

Try a couple things please?

First, I would manually set System Agent to 1.1v do not go over 1.2v. This may help stabilize the memory controller. Default is 1v, I wonder if it's drooping.

Another, would be to manually check and set the timings. There should be a section in the bios that shows spd and xmp timings. Using either, be sure the timings are set manually instead of auto in the dram timings menu. Then, manually set DRAM voltage to it's spec + 0.050v.

You can put your GPU back in, we've determined it's not the issue.

I'd suggest a psu first, a virgin storage drive, then cpu/motherboard.