• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

Question BSOD disappears before a dump file is made

Sep 28, 2021
6
0
10
Got a problem with bluescreens of death. I can be watching youtube or playing a game, but it doesn't seem to happen under low temps. The computer recently began just quickly flashing the bluescreen and then just kinda dying. Not turning off, but just staying in an "on" state, forcing me to do a forced shutdown by holding the power button. The most recent bluescreen that didn't do this was yesterday with the following information:
Bugcheck code: 0x3B (0xC0000005, 0xFFFFF80560CCB52A, 0xFFFFCA052CA4B860, 0x0)
Error: SYSTEM_SERVICE_EXCEPTION
file path: C:\Windows\system32\ntoskrnl.exe

I've tested the RAM, and I can't seem to find issues with it. I would appreciate any hint as to where to go next.
 
Please list system specs including make and model of PSU.

CPU: AMD Ryzen 7 3800X 3.9GHz 8-Core
GPU: GIGABYTE GeForce RTX 3070 GAMING OC - 8GB GDDR6 RA
RAM: Kingston HyperX Predator 32GB DDR4 3200MHz C (2 for 64GB of RAM)
Motherboard: GIGABYTE X570 UD - AMD X570 - AMD AM4 soc
Storage:
  • Samsung SSD 970 EVO 1TB M.2 PCIe
  • Samsung 870 QVO 2.5" SSD - 2TB
PSU: AORUS P850W 80+ GOLD Modular (Model: GP-AP850GM)
 
Being you say "but it doesn't seem to happen under low temps. "

Have you checked the temps?
No. The reason I say that is because the computer runs completely fine under low stress for days on end. I can also use it for some time under high stress. The problem mostly occurs after the computer has been doing something intensive for a while. Rebooting doesn't seem to help, as the bluescreens usually happen in rows. I haven't had any today for example, but I also haven't gotten my CPU over 10% at any point. Yesterday the problem seemed to go away with time, which is why I suspect temps, but I can run some tests if you can give me some directions on what to test and how.
 
Usually when a PC bluescreens under stress it is either too hot or it is a power supply problem.

Download and run HWInfo.

You can check the temps.....as well as a lot of other things.....with it.

In general the CPU should remain under about 75C...GPU under 85C,

You can also check the voltages......which should be in the motherboard section (if the MB supports it).

The main voltages are 12,5 and 3.3 and they should remain within 5% of these values always.
 
Usually when a PC bluescreens under stress it is either too hot or it is a power supply problem.

Download and run HWInfo.

You can check the temps.....as well as a lot of other things.....with it.

In general the CPU should remain under about 75C...GPU under 85C,

You can also check the voltages......which should be in the motherboard section (if the MB supports it).

The main voltages are 12,5 and 3.3 and they should remain within 5% of these values always.
Should I just open the sensor window, start logging and run a stress test? If so, how long?
 
Should I just open the sensor window, start logging and run a stress test? If so, how long?
Usually I choose "sensors only" but it doesn't much matter.
When the sensor window is open you can visually monitor the data.....and it records the high and low values.
You may see the problem right here.
You could also let it run.....as long as the PC doesn't crash....and let it record and see what happens.
If you anticipate a crash.....you can set it up to log and it will log right up to the crash.
Then you can open the file and look at it after.
 
Usually I choose "sensors only" but it doesn't much matter.
When the sensor window is open you can visually monitor the data.....and it records the high and low values.
You may see the problem right here.
You could also let it run.....as long as the PC doesn't crash....and let it record and see what happens.
If you anticipate a crash.....you can set it up to log and it will log right up to the crash.
Then you can open the file and look at it after.
Here's the log file:
Log File

So. The computer crashed when running prime95. Looks like the CPU maxed at about 90 and the voltages maxed at around 3.423 and 12.312.
 
Well <Mod Edit>. Got another BSOD at 60 degrees on the CPU:

Bugcheck code: 0x18 (0xFFFFE68B4ABFE380, 0xFFFFE68B5E237A40, 0x3, 0x7000000000000000)
Error: REFERENCE_BY_POINTER
file path: C:\Windows\system32\ntoskrnl.exe
 
Last edited by a moderator: