Computer crashes while/after using UserBenchmark

[MidiaMay]

Hello! Today I ran UserBenchMark to benchmark CPU, ram, GPU, etc. When it began testing the GPU, a particular video appeared and my PC went black screen. I was forced to restart using the power button.

The boot screen had wide, red, vertical lines through it. When it started to load Windows, the PC crashed and the screen was thin white/grey vertical lines.
Restarting fixed the issue. The GPU and CPU temps were normal. But after playing a game for ~15 minutes, the computer crashed with vertical lines again.

Restarted again, was just on Reddit and the same thing happened for a third time after ~10 minutes.

I tried to install my old GPU but it's not being detected (it's been sitting in a box for 5 years...) so now I'm using motherboard graphics and all has been well for ~2 hours now.

I have had no hardware issues with my PC before today. It never crashes, despite a lot of the parts being 4-5 years old.

How could running a benchmark mess something up? Also, how do I go about finding the real issue (I think it's the GPU but I wanna be 100% sure... could there be a different issue?).

TL;DR Ran UserBenchMark, on GPU testing computer crashed with white/grey vertical lines. It kept crashing afterwards... On motherboard graphics now for ~40 min and no crashes. What could be wrong?

--------------------------------------------------

Computer Specs:
AMD Phenom II X4 925
8 GB Ram
1 TB HDD
600 W Power supply
Radeon HD 6870
Radeon HD 5570 (GPU I tried to install but not detected)
GA-M68MT-S2P rev 3.0

Any help would be much appreciated!


EDIT: Replaced GPU and the issue is solved

 

[Ionizor]

How old is your computer?

Check if the fans are spinning on cpu and gpu. (had the problems of fan not spinning on gpu, if not spinning replace the fan, or buy a new gpu)

I think that benchmark has a discalmer that says its not responsilbe for any damage on your computer, so maybe you overloaded the already strained(old) system.

If onboard graphics work, then its probably gpu. Try with a nother gpu.

 

[Faux_Grey]

Also could be a power supply.

If the power supply is a 600w "cheap brand" and it's 5-6 years old. Could easily not provide enough current.


I recall my Sapphire "toxic" 6870 pulling around 130-140w of current by itself.

 

[MidiaMay]

How old is your computer?

Check if the fans are spinning on cpu and gpu. (had the problems of fan not spinning on gpu, if not spinning replace the fan, or buy a new gpu)

I think that benchmark has a discalmer that says its not responsilbe for any damage on your computer, so maybe you overloaded the already strained(old) system.

If onboard graphics work, then its probably gpu. Try with a nother gpu.

Computer (case, CPU, motherboard, HDD) is 5.5 years old. Ram & GPU are almost 5 years old. Replaced the CPU heatsink/fan a few months ago. And I just replaced the PSU today with a new one.

I tried with different GPU yesterday, seemed fine... but the fan is broken on it so I didn't wanna use it for too long.

All fans have been spinning. I have PC running now, been up for an hour no crash (yet). Not doing anything graphic intensive.

But red and green pixels will randomly appear all over my screen for a few seconds then disappear... Sometimes this causes the screen to freeze but it refreshes and is fine. The audio is fine while this happens (in a skype call..) This is happening every ~10 minutes?

I did uninstall AMD Crimson drivers and am now on CC drivers now.

EDIT: It finally crashed after the screen was full of pixels, but instead of gray line of death screen, Windows blue screened.

Every time the pixel thing happened, this error occurred:
Display driver amdkmdap stopped responding and has successfully recovered.

 

[Ionizor]

How old is your computer?

Check if the fans are spinning on cpu and gpu. (had the problems of fan not spinning on gpu, if not spinning replace the fan, or buy a new gpu)

I think that benchmark has a discalmer that says its not responsilbe for any damage on your computer, so maybe you overloaded the already strained(old) system.

If onboard graphics work, then its probably gpu. Try with a nother gpu.

Computer (case, CPU, motherboard, HDD) is 5.5 years old. Ram & GPU are almost 5 years old. Replaced the CPU heatsink/fan a few months ago. And I just replaced the PSU today with a new one.

I tried with different GPU yesterday, seemed fine... but the fan is broken on it so I didn't wanna use it for too long.

All fans have been spinning. I have PC running now, been up for an hour no crash (yet). Not doing anything graphic intensive.

But red and green pixels will randomly appear all over my screen for a few seconds then disappear... Sometimes this causes the screen to freeze but it refreshes and is fine. The audio is fine while this happens (in a skype call..) This is happening every ~10 minutes?

I did uninstall AMD Crimson drivers and am now on CC drivers now.

EDIT: It finally crashed after the screen was full of pixels, but instead of gray line of death screen, Windows blue screened.

Every time the pixel thing happened, this error occurred:
Display driver amdkmdap stopped responding and has successfully recovered.

Faulty mobo, had similar problems. Artifacts and later BSOD, shorter times for it to crash, and problems with reaching gpu. The best thing you can do now is check it out with a proffesional, but i think for an computer that old is not worth it. I think you can save the files on the old HDD if you install OS on a new HDD or SSD.
If going for a custom build care for: DDR-s on RAMs(mobos have specific ddrs), cpu for mobo(socket), and the size of the casing(so you can fit a larger gpu).

 

[MidiaMay]

Faulty mobo, had similar problems. Artifacts and later BSOD, shorter times for it to crash, and problems with reaching gpu. The best thing you can do now is check it out with a proffesional, but i think for an computer that old is not worth it. I think you can save the files on the old HDD if you install OS on a new HDD or SSD.
If going for a custom build care for: DDR-s on RAMs(mobos have specific ddrs), cpu for mobo(socket), and the size of the casing(so you can fit a larger gpu).

I've had another GPU running for an hour now with no issues. It's old though and the fan does not work so I don't wanna run it for too long or put it under too much stress. (HD 5570)

I couldn't even have my other GPU running for 15 minutes without pixels or some other error under 0% load... They use the exact same drivers.

Makes me think it''s the GPU, not the mobo... but I really wanna be sure. I can't afford to replace more than one part at this time. Also the whole issue started when my GPU in particular was stress tested.

I'm probably just gonna replace the GPU unless there's some way to figure out if the mobo is really the issue.