Question Video Driver Crash: Windows 10, GTX 970. Looking for trouble shooting advice.

[ke5tgd]

Hello all,
I've have a problem that is resulting in what I believe to be a video driver crash. This is a recurring issue that has persisted over many months and through many common fixes.
The Symptoms:
Audio glitch, Audio output holds a note or tone for maybe a tenth of a second when the problem starts. I'm not sure if it happens every time.
Frozen screen, This is temporary and happens every time.
Solid color/Artifacts, This is an odd one, usually the screen will unfreeze momentarily and then freeze again with either a solid color(Black, Red, shades of Blue or Green). Sometime one monitor will be black and the other will be a different color (Never both different colors). Sometimes instead of freezing with a solid color there is an artifact of what was on screen with the pixels shifted in their rows or rearranged in repetitive blocks.

Above is one of the more egregious examples from October. This condition will sometimes resolve its self but other times I have to restart the video driver with Windows+Ctrl+Shift+B.
Software and Application crashes: After the video driver crashes or recovers one or more of my applications running will crash. The software that crashes will normally be what ever was open on either monitor (game + Discord/Web browser). The software that crashes is not always the ones open on either monitor at the time of the crash but I cant point to any trends there.
The Circumstances:
Multiple apps open
One DirectX game open (I've never had this problem with OpenGL or any other graphics API)
Video streaming of some kind (Discord share screens, YouTube, Netflix, Amazon Prime Video, etc.)
Info from crash logs:
Crash logs from games involved are not very helpful either ending abruptly with no sign of trouble or pointing loosely at the graphics driver.
Fixes I've tried:
Fresh install of newest graphics drivers (I wiped the drivers with some software when I did this)
Backdating drives to known good versions
Backdating drives to original 2015 drivers
Clean and reseat graphics card
Lowest graphics settings in games
Different wall outlets, houses, energy companies... (I have moved since the problem started)
Different operating orientations (Trying to eliminate GPU sag as a possible cause)
System specs:
i7-4770K (4.2GHz OC)
ASUS TUF Gryphon Z87
2x8GB Critical Patriot DDR3-1600MHz
EVGA GTX 970 ACX 2.0 SSC
Corsair CS550M 550 watts
MOBO, CPU and RAM are like 9 years old. PSU and GPU are like 7

I appreciate any insights or pointers. This computer has seen its fair share of janky use over the years so no theories too wild or checks too basic.

 

[Colif]

Are you getting any BSOD? Or is it just this weirdness?

Have you got chance to test GPU in another PC and see if it has same problems?

Thats not the best PSU you have there but I doubt it causes this. I would still think about replacing it considering its age.

Could just be age, My GTX 980 stop running cool but I was lucky, had a new GPU arriving that week. And it was only 5 years old. They don't last forever.

 

[Bob.B]

Hello all,
I've have a problem that is resulting in what I believe to be a video driver crash. This is a recurring issue that has persisted over many months and through many common fixes.
The Symptoms:
Audio glitch, Audio output holds a note or tone for maybe a tenth of a second when the problem starts. I'm not sure if it happens every time.
Frozen screen, This is temporary and happens every time.
Solid color/Artifacts, This is an odd one, usually the screen will unfreeze momentarily and then freeze again with either a solid color(Black, Red, shades of Blue or Green). Sometime one monitor will be black and the other will be a different color (Never both different colors). Sometimes instead of freezing with a solid color there is an artifact of what was on screen with the pixels shifted in their rows or rearranged in repetitive blocks.

Above is one of the more egregious examples from October. This condition will sometimes resolve its self but other times I have to restart the video driver with Windows+Ctrl+Shift+B.
Software and Application crashes: After the video driver crashes or recovers one or more of my applications running will crash. The software that crashes will normally be what ever was open on either monitor (game + Discord/Web browser). The software that crashes is not always the ones open on either monitor at the time of the crash but I cant point to any trends there.
The Circumstances:
Multiple apps open
One DirectX game open (I've never had this problem with OpenGL or any other graphics API)
Video streaming of some kind (Discord share screens, YouTube, Netflix, Amazon Prime Video, etc.)
Info from crash logs:
Crash logs from games involved are not very helpful either ending abruptly with no sign of trouble or pointing loosely at the graphics driver.
Fixes I've tried:
Fresh install of newest graphics drivers (I wiped the drivers with some software when I did this)
Backdating drives to known good versions
Backdating drives to original 2015 drivers
Clean and reseat graphics card
Lowest graphics settings in games
Different wall outlets, houses, energy companies... (I have moved since the problem started)
Different operating orientations (Trying to eliminate GPU sag as a possible cause)
System specs:
i7-4770K (4.2GHz OC)
ASUS TUF Gryphon Z87
2x8GB Critical Patriot DDR3-1600MHz
EVGA GTX 970 ACX 2.0 SSC
Corsair CS550M 550 watts
MOBO, CPU and RAM are like 9 years old. PSU and GPU are like 7

I appreciate any insights or pointers. This computer has seen its fair share of janky use over the years so no theories too wild or checks too basic.

Just a test.
Put a copy of memtest86 on a flash stick.
Boot the stick and let it run.....no errors allowed.

 

[ke5tgd]

Are you getting any BSOD? Or is it just this weirdness?

Have you got chance to test GPU in another PC and see if it has same problems?

Thats not the best PSU you have there but I doubt it causes this. I would still think about replacing it considering its age.

Could just be age, My GTX 980 stop running cool but I was lucky, had a new GPU arriving that week. And it was only 5 years old. They don't last forever.

No BSOD. I've had BSODs similar in the past but that doesn't happen anymore and I believe it was caused by a bad USB device.

I dont have any other systems right now that I can test in unfortunately.

I have had the same thought, I turned my CPU overclock off to see if a little less power draw would change anything but nope. Problem still happened with no overclock.

My GPU is very old, had it since 2016. It has always run hotter than I would like (70-80C) but its not gotten hotter. I could crank up the fan curve to get it cooler, it just makes the PC kinda loud.

 

[ke5tgd]

Just a test.
Put a copy of memtest86 on a flash stick.
Boot the stick and let it run.....no errors allowed.

Ok give that a try after work today. Does it need to go on a flash drive or will any bootable partition do? I have an HDD partitioned with a small boot sector that I could use.

 

[Bob.B]

Ok give that a try after work today. Does it need to go on a flash drive or will any bootable partition do? I have an HDD partitioned with a small boot sector that I could use.

Don't know pick up a 16GB flash drive their cheap.

 

[Colif]

memtest has to run from usb as it doesn't boot into windows to run

Try running memtest86 on each of your ram sticks, one stick at a time, up to 4 passes. Only error count you want is 0, any higher could be cause of the BSOD. Remove/replace ram sticks with errors.

Memtest is created as a bootable USB so that you don’t need windows to run it

 

[ke5tgd]

memtest has to run from usb as it doesn't boot into windows to run

Try running memtest86 on each of your ram sticks, one stick at a time, up to 4 passes. Only error count you want is 0, any higher could be cause of the BSOD. Remove/replace ram sticks with errors.

Memtest is created as a bootable USB so that you don’t need windows to run it

Ok I let memtest86 run while I was at work today. Took about three hours but there were no errors.

 

[Colif]

My GPU is very old, had it since 2016. It has always run hotter than I would like (70-80C) but its not gotten hotter. I could crank up the fan curve to get it cooler, it just makes the PC kinda loud.

My heat thing didn't cause any errors, I just noticed it a few days before I replaced card. It was just causing PC fans to run more.

Might need to get repair shop to look at it, they might have a pc they can put GPU in and see if it does same things. It could be its just showing its age.

Have you tried reinstalling windows? it might help.

Does anything show in reliability history that could relate to GPU drivers?

 

[ke5tgd]

My heat thing didn't cause any errors, I just noticed it a few days before I replaced card. It was just causing PC fans to run more.

Might need to get repair shop to look at it, they might have a pc they can put GPU in and see if it does same things. It could be its just showing its age.

Have you tried reinstalling windows? it might help.

Does anything show in reliability history that could relate to GPU drivers?

I really don't want to reinstall windows but I know its one of the few legit fixes I haven't tried. Also I'm not sure what you mean by 'reliability history'.

In regards to taking the card to a shop, I doubt they would be able to replicate the problem in a reasonable amount of time. Sometimes it runs fine for weeks. To know with certainty its not the card would take quite a lot of testing time.

I have this sinking feeling its the RAM on the video card causing the problem although I have no idea how to test that.

 

[Colif]

Also I'm not sure what you mean by 'reliability history'.

How to Check Your Computer Reliability with Reliability Monitor

We show you how to use Reliability Monitor, including how you can view Reliability History in Windows 10 and create a shortcut for easier access.

winbuzzer.com

that link is dumb, all you need to do to access it is search for "reliability history" in search bar. It should show there. No need to make a shortcut.

long shot is the error is caused by a USB device conflict. But that sort of depends on results of the above link.

 

[ke5tgd]

How to Check Your Computer Reliability with Reliability Monitor

We show you how to use Reliability Monitor, including how you can view Reliability History in Windows 10 and create a shortcut for easier access.

winbuzzer.com

that link is dumb, all you need to do to access it is search for "reliability history" in search bar. It should show there. No need to make a shortcut.

long shot is the error is caused by a USB device conflict. But that sort of depends on results of the above link.

Ok. Ill get back to you with the data from this next time the event occurs. I see a few events in here already. A very suspicious 'Hardware error' but I don't remember the time of day the event occurred so it could be unrelated. Might be a few days before it happens again. I would prefer to get it to happen under a repeatable benchmark load but we shall see.

I also found this question posted and I have turned off hardware acceleration in discord. To soon to say if its worked but if it goes a few weeks with no problems I'll mark this as solved

 

[ke5tgd]

Woah. Super lucky, crashed just now in KSP with YouTube on the side monitor. Now my KSP install is modded so take this with a grain of salt but It had all the usual symptoms.

Here is exactly what just happened:
KSP on monitor 1 froze and at the same time audio glitched like I described before.
Monitor 1 went to a navy blue color and monitor 2 went to about 1 FPS.
I hit Alt+Tab to see if monitor 1 was frozen, It was but I could move my mouse on monitor 2
Shortly after that monitor 2 turned black with some artifacts in the bottom left.
Before I could force restart the graphics driver it restarted itself with KSP crashed and chrome in a graphically frozen sate after recovery. Chrome recovered after being minimized and maximized.
Note that aside from the very beginning the audio worked fine the whole time.

Problem details from the two critical events recorded in Reliability monitor
--------------------------------------------------------------------------------------------------------------------------------------
Source
KSP_x64.exe

Summary
Stopped working

Date
‎6/‎1/‎2023 8:59 PM

Status
Report sent

Description
Faulting Application Path: C:\Program Files (x86)\Steam\steamapps\common\Kerbal Space Program\KSP_x64.exe

Problem signature
Problem Event Name: APPCRASH
Application Name: KSP_x64.exe
Application Version: 2019.4.18.4260
Application Timestamp: 5ffdc692
Fault Module Name: UnityPlayer.dll
Fault Module Version: 2019.4.18.4260
Fault Module Timestamp: 5ffdc82f
Exception Code: c0000005
Exception Offset: 00000000004516d3
OS Version: 10.0.19045.2.0.0.256.4
Locale ID: 1033
Additional Information 1: 372d
Additional Information 2: 372da45f6e4c97998c8ef25e0fefaf58
Additional Information 3: aa10
Additional Information 4: aa1005303018e8ce479932ed24774c07

Extra information about the problem
Bucket ID: 7298f35aad0597de2d578318eb6a3409 (2114302693125796873)
--------------------------------------------------------------------------------------------------------------------------------------
Source
Windows

Summary
Hardware error

Date
‎6/‎1/‎2023 8:59 PM

Status
Not reported

Description
A problem with your hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
Code: 117
Parameter 1: ffffd58180a70010
Parameter 2: fffff80052061690
Parameter 3: 0
Parameter 4: 480
OS version: 10_0_19045
Service Pack: 0_0
Product: 256_1
OS Version: 10.0.19045.2.0.0.256.4
Locale ID: 1033