What's Wrong? Games Cause System to Flicker then Crash.

Associate
Joined
28 Feb 2012
Posts
1,832
Location
London, UK
How can I go about finding what's wrong?

Played F1 2018 last night and switched PC off. Turned it back on this morning and this afternoon fired up F1 again and as soon as it loads in, the entire system flickers and after a few seconds will crash. Picture of what happens.

k1340OIl.jpg.png

To make sure no windows updates broke my system, I restored to the day before. Issue still happens. Have tried loading up F1 2017 and same issue.

I would guess my GPU is done for, 780, run for years but is there anything I can do to check it isn't the GPU as I would hate to get a new one to find out something's wrong somewhere else.
 
Last edited:
Associate
OP
Joined
28 Feb 2012
Posts
1,832
Location
London, UK
Does it load into windows ok? Could be the PSU is on its last legs if not the GPU.

Sometimes it does, sometimes it doesn't. At first it was artifacting in games and now sometimes before logging in the screen will start flickering.

I guess the only way to know if it's the PSU is to swap it out? Either way I've got to buy one component and try?

Would hate if it was the PSU as it's a AX860 which is perfect for me but sinking money into a new PSU would be terrible as I need a new GPU anyway.
 
Associate
OP
Joined
28 Feb 2012
Posts
1,832
Location
London, UK
How old is the PSU? If you could get hold of another to test at least it will rule it out. Although id put money on it being the GFX card, ive seen these symptoms before and a new PSU fixed it.

Corsair AX860, just past 5 years old and probably used for about 3 years (spent a lot of time away from home).

Would a CPU stress test tax the PSU enough to cause it to show if it is the issue?
 
Joined
10 Jan 2004
Posts
9,839
Location
Poland
It could do but the pci-e is probably on a separate rail so might not show up. Give it a shot, and also check your gfx temps too if you haven't already, try running some gfx benchmarks if you can to see where it falls over then check Event Viewer, see if it throws up some useful info like power problems, hardware etc.
 
Associate
OP
Joined
28 Feb 2012
Posts
1,832
Location
London, UK
It could do but the pci-e is probably on a separate rail so might not show up. Give it a shot, and also check your gfx temps too if you haven't already, try running some gfx benchmarks if you can to see where it falls over then check Event Viewer, see if it throws up some useful info like power problems, hardware etc.

Appreciate the help btw.

Ran F1 2018. What normally happens is it will start flickering and then proceed to crash. So I ran it long enough to start flickering and alt+f4'ed out. Checked event viewer and have been flooded with the same error.

Code:
The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

The source is from the NVIDIA driver which I have reinstalled twice after wiping the drivers with DDU.
After the last line of the above, each error has something to do with a "Header".

Code:
Graphics Exception: Shader Program Header 18 Error

Code:
Graphics Exception: Shader Program Header 3 Error

etc.

Ran Prime95 for about an hour, stable temps, no problems.

My GPU is idling at 30C. One thing to note is that this card was factory OCed but was never stable on factory OC. I have run this card at a higher power limit and -100MHz on core clock. That has been stable for years.

Will run some GPU benchmarks and will report back.
 
Associate
OP
Joined
28 Feb 2012
Posts
1,832
Location
London, UK
No worries mate, i know how frustrating it can be pinning something down!

Interestingly i looked up that shader program error and apparently ram stability can cause it too! Maybe running windows memory diagnostic or memtest wont hurt. I feel this problem could be a whole can of worms! :/

So 99% sure it's the GPU now. Windows Memory Diagnostic came back as 0 issues. I would have run Memtest but can't seem to find a disk or memory stick to put it on.

I ran Heaven and artifacted instantly and system crashed.

Luckily managed to find an old 680 so after a driver cleanup, ran Heaven with that and no issues at all.

This would be the GPU right? Or is there a chance the 680 isn't causing a potential PSU issue to come up? If the PSU could be an issue then I might just take my GPU into work and test on my workstation although that would be a bit of a pain.
 
Joined
10 Jan 2004
Posts
9,839
Location
Poland
Yeah its looking like its the GPU, the 780 draws about 250w compared to 190? for the 680 but what you describe makes me think its dying. I guess you could underclock the hell out of it to see if it stays stable but still, it wont be fit for purpose either way. As much as a pain as it is, if you can test it elsewhere it will save you a lot of hassle in the long run
 
Associate
OP
Joined
28 Feb 2012
Posts
1,832
Location
London, UK
Yeah its looking like its the GPU, the 780 draws about 250w compared to 190? for the 680 but what you describe makes me think its dying. I guess you could underclock the hell out of it to see if it stays stable but still, it wont be fit for purpose either way. As much as a pain as it is, if you can test it elsewhere it will save you a lot of hassle in the long run

Well took my GPU into work and got to test it in my lunch break. Thankfully it is the GPU that is gone, same issue at work. Now I have an excuse to upgrade to 1440p and will probably grab a 2070 or something.
 
Back
Top Bottom