[SOLVED] [GPU Crash] Monitor > No Signal - Can still hear sound (discord etc)

[Bartt]

Hello!

Lately I've been having a rather strange problem with my pc.

While under load (playing games like ark, rainbow six siege - but also happened once with Bloons TD6) my monitors (I have 2 monitors) give a 'No Signal' message and sometimes my fans also start spinning really loud. During that I can still hear sounds (discord, in-game sound etc) but my monitors have become unresponsive.
- This happens only under load as far as I know, but also then it seems to have periods of not crashing (yesterday I played a few hours Rainbow, and last night a few hours Ark without crashing)

Does anyone know what this might be or how I could figure out exactly what component is causing this? I've tried:

• Safe mode, doesn't work
• CMOS / BIOS reset, doesn't work
• Checked dust in PC, tried putting only 1 ram stick, tried putting the GPU into a different PCIE slot, doesn't work
• New GPU drivers (with DDU), doesn't work
  Edit: since I have 2 monitors and they both crash at the same time it's also not my HDMI/DP cable and/or monitor itself

I've checked the 12v rail voltages, and those seem to be on the low side (sometimes as low as 11.3, but usually 11.6/11.7)

With those lower PSU voltages I've started leaning towards thinking of a PSU problem but I'm just not quite sure because 11.6 is in the '0.6v range of 12v, and to add to that: I feel like when I'm not in call with my friends it happens less/not: when playing Rainbow on my own I don't think I have experienced much crashes but as soon as I join with my friends it crashes much more often.
- It just seems really crazy to me that audio could cause an GPU crash but maybe you guys know more about this (+ I was playing Ark with friends for hours, so there the audio definitely was not a problem)

Since I've basically kept crashing this evening (after all the things I've tried) I've come here to ask for some support..

Does anyone know what I can still try to figure out exactly what component is the broken one?

Thanks in advance!
Bart


Edit: Specs:
CPU - Intel i5-8600
GPU - Gigabyte GeForce GTX 1080 WINDFORCE
PSU - Corsair RM550x (2018) (EU-model)
MOBO - Asus TUF B360-Pro Gaming (Wifi)
RAM - Crucial Ballistix Sport LT 2x8GB Sticks @ 2666 Mhz

 

[Giant Hunger]

I'm not an expert in computers. But I think you could have faulty vram. Try install msi afterburner and lower the core clock and memery clock. After that try to play your usual games to see if it crash or not.

 

[nitrocide]

Had the same issue for months, tried everything you mentioned and more and searched around forever and saw so many different solutions, ended up getting a nice new PSU and it didn't fix it. The way I ended up "fixing" it for now to hold me over until a new GPU is by lowering the core and memory clocks in MSI Afterburner as much as the program allows, as well as setting the power limit to 75%. I might be able to set these numbers higher but I haven't had problems and don't notice a performance decrease although there probably is.

 

[Bartt]

Had the same issue for months, tried everything you mentioned and more and searched around forever and saw so many different solutions, ended up getting a nice new PSU and it didn't fix it. The way I ended up "fixing" it for now to hold me over until a new GPU is by lowering the core and memory clocks in MSI Afterburner as much as the program allows, as well as setting the power limit to 75%. I might be able to set these numbers higher but I haven't had problems and don't notice a performance decrease although there probably is.

I actually already underclocked the GPU a tad bit a few hours before you sent this message but right after I've underclocked it as much as possible without succes - I've just crashed again.

I now however can't even start up my PC, the white VGA led is on indicating that something is wrong with the video output. I think I can now rightfully say that my GPU has died and that my PSU is working just fine. I'll be looking at the gpu/cables/monitors for a bit but I think that was the last blow.

Kinda disappointing knowing how expensive GPUs currently are with all the mining/corona stuff, but atleast I finally figured out the problem (and well it's really no surprise that it is the GPU)

 

[Giant Hunger]

I actually already underclocked the GPU a tad bit a few hours before you sent this message but right after I've underclocked it as much as possible without succes - I've just crashed again.

I now however can't even start up my PC, the white VGA led is on indicating that something is wrong with the video output. I think I can now rightfully say that my GPU has died and that my PSU is working just fine. I'll be looking at the gpu/cables/monitors for a bit but I think that was the last blow.

Kinda disappointing knowing how expensive GPUs currently are with all the mining/corona stuff, but atleast I finally figured out the problem (and well it's really no surprise that it is the GPU)

Well if your gpu died, you could try and bake your gpu. This actually works for me, I have a gtx 780 that has artifacts on it, so I used my heat gun to heat up the gpu, after that it stop artifacting. Make sure this is your last resort if your going for it

 

[AB2647]

Same thing happened to my Sapphire HD 6950 1GB. Under-clocked it, worked for about three months, and then same thing happened all over again. Flickering images at first, then lead to black screens wit sounds and everything on, just no video output. Game still playable. Now, I can't even watch a 1080p video. PSU was a bad one, 600W, only 348W on +12V rail. The 6950 already has 2 6 pins which the PSU didn't offer, and had to use the cables that came with the PSU.

 

[Bartt]

Same thing happened to my Sapphire HD 6950 1GB. Under-clocked it, worked for about three months, and then same thing happened all over again. Flickering images at first, then lead to black screens wit sounds and everything on, just no video output. Game still playable. Now, I can't even watch a 1080p video. PSU was a bad one, 600W, only 348W on +12V rail. The 6950 already has 2 6 pins which the PSU didn't offer, and had to use the cables that came with the PSU.

So you had a faulty PSU, I'll probably buy a new PSU aswell and see if that fixes it.

I've now fixed the white VGA led, so my pc boots up again but I still had a crash. Weirdly enough tho I crashed the first second running furmark but after I rebooted I ran it like 5 times without it crashing which makes it even more weird. Now I've played some Rainbow for a bit without crashing aswell.

It really seems to crash more often when I play games with friends for some odd reason, which makes this all more complex to understand. Like: why does my GPU crash more often when I'm playing games with my friends? Why did furmark crash the first time within a second but after that could run 5+ times GPU tests fine?

Edit: so that's kinda my main point where I'm bothered about: it's not consistent. When I don't play with friends (seemingly) on discord it does not crash, sometimes it crashes within a second on Furmark but sometimes I can run the GPU test 5+ times... This makes it so much harder to figure out what's going on.

I don't believe it has to do with audio, that just seems odd (I did update my audio drivers just in case, but obv. that didn't fix it) but it's hard to figure out since I can't recreate it on purpose..

 

[Bartt]

Same thing happened to my Sapphire HD 6950 1GB. Under-clocked it, worked for about three months, and then same thing happened all over again. Flickering images at first, then lead to black screens wit sounds and everything on, just no video output. Game still playable. Now, I can't even watch a 1080p video. PSU was a bad one, 600W, only 348W on +12V rail. The 6950 already has 2 6 pins which the PSU didn't offer, and had to use the cables that came with the PSU.

Yes!! New PSU seems to have fixed it🤞

Yesterday my PC didn't want to boot at all anymore with the GPU inserted (VGA led), so I tried it without the external GPU. This got me in the bios but with a 'CPU Fan error'. The most basic problem that could cause that is a broken CPU fan (/cooler) or not properly installed pins but that would very coincidental so I figured it had to be the PSU.

In the bios the 12v rail also had very low voltages which more or less confirmed my assumptions of a broken PSU. That considered + the message that you also had a faulty PSU made me buy a new one and so far so good.

 

[AB2647]

Yes!! New PSU seems to have fixed it🤞

Yesterday my PC didn't want to boot at all anymore with the GPU inserted (VGA led), so I tried it without the external GPU. This got me in the bios but with a 'CPU Fan error'. The most basic problem that could cause that is a broken CPU fan (/cooler) or not properly installed pins but that would very coincidental so I figured it had to be the PSU.

In the bios the 12v rail also had very low voltages which more or less confirmed my assumptions of a broken PSU. That considered + the message that you also had a faulty PSU made me buy a new one and so far so good.

Good news, hope it stays that way. Take care of the gpu. Finest pieces of art from our engineers.