1080TI Causing Crashing & BSOD

[johnace01]

Hi all, long time no post!

I currently have a 1080TI (EVGA Founders Edition) which has been getting worse for a few months now, it will crash, usually, everything freezes and becomes unresponsive (audio will continue) then I get a black screen, then BSOD usually critical_process_died unhandled_store_exception once a ram management one. Like I say this was occasionally, but now it is every 30 seconds.

I have done all the usual testing, updated drivers, reinstalled Windows, tried each stick of ram in different slots, tried the GPU in different PCIE slots, changed BIOS to PCIE Gen3 instead of Auto and now tried a 980TI which seems to work fine. When I reseat the card it generally tends to be OK for a couple of days.

Looked online and info is pretty scarce however people are having this issue! Has anyone had this and been able to solve it? Happy to open it up etc. not so happy to buy a new card - now is not the time for gamers! 

Any help or further info is much appreciated!

Full specs:
i7 6700K
EVGA Z270 FTWK Mobo
16GB HyperX Fury 
EVGA 850w PSU
Also have SoundBlaster AE-9 however this has been an issue pre-that-card

[ir_cow]

Sorry to hear about your troubles. Good news is i'm sure we can get to the bottom of the actual problem if its the 1080 TI or not and go from there.

To rule out the power supply, just in case. Download MSI Kombustor https://geeks3d.com/furmark/kombustor/. Run it with the GTX 980 Ti. Since both cards have the same power draw at stock it will be easy to eliminate the PSU being the problem. If the GTX  980 Ti goes for 10 minutes, give the GTX 1080 TI a try. Seems like from what your describe its pretty constant on the crashing (30 seconds you say). It probably won't last that long. But keep on eye on the temps just in case. 

If it ends up being the GTX 1080 Ti and seeing you done all the system trouble shooting. Evening going as far to reinstall windows. The only thing left to do is run GPU-Z while stress testing and seeing if its overheating. The fan could have died, or something like that.

In any case I'm afraid if its not a PSU issue, you could narrow down what is the problem, but no way to really fix it in the end.... Bad VRAM or bad Silicon can't be replaced without proper lab tools. Which leaves just some capacitors and MOSFETS that can be replaced with a soldering iron.

In the event that you cannot get the GTX 1080 Ti boot long enough to trouble shoot and the GTX 980 Ti works fine, Other than new thermal paste, not much more you can do without running around with a multimeter and testing all the capacitors.

 

[ir_cow]

Also I see you said re-seating the card is a short term solution. I'm not sure how that factors into this yet. If I was wildly guessing why this works and the card indeed is fine for X amount of time once your re-seat it in ANY PCIE slot, it could be a broken tracer. As the PCB sages and settles in over hours / days, the tracer pulls apart and that is your issue.

However I wouldn't go that far just yet.... focus on one thing at a time.

[johnace01]

Thanks for the help!

I have already done a Furmark test on both cards for roughly 10 mins as you suggested with GPU-Z on the go, nothing foul to report from it really unfortunately. The 1080TI has been in now since yesterday had no problems with it after reseating so far (though I know this WILL change.)

This seems like its going to be quite an issue, usually I'd use it as a cheap excuse to upgrade however given the extortionate prices of GPUs at the moment I will use my laptop! Unfortunately my Dad uses the 980TI in his PC and I don't have any more graphics cards hanging around - quite how that happened I don't know I'm sure I had thousands of the buggers at one point!

[ir_cow]

Hmm. I didn't think you had another computer handy. If you haven't already, pop that GTX 1080 Ti into your dads computer if you can. When it BSOD than you know it cannot be anything else related to your system.

I'm hung up on why it works after being re-seated for a few days. It is such a odd thing.

I don't like telling people they have bad hardware because the amount of times I thought I had something bad over the years is a lot. Only to find out it was something really odd and unrelated. But it is starting to look that way for you. What is causing your card to fail is beyond my knowledge. I'm good at troubleshooting down to the hardware in question.

[sticknstone]

Been reading about this after a web search.  The windows minidump file will tell you what the problem is.  I searched the error code the BSOD left.
Whatever it tells you can be researched further.

[johnace01]

OK so it happened again today as expected, it will just get more frequent as of now.

I definitely already did grab the dump and search it up before (maybe 6 months+) this has been going on for some time, however I got absolutely nothing that helped me at all - hence why I thought "hang on a second what about OCC."  

I will let it happen again then get the dump and find some more info, at my wits end to be honest - fortunately my laptop is sufficient to do what I want to do.

[sticknstone]

Post back here with the info on the dump if I can help I will.
Nothing you mentioned makes me think the graphics card is faulty because when I searched the BSOD message others had a driver issue or hard drive issue.  The message is more general until the dump file is analyzed.
The CPU does have integrated graphics.

[ir_cow]

@sticknstone Wouldn't you think a clean windows install would solve all the driver issues? I guess it could be that Soundblaster is conflicting with NVIDIA Audio. I read so many problems with sound cards these days. I would also suggest removing the card if the dump file doesn't give any better clues.

Heck I have a Soundblaster Audigy 2. I still remember to this day fighting the drivers weekly.

[sticknstone]

I thought the same thing, but the BSOD came back.  Until the dump file gets analyzed no one will know the source of the problem.

[johnace01]

Right here we are more than a month later,

I changed the thermal paste despite the temps being fine - this did result in slightly lower temps which is a winner I suppose however after a few days low and behold, BSOD.

So, I did a little digging and something I hadn't noticed before was that the lane it was in is a 16x - 8x when using certain slots - on my mobo that is slot 4, which, my soundcard was in slot 4 despite only needing the smaller 1x slot this was to make it as far away from the GPU as possible to reduce any noise heard through it. I have changed that so it is now in the 1x slot next to the GPU and I haven't had a BSOD since doing that - maybe 2 weeks now. SORTED! (hopefully, let's not jump the gun!)

The only downside to this is there is very minimal noise through the sound card now which as someone who loves a flac is slightly annoying but certainly less annoying than the BSOD's and its only really when listening very quietly, I am going to keep it like this for a while until I can confirm it is all working then maybe fiddle with some of the other GPU slots to see what happens.

Very bizarre all in all - I would had expected the GPU running at 8x would had been fine but clearly not, maybe its a bios update away from being sorted? I know for now at least everything is working fine, if anyone is interested in real life application of 8x vs 16x I can't say FPS in games has been affected at all.

Thanks for everyone's help much appreciated through and through hopefully this is the last time I will be posting on this thread!