GPU fails randomly

[Arzor529]

Hey everyone,

I was wondering if anyone could help me out as i've never come across a problem like this before. Basically over the past month my GPU has been randomly turning off. This involves my two monitors being disconnected saying there's no signal and the GPU fans throttle up to 100% forcing me to restart the PC. The reason I believe the GPU to be failing is because my motherboard is still lit up, my fans are all still spinning and I can hear in-game audio and can still speak to friends through my mic when the screen goes black. This has been happening at random times. Firstly it was once every couple of weeks when I was gaming but it has began to happen more often and just tonight it has happened 4 times, one of these I wasn't playing any games. It has also happened across multiple games of varying demands eg. Elite Dangerous, Arma 3, Dragon Age: Origins and Star Wars Battlefront 2.

I originally thought it was an overheating issue so I have been monitoring my temps constantly for a couple weeks and my GPU under 100% load for extended amounts of time has not surpassed 68 degrees celsius. I also cleaned all of the dust out of my PC as i thought possibly the PSU was clogged causing shorting issues. The GPU was also cleaned at this point. Next port of call was to update my GPU drivers and even did a clean install but the issue has remained. I also reseated the PSU and checked the cables connecting to it. The only thing I have noticed is the GPU sags a bit at the front, could this be an issue?

My specs:

Intel core i5 8600k
MSI B360 Gaming Plus Motherboard
CoolerMaster MasterAir MA410P
Nvidia Geforce GTX 1060 6GB (MSI Armor OCV1)
16GB Corsair Vengence DDR4 2666mhz (2 sticks of 8GB)
Corsair CS650m PSU
4 x 120mm fans (air coming in the front, out the back)

I'm really at a loss with this one because i've tried to replicate the issue by running the Heaven benchmark to see if it would fail and there have been times i've ran it for hours and it's been fine, other times it fails after 5 mins. Any help would be greatly appreciaated!

 

[basco]

my guess goes towards the power supply or maybe memory
did ya try memtest for your ram. and how is your ram configured? running xmp?

Memtest86+ | The Open-Source Memory Testing Tool

Memtest86+ is an advanced, free, open-source, stand-alone memory tester for 32- and 64-bit computers (UEFI & BIOS supported)

www.memtest.org

plz fill out your complete system specs for easier helping

if ya have a friend with computer parts try to counter check

 

[Arzor529]

I have filled out my specs and edited above (sorry it's been a long night!)

I'm running memtest currently so should see the results soon.

I'm running XMP and I actually only turned it on yesterday as I didn't know about it.

If the memtest doesn't show any issues i'll try swapping parts with my brother to see if that narrows it down though it's hard to know what to swap as the issue is so random it's hard to get a consistent point of failure. Testing could take 5 mins with a new part or 5 hours just to make sure.

 

[basco]

no probs men all is good.
if your ram was running default until yesterday then i would test the power supply.
normally random errors points to it.

just do what is easy to swap first-so if your brother has other ddr4 then try this first and afterwards the power supply.

normally gpu´s sag at the end not the front-maybe show us a picture?

 

[Arzor529]

Okay i'll try swapping out parts, could maybe even try his power supply to see if that helps.

Maybe i'm thinking of the GPU backwards? The end furthest from the I/O is the part sagging though it's not hanging massively.

 

[basco]

little sag should be no prob.
is this nvidia reference 1060 ?

test only one at a time so ya know which part is at fault.

 

[Arzor529]

It's the MSI GTX 1060 Armor 6GB OCV1

Completed the memtest and got 0 errors with the RAM.

 

[Regeneration]

Try another PSU.

 

[Arzor529]

Try another PSU.

This will probably be my next attempt but I won't be able to test another PSU for a few days maybe. I'll make sure to update this post with my progress.

Update:

Since I can't test a new PSU straight away I decided to do a double check of my PCIe cable from the PSU to my GPU to ensure it was connected properly and while I was doing this I decided to swap the cable out for a new one just to cover that possibility. So far after about 4 hours of running stress tests on the PSU and GPU using OCCT I haven't had a crash. I won't say this is fixed just yet due to the sporadic nature of the issue but I'll keep an eye on it over the next while before I test a new PSU and see if the problem persists.

 

[Arzor529]

Update:

So after a week of running the PC i actually ran into a lot more issues with other aspects of the PC including lock ups, freezes, programs not responding issues with audio I/O not being detected and the PC was unable to restart or shutdown requiring a manual power down everytime.

I ended up updating the motherboard BIOS, clearing the CMOS and doing a fresh install of Windows 10.

It's been about 5 days since the reset of the PC and so far no issues including the GPU issue have appeared. I'll continue to keep an eye on this and will update here of the GPU failure happens again in which case i'll be buying a new PSU.