Question Is my 1070 irreparably damaged?

[scipsych]

It appears my GTX 1070 GPU overheated to the point of burning a portion of its circuit board last night. I was playing a game when my PC shutdown without warning and then attempted to reboot itself several times before shutting off completely. This all happened in the span of several seconds. This was quite startling, as the game (Halo Wars 2) had been playing smoothly and I hadn’t noticed any indication that anything might be wrong.

As soon as I crouched down to take a closer look at the machine, I noticed a strong burning smell that seemed to be coming from the middle interior of the case. I immediately unplugged the computer and switched off the power supply. The case was rather warm to the touch, so I let it cool before opening it up to determine what component had overheated. When I did so, I found the source of the burning smell rather quickly. I’ve linked images below to show what I found.

https://imagizer.imageshack.com/img921/4672/ea1vq1.jpg

https://imagizer.imageshack.com/img922/163/mnM70r.jpg

https://imagizer.imageshack.com/img924/6885/03Of1N.jpg

https://imagizer.imageshack.com/img921/121/m9l78E.jpg

As you can see, a portion of the graphics card has suffered significant damage due to overheating. I’ve had this GPU (the entire machine, in fact) since December 2016 and have never had problems with any of the components overheating. I have never attempted to overclock the CPU or GPU, either.

Among other concerns, I’m left wondering what might have caused this and whether there is any chance the GPU can be salvaged. I would greatly appreciate any insight anyone may be able to offer.

As the next two images illustrate, I haven’t been able to find any other evidence of damage to other components. If it would be helpful for determining whether there may be further damage, I could remove the GPU and upload additional pictures.

https://imagizer.imageshack.com/img921/496/31CI5r.jpg

https://imagizer.imageshack.com/img923/7382/odraqS.jpg

I haven’t attempted to boot the computer since this occurred. Given that the damage seems to be localized to a small section of the GPU, is there any reason to be concerned that other components might have been damaged? Would it be safe to attempt to reboot the computer after removing the damaged GPU? I’m concerned about losing data and don’t want to risk any further damage to the machine.

Specs are as follows:
GPU: EVGA GeForce GTX 1070 8GB
CPU: Intel Core i5-6600K 3.5GHz (CRYORIG H7 49 CFM Cooler)
MOBO: MSI-Z170A KRAIT GAMING 3X LGA 1151
PSU: EVGA – SuperNOVA G2 650W 80+ Gold Certified
RAM: Corsair Vengeance LPX 16GB DDR4-3000
HD: Crucial MX300 525 SSD; WD Caviar Blue 1TB HDD
OS: Windows 10 Home

Sorry for the long post; I just wanted to include as many relevant details as possible. Thanks for reading this far!

 

[Ralston18]

No way to be sure what may happen.

My thought is that that label just caused the covered GPU components to just run hotter than they should have and that the cumulative heating effect finally degraded them to the point of failure.

With no other signs of damage the problem is most likely localized to the GPU.

You should be able to remove the GPU and use either the motherboard's DVI or HDMI port for video.

Most likely you can pull the drive and install it in and external USB case with its own power. Then copy the drive(s) (clone or image) or just the files you wish to have backed up.

SSD is boot drive, HDD is data - correct?

Then reinstall just the boot drive, connect monitor via one of the motherboard's two video ports, and hope for the best.

 

[NightHawkRMX]

I wouldnt try that card since it could cause more damage. You could try taking the cooler off of the card to see what got damaged on the other side of the pcbm

 

[scipsych]

No way to be sure what may happen.

My thought is that that label just caused the covered GPU components to just run hotter than they should have and that the cumulative heating effect finally degraded them to the point of failure.

With no other signs of damage the problem is most likely localized to the GPU.

You should be able to remove the GPU and use either the motherboard's DVI or HDMI port for video.

Most likely you can pull the drive and install it in and external USB case with its own power. Then copy the drive(s) (clone or image) or just the files you wish to have backed up.

SSD is boot drive, HDD is data - correct?

Then reinstall just the boot drive, connect monitor via one of the motherboard's two video ports, and hope for the best.

Thank you for replying so quickly!

Yes, SSD is boot and HDD is data. I don't have an external USB case but will purchase one so that I can back up the drives as you've instructed. Just to be clear, your recommendation is to not attempt to boot the computer even without the GPU installed until I have backed up the drives, correct? I had hoped that if the damage is localized to only the GPU, then I would be able to remove that component and then boot normally using the motherboard's video port and then back up the data using external USB drives. It sounds like this may be risky, however.

My first thought on seeing the damage was that the label may have contributed, but since I had professional assistance with building the machine I presumed that leaving such labels affixed to the card was common practice and would have been accounted for in the manufacturing process. I'll not make such presumptions in the future.

Assuming the damage was caused at least in part due to the label having been left in place, do you happen to know if overheating due to this sort of oversight would typically be covered in the manufacturer warranty? Since the standard warranty is for three years, the card should still be covered. Given that this damage has rendered the card inoperable, I intend to contact the manufacturer regarding the possibility of replacing or repairing the card.

 

[NightHawkRMX]

Most gpus have a few stickers over them. This wouldnt cause this. Id try to get it replaced under warranty before trying to repair or even use the card again.

 

[scipsych]

I wouldnt try that card since it could cause more damage. You could try taking the cooler off of the card to see what got damaged on the other side of the pcbm

Thank you for the warning and for replying so quickly. I use this machine for almost everything and it was a large investment for me, so I've been hesitant to do anything before consulting others more informed than myself. I'm glad my caution paid off for once and prevented me from immediately trying to boot it back up

As for taking a closer look at the card itself, I'm concerned this may void the manufacturer warranty. That's assuming this sort of damage would be covered in the first place, of course. If it isn't, do you think there is any chance the card could be repaired at a cost less than purchasing a replacement? Being a novice myself, I would not have confidence that I could effectively perform diagnostic work on the card, to say nothing of actually repairing it myself.

 

[scipsych]

Most gpus have a few stickers over them. This wouldnt cause this. Id try to get it replaced under warranty before trying to repair or even use the card again.

Ah, you've answered the question I had next before I even asked it! Thank you. I intend to contact the manufacturer tomorrow to see about my options regarding repair/replacement.

 

[Ralston18]

Boot without the GPU and only the SSD.

And do pursue the warranty on the GPU.

I do agree with @remix that that the sticker should not necessarily have caused the overheating. However, the sticker's placement seems just to coincidental and I really do not have any idea if that is where the sticker should have been placed or how much pressure etc. was applied to get the sticker stuck down.

One suggestion: go to EVGA's website and check their forums and FAQs before you contact them.

Wondering if anyone else is reporting, asking, complaining about similar problems. May help your build "your case".

Will be very interested in how the EVGA responds... please post accordingly.

 

[Racksmith]

Damage such as that is classed as "component failure" you should be good for an RMA.

 

[scipsych]

Boot without the GPU and only the SSD.

And do pursue the warranty on the GPU.

I do agree with @remix that that the sticker should not necessarily have caused the overheating. However, the sticker's placement seems just to coincidental and I really do not have any idea if that is where the sticker should have been placed or how much pressure etc. was applied to get the sticker stuck down.

One suggestion: go to EVGA's website and check their forums and FAQs before you contact them.

Wondering if anyone else is reporting, asking, complaining about similar problems. May help your build "your case".

Will be very interested in how the EVGA responds... please post accordingly.

Damage such as that is classed as "component failure" you should be good for an RMA.

Thank you, Ralston18 and Racksmith. I have submitted a report and request for RMA service to EVGA. I verified that the GPU is still under warranty, so hopefully EVGA will repair or replace it. I will keep this thread updated as I receive more information from EVGA. I included the same images and information as I posted here with my report, so EVGA should have all they need to assess my request for service.

Meanwhile, I've purchased an external USB drive case so that I can check and back up the data on my HDD and SSD (the SSD is the boot drive, but I have data saved to it as well). Once I've been able to back up the data, I'll remove the GPU and HDD and attempt to boot the computer using only the SSD and the video output from the motherboard.

Also, I wasn't able to find any questions about similar issues on EVGA's FAQ. Perhaps this issue is a rare one? I'm hoping EVGA will be able to tell me exactly what went wrong when they examine the GPU, though I don't know if they share that information as part of the usual RMA process.

Thank you all for your advice! I'll keep this thread updated as I learn more.