Hello,
I've just recieved my 1080 FTW card (I am a second owner), and the problem is that it crashes when it reaches 72 Celsius degrees. No matter the game (Witcher 3) or application (Valley, TimeSpy). Every single time it hits 72 the game/application freezes, no reboot, no blue screen, just freezes and takes me to the desktop). Just before it crashes there are some artifacts poping on the screen - black squares to be precise.
If I change power limit in MSI Afterburner to minimum, all is well, nothing crashes, temps are stable at 55 degrees. If I max it out it crashes as at normal settings.
No OC is applied to GPU it runs at stock values.
 
I used DDU before I install the card and new (latest) drivers. My PSU is 650 Watt from BeQuiet.
 
Is my brand new (old) card dying?
 
Thanks for any help.

Thanks for the reply. I will check the performance on all default settings/no OC.
 
At the moment case is open, no side panel. Inside I have 5 Silent Wings 140mm, 2 pushing in and 3 pulling out.
 
Still, it is only 72 degrees, card limit is at 83, is not it? I am way below the limit. Should not it perform as usual at 72 no matter the heat inside (as long as card is below 83)?
 
I have X99 platform with 6800K mildly OCed to 4000MHz and cooled with BeQuiet Dark Rock Pro 3. Also Corsair Vengance LPX 3000MHz CL15.
 

kingisdead
Hello,
I've just recieved my 1080 FTW card (I am a second owner), and the problem is that it crashes when it reaches 72 Celsius degrees. No matter the game (Witcher 3) or application (Valley, TimeSpy). Every single time it hits 72 the game/application freezes, no reboot, no blue screen, just freezes and takes me to the desktop). Just before it crashes there are some artifacts poping on the screen - black squares to be precise.
If I change power limit in MSI Afterburner to minimum, all is well, nothing crashes, temps are stable at 55 degrees. If I max it out it crashes as at normal settings.
No OC is applied to GPU it runs at stock values.
 
I used DDU before I install the card and new (latest) drivers. My PSU is 650 Watt from BeQuiet.
 
Is my brand new (old) card dying?
 
Thanks for any help.

You're describing the classic symptoms of a failing GPU. While there's a minuscule chance it's your PSU not providing stable power under maximum load, the only way to know that for sure is to swap out hardware. The 99% solution is to RMA the GPU.
https://www.evga.com/warranty/rmaprocess/

kingisdead
The case is it works stable (@ ~1950MHz) under maximum load for a few minutes until it gets as hot as 72 degrees, and then crashes. I would assume it gets stable power as it does not crash before hitting 72 degrees.
 
Is it possible that there is something not right with VRAM or RAM cooling? And either VRAM or RAM temps are causing the trouble?

I read your first post and your following post. My advice is exactly the same. Unless you're going to disassemble the card yourself, what point is there in discussing what's 'possibly' wrong? If you can't repair it, it doesn't matter.

Artifacts is bad in general, but strange that it's crash at 72 degree al the time. GPU have top temp 91 (GP102). So, maybe something else is overheating on the card?
 
Maybe you may show some more sceens (how hot MEM or PWR)? All fans are spin when card is working?
 
>>>Well, I can disassemble the card and reapply pads on vram and ram and change the paste on die, no problem with that.
 
I did so, and i may say it is a problem. Standard pads may leak and clean PCB and other elements on PCB is a PAIN. After that need to not only do custom cut pads, but also know what size they are. I used 1.5 and 1.0mm, at one spot was 2.0mm.

kingisdead
How can I check MEM and PWR temps?

Install HWiNFO64 for example (it's have Sensor button in menu). Run sensors monitor with opened window start benchmark or game to force-to-crash and then return to HWInfo. Look for maximum temps.

kingisdead
I've got HWiNFO, not sure where the temps you are reffering to are.

OMG, i am stupid. Some EVGA cards have got iCX Technology, some have got iCX Cooler. It's not the same - iCX Technology have got 9 additional sensors all other the card. I see you screenshot and can't see that data (obviously!). But this screenshot is way better - we may see more data about card. Voltage, load, power usage.
 
Can you test card on another PC? With another (probably better PSU)? If first owner didn't have such problem and you got it - maybe problem in you system. Maybe GPU really requesting more power from PSU, but it's can't give it and card crash.

Thank you all for the help, but this story ends here, as I decided to send the card back to the seller. Luckily he does not make a problem with that. I do not have time for RMA and all.
This forum is the best I have ever used, hence I guess I will go with brand new 2070 FTW.