Help, my GPU died!?

[Onomatopesha]

Hey guys, i bought myself a few weeks ago an XFX DD 7970 (TDJC) and so far, it was working as expected, a bit high temperatures (86°C and 90°C on extreme cases - these of course are when stress testing the GPU) but i was having no problems beyond that.

A few days ago, the PC just stopped working, and wouldn't turn on. After a few hours of testing (trying different PSUs, different RAM sticks, other mobo and the like) i noticed that whenever the GPU was plugged in with its power connectors, the PSU would automatically shut off. When the GPU was not powered up (plugged in only in the pcie slot), the fans would spin, but of course i would get no detection nor display from it. 

 

For the moment, i'm running on my older gpu (a Sapphire 7850 OC Edition GPU) on the same slot, with the same setup i was running prior to it, and having no issues at all.

The setup in question is the following:
Mobo: Gigabyte g1.sniper m3 LGA 1155

CPU: i5 3570k 3.8GHZ (stock cooling for now)

RAM: Kingston HyperX Blu DDR3 2x4GB 1600MHZ
GPU: -new one- XFX 7970 DD TDJC

         -old one, currently using- Sapphire 7850 OC Edition

PSU: Corsair AX760

The question here is, or rather, why i'm asking for help, is to diagnose, if at all possible what could've made the GPU died, or if it's dead at all (considering the fans spinning when not plugged in) before i send it to RMA and warranty, who they haven't responded yet, so as to give them a solid reason instead of just saying "whaa my gaming thingy doesn't work!"

I'll try to update the thread when the GPU has been shipped, so if you guys think of anything i could try with it before i do, let me know.
Any suggestions are welcome.
Thanks!

[Milesgaming]

Try a different gpu, in the exact same system

[Bluejay0]

Well if it's the GPU, XFX has a Lifetime Warranty so no worries.

[LDG]

Try a different psu or a low power card.

[Onomatopesha]

I have tested the 7850 (i'm using it right now) on that same system. I also tried a sentey GSP-650W (54Amps on the 12v rail, 80+ gold rating) with both gpus, and whenever i plugged in the 7970, the PSU's failsafe would trigger and shut off the system.

[airdeano]

I have tested the 7850 (i'm using it right now) on that same system. I also tried a sentey GSP-650W (54Amps on the 12v rail, 80+ gold rating) with both gpus, and whenever i plugged in the 7970, the PSU's failsafe would trigger and shut off the system.

 

sounds like the GPU went to GPU heaven. was there any loud cracks or pops

before the last power down? sounds like the VRM went down. do you see any

external scarring or burn marks on the back of PCB? do not take card apart before

contacting manufacturer! let them make the determination. some are really picky

over GPU tampering and can void warranties.

[Onomatopesha]

sounds like the GPU went to GPU heaven. was there any loud cracks or pops

before the last power down? sounds like the VRM went down. do you see any

external scarring or burn marks on the back of PCB? do not take card apart before

contacting manufacturer! let them make the determination. some are really picky

over GPU tampering and can void warranties.

I haven't noticed anything weird in the VRM's (looking from the backplate), however i did notice some scarring (like lines that don't follow the circuit board, and are not simply marks made by my fingers apparently) around a few capacitors near the display ports and right in the middle at that same height (again, in a few capacitors). And now that i pay a closer look into it, it seems that the power plugs part of the PCB seems like it had a really close encounter with heat (the soldering is ok, but there's a lot of scarring around each pin).

It's still kinda strange that the GPU's fans still spin while plugged into the pcie slot.

And yeah, thankfully i didn't even try to pull it apart, it's way too risky and i don't feel like voiding my warranty right before i ship it.

[airdeano]

I haven't noticed anything weird in the VRM's (looking from the backplate), however i did notice some scarring (like lines that don't follow the circuit board, and are not simply marks made by my fingers apparently) around a few capacitors near the display ports and right in the middle at that same height (again, in a few capacitors). And now that i pay a closer look into it, it seems that the power plugs part of the PCB seems like it had a really close encounter with heat (the soldering is ok, but there's a lot of scarring around each pin).

It's still kinda strange that the GPU's fans still spin while plugged into the pcie slot.

And yeah, thankfully i didn't even try to pull it apart, it's way too risky and i don't feel like voiding my warranty right before i ship it.

 

the GPU also gets support voltage through the PCIe slot (voltage source).

then let the support of the manufacturer figure the issue out so your warranty stays intact.

[Onomatopesha]

I know this is late, and i'm kinda necroing by now, so my apologies but....

After this thread and many other in different forums, i contacted the retailer, sent them the GPU which thankfully, they managed to send back to XFX (as XFX contacted me saying that, as a final user, shipping would be impossible due to import taxes and policies and the difficulty those things imply to them and myself); After quite some time, they contacted me saying that XFX just shipped them a freshly new GPU, and it should be arriving back here in a few days, so thankfully this has been solved with no issue.  

 

I'll try to contact them once again and see if i can get a response in regards to what exactly was the issue with the GPU, just out of curiosity. If i do, i'll update the thread with the answer.  

 

Thanks a lot for the help!