Questions after PowerColor HD 7990 Overheated and Died

cwmabray

Honorable
Oct 17, 2013
1
0
10,510
Hello there, I was wondering if anyone knew what would cuase a Hd7990 overheat and brick with only about 30mins of game play?

Here is some background
I just received a new Powercolor HD 7990 http:// two days ago from NewEgg. Yesterday I was working mainly on MS Office and I noticed that the temps were staying around 40c (from catalyst control center). I thought this was normal idle temps, so today I thought I would try the card out on BioShock Infinite. I played on ultra at 1080p for about 30 mins and just about made it through the tutorial; when I heard a series 3 or 4 pops, then my computer crashed and I started smelling this terrible burning smell. I immediately unplugged the PSU and opened the case, and found that card was EXTREMELY hot and was emanating a pungent smell. After cooling under some external fans for about an hour I plugged the computer back in and found that the power light would only flash and make a click and not even make it to BIOS. I swapped out cards back to my old HD 5670 and my computer was working just fine. Later I tried the card in another computer and that made that computer unbootable as well. I sent it for a RMA replacement and it should be here sometime next week.

So before my other card arrives I have a few questions:
After thinking about, I don’t think I heard the fans blowing as hard as they should have been, so have you ever seen a diver fail to control the fans or something? Plus I thought GPUs would shut off before frying?

Do you think there is hardware issue with my computer that caused this burn out?

Do you think that there is enough air flow within my NZXT Phantom ATX case? I have 1 front 140mm, 2 side 140mm plus a side 200mm, rear 140mm exhaust fan, and top 200mm exhaust.

I was also wondering, for the other 200mm for the second fan port top, would it be better for it to be an exhaust or intake?

Specs:
i7 2600k (stock cooler)
Asus P8P67
8GB of corsair Dominator GT RAM @ 1866MHz (with a ram cooler)
720GB 7200RPM HDD WD
Corsair CMPSU-750TX 750-Watt
NZXT Phantom ATX Full Tower

 
Solution
Difficult to say. You could have simply been shipped a product that passed quality control check one to fail shortly afterwards. No mass produced product ever made has ever had a 100% flawless track record and there will never be such a product ever. Hopefully, the next one will not give you any issues.

I recall reading a post at HardForum where one unlucky guy had 3 different graphic cards die in 1 year. Somehow eventually it was diagnosed that the PCI-e port or the 6 pin connector (I forgot which) was providing too much power and fried the cards.
I suspect overheating wasn't what happened, most cards can run well over 100C without damage (very short term obviously). Whenever I've had a card overheat, both Nvidia and AMD, the system black screened and wacked the fans up to full, that would probably have been what would have happened if it were overheating. It probably had some other issue and just happened to go pop.

 
Difficult to say. You could have simply been shipped a product that passed quality control check one to fail shortly afterwards. No mass produced product ever made has ever had a 100% flawless track record and there will never be such a product ever. Hopefully, the next one will not give you any issues.

I recall reading a post at HardForum where one unlucky guy had 3 different graphic cards die in 1 year. Somehow eventually it was diagnosed that the PCI-e port or the 6 pin connector (I forgot which) was providing too much power and fried the cards.
 
Solution