Both Brand New GTX 980m failed

I was stunned when I turned on my PC and was greeted with black screen. Several hard reboots later pc came on and there were lines all over my screen. Again pc shut down. When pc turned back on both 980ms not recognized in device manager, or bios. Thought it may have been motherboard problem, but when I put my 880ms back it they worked fine. Both my brand new 980m GPUs have failed. I didn't even get around to do any bench marks. I tried the following

-Removed drivers, was using 347.09

-Tried FN+F5

-removed both 980 cards started pc. Put one in restarted-no good. Moved card to other slot, restarted same result. Did same thing with other card it did not work

-reformatted Hdd, reinstalled win 8-no good

Had friend bring over his 980ms inserted they both worked in my pc. In his pc my cards DEAD!


Not a MB problem just got new one, and both my 880ms work fine.

Anyone have any thoughts as I start the warranty wars.

What a great new year

 

Those cards get 30 days right? You should be fine for exchange.

 

I had it happen to me. Just send them back to get replaced.

 

Wow. Dead 980Ms like that. I'm a bit afraid now; it'd cost an arm and a leg if I ever got them.

 

Let me guess..... HIDevolution?

If yes, I can relate my woes as well http://forum.notebookreview.com/alienware-17-m17x/767611-hidevolution-never-again-5.html

 

I have to ask. I sure hope you did not try to use your old back plates on these cards.

 

Oh man, my condolences. I hope you used thermal pads on more than just the VRAM and ensured they touched the heatsinks properly.

 

Johnksss, not sure what you mean by backplates, but if your referring to the heat sinks those were fully modded as the pics said to do.

Cloudfire, everything that was a black square that rose above the card was covered with a thermal pad, this includes anything that was on the back of the cards.

Billyxq, yes these cards were gotten from HIDevolution. Needless to say that if I had seen your thread I may have gone another way.

Zero989, I've had these cards less than 2 weeks. Also, they have a one year warranty.

 

Both 980M's are dead? That's what I'm afraid of with all this "tweaking" and modification of drivers to get them working in the first place.

 

Yeah, not referring to the heat sinks, so that could pose a serious problem.
If you used this back plate with out adding the tape

Then chances are like 95 percent you fried both cards.
If you used this one.

Then chances are you installed the cards while there was still some power left in the board. And the shock to the cards took them both out.

Or if you flashed them both and tried to re-install them in legacy mode, they will not work either. (Alienware 18)

Speculation of course.

 

I didnt use backplate on my 970Ms. I just slapped on some thermal pads on the 6 VRAM on the backside and thought that would do just fine because they are like right next to the fan.
I did use x-bracket though.

 

I also used normal X-Bracket. But i have not put any thermal pad on the backside and all is good. Also temps are very nice.

 

Shouldnt need any either since thermal pads are mostly for transfer heat between two components. And its just pcb right next to them anyway. But I had some leftovers so why not

What HWinfo fan settings are you using btw? Have you found anyone thats not anoying lol?

 

Mine did not come with back plates, and I did not have any to use. The X-bracket was already installed and I used thermal pads on all the square black things on the back, and front of the cards (sorry for being all technical). The 980s I installed had way more thermal pads than the 880s they replaced. As possibility of spare juice, I hold the power button down for a minute than hold for another just in case. Didn't flash them, don't know how . They worked for 3 days (not even entire days) and died.

 

I am using the Clevo x-bracket (support plates) that came pre-attached to the GPU and no thermal pads on the memory. I didn't see any point in putting a pad on a component that has no place to send the heat. Putting a pad on might even increase vRAM temps without a heat sink. I figured air circulation would be more effective at cooling than using a pad to trap heat like a blanket. I think the Dell support plates with the memory heat spreaders would be better, but as Brother John pointed out, unless you insulate the corners you run a big risk of creating a short circuit.

You may have received GPUs that were just no good, or maybe were mishandled before you received them. I recently did some work for another Alienware owner and upgraded his M18xR2 to 780M SLI. One of the 780M cards lasted about an hour before it gave up the ghost. The replacement GPU was fine.

 

Mr. Fox are you saying next set of cards I get, DO NOT put thermal pads on the back of the cards? And what do you mean insulate the corners, I basically took my queue from how the 880s looked before I removed them.

 

Good deal!

As to pads on the back of the card. They do not need them.

Interesting though. Our cards came from rjtech. And everyone having problems came from hidevolution.? Will need more to chime in on this.

 

That's correct. Unless you get video cards that do not have a support plate attached to them already, in which case I would use new Dell support plates with heat spreaders as long as the corners are taped to prevent a short circuit. I do not recommend taking the support plates off of an old GPU and re-using them on a different GPU.

Here is what John and I mean by taping or insulating the corners: [Guide] Curing Shorted Support Plate on NVIDIA MXM Modules (Symptoms: Display "Static" or "Snow" Artifacts - Error Code 43 - Revision FF error)