Crap you guys not, get my two Titan X's installed, get into Windows, get the latest NV driver installed, reboot, turn on SLI, go into Furmark, and start the 1080 benchmark.
 
PSU clicks off immediately.
 
Hmmmm...ok...maybe just a fluke.  Hit the power button again...nothing PSU is staying off.  I unplug the PCIe power connectors from both cards and try to turn it on again.  *zap* and smoke comes out of the top card. 
 
Sigh...really?!  I can't find any blown components on the backside of the PCB so I'm guessing it happened somewhere under the fan area.
 
Now I have to figure out if I bother with EVGA support or Newegg support.  Which one will get me a replacement faster?

Didn't even realize people were getting the EVGA cards in, lol. I thought they were still coming from Nvidia only after seeing someone say they would ship next week (Monday).

Sending it back to Newegg or through EVGA will be easy. They both seem to have great support honestly.

I thought that FurMark was a tool that we should no longer be using on our Graphics Cards.
Also "5960X 4.6GHz" I think it would be best to reset your Bios to the Default or Optimized settings for a New Video Card.

More likely than not the card wasn't put together right and a solder stringer or piece of the heatsink dead shorted itself.  The PSU clicking off was probably the overcurrent protection.  Not sure why it fully exploded when I unplugged the PCIe power connectors from the card unless the overcurrent protection for the PCIe lanes through the 24pin connector has different protection.

Scarlet-Tech
Unless changing the strap to 103 or something ridiculous, the cpu overclocking doesn't usually cause video cards to explode.

As far as I understand, the strap only effects certain things, like sata ports and such, which get over clocked as well. It doesn't overclocking the PCB lanes, unless I understood incorrect.

It would include the GEN to Auto and not knowing more it is hard to say.
More so setting the BCLK as some users do to overclock, having 125 or 135 BCLK could cause issues on the PCIe Channels.
The OP did say that he pulled the Power Cables off of the Graphics Cards and then turned on the computer and then the cards smoked.
Myself I would reset my Bios and start a new Overclocking process with new Cards.
Like was a tool install that set the voltage or clocks as the OP did not say.

bcavnaugh

Scarlet-Tech
Unless changing the strap to 103 or something ridiculous, the cpu overclocking doesn't usually cause video cards to explode.

As far as I understand, the strap only effects certain things, like sata ports and such, which get over clocked as well. It doesn't overclocking the PCB lanes, unless I understood incorrect.

It would include the GEN to Auto and not knowing more it is hard to say.
More so setting the BCLK as some users do to overclock, having 125 or 135 BCLK could cause issues on the PCIe Channels.
The OP did say that he pulled the Power Cables off of the Graphics Cards and then turned on the computer and then the cards smoked.
Myself I would reset my Bios and start a new Overclocking process with new Cards.
Like was a tool install that set the voltage or clocks as the OP did not say.

The voltage is locked on Titan X's right now.  I had just installed them and hadn't even messed with any of the clocks yet.  The second card is what I'm using right now and it overclocks like a monster though so we're 1 for 2 :)

Scarlet-Tech
Are you 100% positive you disconnected the car battery from the card as well? I heard connected a deep cycle gel cell battery would add up to 400mhz on the core without changing settings.. But it also requires the PCB cables to be plugged in. If you forgot to disconnect the car battery, that may have been the problem....

LoL. Or chalk it up to, "weird crap occurs," and pull the shroud real quick and get some pictures.

Sometimes I just don't get you.

It's called a joke. People make them occasionally in light of a bad situation. Basically, crap occurs and the card blew. Plain and simple, it happens. Pull the shroud and see what gave out. (this part is more of a request, although it looks like a demand, lol)

AnonymousGuy
I Wish you luck with your new cards. Later

Why would you use furmark on a Titan?

bcavnaugh
I thought that FurMark was a tool that we should no longer be using on our Graphics Cards.

 
Yeah, we were supposed to stop using FurMark because it caused voltage regulator components to fry on brand new cards in the past.

I guarantee that the OP's issue is a fried MOSFET.  The component overheated, shorted, and caused the PSU to trip off.  When he disconnected the GPU power cables, the load was low enough to allow the PSU to turn on, but the power fed through the PCI-E connector into the card and continued to burn components on the card until it was noticed.
 
I am sure that there is at least one fried MOSFET looking something like this somewhere on his card.

 
Whose fault is it?  NVIDIA's.  The card should accept anything you throw at it, self-control, and self-regulate.

Sajin
Actually the NVIDIA driver throttles the card down when running furmark as furmark is blacklisted.

Maybe not enough? Maybe the card has a defective or poorly engineered protection circuit? Maybe his card was defective? Regardless of the cause, I am almost positive of what failed given the description and past experience.

Sorry to hear about this problem with the Titan X and it's likely just a one time occurrence.
Hope the OP gets an RMA replacement ASAP, But this does make me want to wait awhile longer before I decide on any purchases.

I found what component died.
 
Pics coming shortly.
 
Also EK's instructions for how to remove the stock cooler are missing the 2 screws on the PCI-e face plate.

You links to your images to not link to the Images. FYI Only
 
So did the Cooler cause this or was 1R0 cracked and that caused it or was the burn out/short out cause it to crack?