Yep, it's the motherboard, you need to use Power Boost.
http://www.evga.com/products/moreInfo.asp?pn=100-MB-PB01-BR&family=Accessories%20-%20Hardware&sw=4

HotRodPolk

Ouch! You better replace the mobo as well. Where it's at suggests it's mobo related. If it were the supply, it'd be burned at the other end...

What do you mean it will be burned at the other end? The mobo end or the power supply end? Because to me both sides seemed burned. Could the problem be caused by overloading? I normally don't turn my computer off and i leave it folding at nighttime. 

I forgot to post the picture for the 24pin topside view.

So in your opinion do you think i can rma both the mobo and the psu? The psu i don't smell burned inside it, so i think it's fine, it's just the wire i think then that is melted or the sheilding around the wires is melted.

Are there any additional molex ports for additional power to your PCIe x16 lanes on that Gigabyte board? If so are you plugging molex leads into them? If folding for long periods of time on all PCIe x16 lanes (i.e. many GPUs on that motherboard) and there isn't enough power getting to the lanes that is probably why your 24 pin has melted on that pin as suggested above.
 
If you find molex ports that you never plugged a molex lead into (typically located just above and to the left of your top most expansion lane) then plug a molex in there (can also be located underneath bottom most expansion lane).
 
Either that or buy a power-boost from EVGA as previously mentioned.
 
Good luck with the RMA(s)!!

Hey guys, is that PCIe add-on power adapter needed if you only have one 580 and a sound card in your PCIe slots? I am getting all paranoid after reading this thread now.

chrisdglong

Hey guys, is that PCIe add-on power adapter needed if you only have one 580 and a sound card in your PCIe slots? I am getting all paranoid after reading this thread now.

 
I've never seen pins 10 and 11 melted on the 24-pin connector on a non-SLI'd system.  This only happens on multi-gpu systems.  It's not a fault of the power supply.  It's just the PCI-E subsystem drawing more power than the +12V connectors were designed to handle on the 24-pin connector.  Power Boost will solve this issue as will any mobo with supplemental power to the PCI-E subsystem.....either molex or 6-pin connector.  If you have a single card there is nothing to worry about.

And I bet it'll be years before motherboard manufactures actually decide to redesign the whole way a motherboard works so this won't happen again

i know this post is kind of old, but i didnt find the recent one, the one where are more details about 24 pins burned, i had this problem also, well not really burned but almost did burn it, the problem is the power supply, i had a ms-tech 750w val, on a q6600 clocked to 3.2, evga gtx 460 stock clock 720 small overclock to 800 core and 2000 memory... so the 24 pin was getting pretty hot i didnt know why... but i seen problems like freezes, low voltage and blue screens, so i changed the psu to a fortron hyper 700w, now same settings overclocked, and the 24 pin is ice cold np now... so in other words for me was the powersupply the problem on 24 pin getting hot... 
 
 it seems the problem comes when your components suck too much power and the psu cannot give enough...

hmm i think there is another problem... not that it draws to much power... i think it happens when the power supply cannot deliver enough for all components but then again after i put the new power supply no cable got hot... and with old one the 6 pins or the 8 pin for cpu did not get hot instead 24 pin was... and yep you are right like the quality is not good... for example the old one says 750w but i doubt it it was delivering 750w... not even 500 i dont think it could... and my motherboard dont have a molex connector on it... now like i sayd its ice cold with a good quality psu... and i dont think it helps much if every 12v connector have different lanes... it could cause instability in my opinion...