Build was new in August. I've been having hard lockups in Trine 2 (and only Trine 2; code 41 errors mostly, I think) and had been stepping the graphics down to see if it would stop; this was weird enough already, since a 660ti should really be able to handle Trine 2 without even blinking. The system's run Assassin's Creed Revelations and Borderlands 2 at max, as it should.
Well, this last time it crashed, the screen went bright red, I powered it off with the button, and when I turned it back on it sputtered, whirred, single-long-beeped and shut back down again. Opened it up, discovered that the hard drives were spinning up but the fans on the CPU and GPU weren't spinning. Took the GPU out and booted the system again; everything seems to be fine for about ten minutes.
Then it crashes again, just as hard. I start poking at cables; I notice the ATX cable (24-pin) doesn't look quite right. Start struggling to get it out and discover that the +12V pin has *melted into the port*, at which point I start freaking out. I haven't melted a component in over a decade, and those were ancient parts.
So, what happened here? Do I RMA the mobo, the PSU, or both? Will that fix anything or will this just happen again? I haven't found anyone else that seems to have this problem that quite fits, or offers a useful solution.
Mobo: ASRock Z77 Extreme4 Z77
GPU: Gigabyte 660ti
PSU: Corsair TX650 V2 650W
Cooler: Cooler Master Hyper 212
SSD: 128gb Samsung MZ-7PC128D/AM
RAM: 4gb x2 GSkill
A variety of other HDDs and a Samsung DVD/RW
Well, Newegg's RMA window on these parts is 30 days, so they told me to contact Corsair and Asrock myself. Neither of them is answering the phone, and I'm not sure how to go about this anyway since i suspect they'll both try to blame the other part. I'm a web developer; I have clients waiting for me. Does anyone have any suggestions?
I did try cleaning out the ports as a stopgap if nothing else, but there's just too much cold melted plastic in the female end to make a connection.