From what Magus saw, I did notice that the NTSTATUS error code for all the 0x116 bugchecks (Arg2 of the bugcheck) was the same: c000009a, which means "Insufficient system resources exist to complete the API.
" This means either that a request was sent to the video drivers that required more resources than was allocated to the drivers, or that the amount allocated to the drivers was smaller than what was requested. This sounds like the same thing mentioned both ways, but it means the request was either set too large, or the resources allocated was set too small. There's also a third possibility in that whatever resources allocated were sufficient, but were inaccessible at the time (most likely due to hardware failure). Either way, the driver as a result caved in, couldn't handle the request, and bugged out. Most likely we won't be able to figure out more details from just these minidumps.
You will want to first ensure that you aren't running out of neither virtual memory (paging file) nor physical memory (RAM). Second, to ensure that the VRAM on your GPU isn't whacking out, you should run MemtestCL
on it a few times. I am aware that you got the video card sent back to you, but honestly, I think if your card has been permitted to run several times at that high of temperature, that I wouldn't doubt there'd be some manor of permanent damage wrought on it.
In addition to the previous Memtest, you'll want to run regular Memtest86+
on your RAM to ensure it's not that resource it's talking about that was bugging out.