Please reset BIOS to factory defaults, and that especially includes your OC settings. I've noticed that you are trying to reduce it but if you're currently diagnosing a stability issue with your system the first thing you should do is get rid of all OC settings, regardless if "it worked before". It will just make diagnosis a lot easier when potential candidates for problems like overclocking are taken care of first. If you must remember how you set them up previously just write down the settings before resetting them. Don't forget to reset your OC settings on your graphics card as well if you have OCed that and haven't reset it by now.
Afterwards, turn on
Driver Verifier and let it crash once or twice afterwards and send us the crashdumps. Make sure to read the entire article and follow all bullet points for the setup. There are some checks that will cause false positives, so make sure only to select those that are listed.
I may running on a rabbit trail here, but what connection are you using to hook to your monitor? If you're using HDMI, try switching to DVI or even analog. The reason why I mention this is because one of the DPCs that's showing up that appears involved with causing these crashes is from the HD Audio Bus for the graphics card, which I've actually seen cause problems for some other people. Of course, you can try a beta driver for Nvidia as well for Windows 8. Overall I think either the Nvidia drivers are buggy, or it may be getting held up by another driver like your audio drivers.
I'd like to peruse this further but I think minidumps aren't giving us all the info we need. With Driver Verifier on and producing more viable crashdumps we may get something from that, but we may even need to go a step further and get a kernel dump from one of those, which is the
MEMORY.DMP file located in your Windows directory. Zip up and upload to some site like Mirrorcreator.com. The kernel dump only is the dump for the latest crash that could successfully create a crashdump, and will be overwritten every time a crash occurs.
Analysts:
Each crashdump is identical. Raw stack output for one:
Code:
fffff801`95df3b10 00000000`00000003
fffff801`95df3b18 fffff801`95171f80 nt!KiInitialPCR+0x2f80
fffff801`95df3b20 fffff880`04547b70 nvlddmkm+0x12ab70
fffff801`95df3b28 000601db`000601da
[B]fffff801`95df3b30 fffff880`0511259c HDAudBus!HdaController::NotificationDpc[/B]
fffff801`95df3b38 000601db`000601db
fffff801`95df3b40 fffff880`04547b70 nvlddmkm+0x12ab70
fffff801`95df3b48 00000000`000601db
fffff801`95df3b50 00000000`00000000
Not sure if this is current DPC or a preexisting one for some previous I/O interrupt that got left behind on the thread stack, but it's a potential candidate. I can't see any other at the moment that could be involved.
!dpcs extension can help here but only available on kernel dump. Though I'm pretty sure this thread is the DPC.