Yeah... I'd also go for a RAM test on this beast... Just out of interest, I'd also be interested in what wattage PSU you've got in it... Those 8800GTs use quite a bit of juice. You might just be suffering drops in the supply voltage or the effects of noise on the mains supply. Particularly if you're pushing the capabilities of your PSU.
Hiper aren't known for their reliability... not lately anyway... in the past they were acceptable.
Go with ram tests, if they come out positive try stressing your CPU with Prime95 to see if that trigers a bluescreen. Could also be your video card getting old... I've seen the IRQL and page fault BSODs before from overly ambitious overclocks, so I'm guessing your RAM's the issue.
If memtest does not come back as having failed. Re-install your OS (do a clean wipe of the disks), I have seen some issues with the windows install becoming corrupt and these BSOD's popping up:
page fault in non paged area
bad pool caller
Could also be down to a timing issue... I had this a couple of years ago on a Barton 2500+. Existing memory was good (2x256MB). So was the new stuff (1x512MB). I knew this as I'd tested the 512MB stick on another board at the local computer club where I'd brought it second hand. The problems started when I added the 512MB stick to what was already there, so I knew the existing RAM was good. But when both were running together, I was consistantly getting errors when running MemTest x86+. So I repeated both tests, first with the existing 2x256MB and it passed, then with the 1x512MB on it's own, and it passed.
All sticks were rated at the same speed, but the 512MB device was a different manufacturer. Turned out to be timing causing the problem here.
That's not the strangest issue with RAM that I've experienced though... I'd built this home-brew server for the back room from a 1.3GHz P3 mainboard and a couple of HDDs, one 13GB for the Linux OS, software & swap etc and a 60GB for the /home partition which contained the website and other stuff for sharing on the local LAN only. I had it running Debian for a while with 256MB of RAM and was planning to upgrade it to 512MB. I was in the process of mix 'n matching RAM for this machine, again from known good stocks at the local computer club and was getting some weird and wonderful reports from the BIOS. Like 256MB + 64MB = 128MB.
A BIOS flash fixed this board... What I found really interesting though was taking the backup dump of the existing BIOS that the flash program saved and the file that it programmed into the chip and putting them through a CRC comparison program produced an exact match, so why re-flashing the board fixed it is still a mystery.