Most of the crashdumps are WHEA errors reporting that there was failure of bank 5 in the L1 cache on your CPU. The banks are pretty consistent, which would have mean evidence leans heavily on it very well being your CPU suffering permanent damage. The other couple of crashdumps that aren't WHEA errors are inconclusive, as minidumps do not retain the information necessary to do a deeper dive.
First of all, before anything else, if you have your system overclocked in any way, shape or fashion, reset all
of it to factory defaults. Even if your system has previously been stable with OCed settings, it can aggressively cause wear on your CPU and other hardware that will reduce their life, and your CPU may have just started suffering from it as a result.
Just to be extra sure, first I noticed that you have some motherboard utility software installed (like Asus PCprobe). Often times I've found WHEA errors that report bad CPU issues have actually been caused by bugs in motherboard utility software. Usually this software is gimmicky, or is better left done with alternative 3rd-party software options that are more reliable. If you still feel determined to keep your motherboard software installed, at least go to the appropriate website (Asus) and download any updates for all of your motherboard's related software, its drivers
, and its BIOS (just to be on the safe side). If problems persist, you may need to remove the motherboard utilities just in case (keep all the drivers).
In addition to this, on rare occasion I've found video driver
bugs to cause this (which is unusual from a technical standpoint, but happens). I did notice your ATI drivers are from Dec 2011, so you may wanna try and update those.
Lastly, you may check your CPU with Prime95
. Set it to Torture Test with Large FFTs and let it run for about 9+ hours. If your system BSODs during it, send the resulting crashdump. Otherwise, if you see any errors you can start betting that it's the CPU that's being problematic (though no errors at all does not
100% ensure your CPU is ok). If you wish to see if voltage (PSU) or temp is being an issue, you can use HWinfo
with Sensors only
option checked to see. Log two 30-minute sessions: one during idle and one during high load. Send the resulting logs. If your system crashes during one of them, don't worry, it'll log up to the latest point before the crash. Send the resulting logs for analysis.
Given the information so far, I'm pretty heavily set on it being CPU failure, but these are all just measures to make sure anything else isn't the cause.