>>31448
You mean if the machine's *not* stressed it crashes? Well now.

This is not necessarily a processor defect. The watchdog timeout means something is going tits up with interrupt processing, which can happen if the processor enters a low power state it doesn't get out of, or if there's a third-party driver that's inappropriately futzing with interrupts.

First make sure you have the latest BIOS and drivers for your motherboard. Then boot Windows into safe/recovery mode (or even better, do a completely clean install with no funky third-party or "vendor" software, but I'm guessing it's already clean since you're stressing it). If that doesn't do anything, try fiddling around with the power options (specifically, turn off any power saving measures either in BIOS or in Windows). If the power options do the trick, you still don't know if it's the processor or the motherboard.

>>31448
Are you overclocking it? If so, reset the mobo to use optimized defaults. Make sure you have all your chipset drivers installed.

>>31467
Yeah, my reaction exactly.
I forgot a full list of what I've given a shot at I guess, so here goes (as best I can remember)
-Started noticing the problem
-uninstalled drivers
-reinstalled drivers
-reset windows down to barebones OS
-reset bios settings
-cleared cmos
-reset settings again
-moved the gpu from one slot to the next one down
-moved ram A to slot B and ram B to slot A
-reinstalled drivers again
-uninstalled drivers again
-decided to run stress programs to try and figure out exactly what I have to replace
-stressing CPU magically solves problem

So anyway.
Gimme a few and I'll check for power settings, I've already disabled sleeping on Windows, ever, because it freezes it every time trying to wake up.

>>31448
Nope, was planning to at some point down the road but currently not in the itinerary.

>>31481
>I've already disabled sleeping on Windows, ever, because it freezes it every time trying to wake up
Heyo, big red flag here. This means suspend states aren't working properly, so that combined with the "it hangs when not stressed" definitely narrows it down to a power management problem.

See if suspend/resume works from a Linux USB stick. Yes, I know, but Linux is actually really useful for troubleshooting stuff even if you have no intention of running it on your desktop. If Linux hangs/crashes too, you definitely know it's a BIOS or hardware problem and not a chipset driver. I'm still not sure if it's possible to tell whether the processor or the motherboard is at fault without replacing either. Of course, you might just RMA them both if you got them together...

>>31493
I will dig out my very outdated Ubuntu stick and let you know then.

>>31502
mprime is a good way to stress the system on linux

>>31493
Well, after dicking around just browsing for about an hour and a half, no freeze so far booting into Linux off a stick. This implies some kind of driver conflict in Windows if I understand the diagnosis correctly?

>>31543
Yes it looks like its a problem with your install of windows. Run some more tests in linux to make sure. If so, do a fresh install of windows and install ALL needed drivers (gpu, chipsets, ...)

>>31585
I installed from a disc, will it allow me to wipe it off the system and then reinstall if I've authenticated once already? Or would doing the windows Reset be similarly effective?

>>31589
Windows 10 is locked to your system.

You don't even need to enter a product key; it asks the server if the computer you're on is licensed, and as long as you install the correct version, the server will say yes.

>>31591
Cool.
Well, time to venture back into hell. Here goes.