As someone else said, PSU is a prime suspect here.
Including that you should go every one of your connections. I've repeatedly run into a case where I setup a system in an area with a lot of room and light and then moved it to its final resting spot only to have sporadic failures.
In each of those cases, one of the modular PSU connections was either only partly in from the beginning or had come loose when I moved the system.
Go over your cables and connectors carefully, re-seating as needed.
Next, I'd run a memory test. memtetst, aid64, or linux+stressapptest if you can get that setup (I often use a thumb-drive "live OS" linux setup for this).
Of course when you are checking all your cables, check your cards and DIMMs as well. Sometimes cases are poorly shaped and put enormous stress on the PCIe slots. So, over time the card pulls out a little. Any card with the right ground or power pin lacking a good connection could ultimately send your system off into the weeds...
Lastly, I had a 5930k that started life in my R5E just, um, defecate the bed... It was never OC'd as It was not stable from day=1. I moved it around trying to eliminate everything (memory, psu, mb, etc...) and finally I gave up and RMA'd the CPU through Intel. They are pretty easy to work with.
That same R5E is still going strong, now with a 6950x in it, but its also hosted 5960x's and the replacement 5930k from Intel without issue.
