It sounds like you've checked the obvious connectors and sockets. Have you tried running with just one DIMM to see if it still crashes? Have you contacted Gigabyte tech support about the issue? Have you checked the CPU core temp under heavy load like gaming or P95 stress testing? Are you manually setting the RAM timings in the BIOS to match whatever the RAM maker suggests?
A couple settings that seem to help many people with the 8-core FX processors are:
increase the RAM voltage ~.05V above the specified voltage
HT = 1.25v
HT frequency = 2600 MHz.
If you don't need the power saving options you can manually set the CPU vcore and disable C&Q, C6, C1E, Application Power Management, CPU internal throttle (only if the CPU cooling is excellent and fan speed appropriate). Gigabyte may use different names for some of these BIOS settings.
Adjust the Load Line Conditioning (LLC) or similarly named control so it keeps the vcore voltage as close as possible to the default voltage (as shown in CPU-Z), when the CPU is under heavy load. You can also increase the CPU vcore if need be as "auto" mode does not always work even though it should.
It's also possible Gigabyte has a BIOS issue with that board as the XFA models are pretty new and they changed the VRM circuit because the prior models would over-heat the VRM circuit with 8-core FX CPUs.
While it's remotely possible, I highly doubt the CPU has gone bad. It's usually a BIOS setting that isn't 100% right for the CPU/mobo/RAM and it causes the problem. Corrupted GPU drivers are also a common problem and sometimes require extensive cleaning and re-installation.
Building a reliable PC involves more than just assembling the parts. You need to be able to configure all of the BIOS settings appropriately. This can be quite involved and frustrating as it can require a lot of trial and error with stress testing. It is however often the only means to get a 100% reliable PC.