If your issue isn't fixed by now, is your GPU overclocked? This can also be caused by an unstable overclock on your GPU. Though i think you'd probably of already known that if that was what it was. Another thing is it could be getting entirely too hot and had hardware failure due to that. Its also worth checking to see if your GPU is fully seated into the motherboard. That may sound silly but i've encountered it before. I always use NVidia GPU's in my builds so i'm not sure if you have this but if you have a control panel for your GPU's driver settings, check that and see if anything looks odd. Although thats probably unlikely unless you made changes yourself. Using an older driver as suggested above may work for you as well, as sometimes AMD and Nvidia just release buggy drivers and you have to wait for the next release for them to fix it. Its rare to hear about a GPU driver crashing just due to stress unless its heat related... The last thing is it could just be hardware failure. HWMonitor is a program that can monitor the temperatures of your hardware. While using that, run a stress test and see what happens. Heaven Benchmark 4.0 is a good one. If your temps go above 85c, that is likely the culprit. If not, it doesn't necessarily mean bad hardware. I'd do some googling on it if it hasn't been fixed yet.

If your entire system crashes after the driver stops working it is likely either the GPU you got is damaged or is getting too hot and throttling isn't keeping it cool enough, so the system shuts down to protect itself. Rare with GPU's though, they usually just throttle. HWmonitor will tell you the GPU fan speeds aswell to make sure those are running. Thats all i got