Add suport for int8 matmul in afcuda #1656

I see these new features in cuda 8:
"Native FP16 and INT8 computation for deep learning and other workloads;" :
https://devblogs.nvidia.com/parallelforall/cuda-8-features-revealed/
This feature request is to be able to use it through AF at least for matmul and arithmetic (+ - * /)
Thanks
WT.