Hi ,
Downloaded PFAC code and trying to compile code on machine which contains CUDA Tool Kit version 8.0 (latest). But facing following build error message,
nvcc fatal : Value 'sm_13' is not defined for option 'gpu-architecture'
Current CUDA tool kit version 8.0 supports arch 'sm_20' onwards. PFAC document recommends to use only CUDA version 3.2 . To upgrade PFAC to CUDA 8.0 requires any major logic changes ?
Or only "Makefile" changes enough to compile the same ?
Thanks
Mohanraj