2018-07-27

Which PTX and binary code gets embedded in a CUDA C application is controlled by the -arch and -code compiler options or the -gencode compiler option as detailed in the nvcc user manual. For example,


nvcc x.cu

        -gencode arch=compute_35,code=sm_35

        -gencode arch=compute_50,code=sm_50

        -gencode arch=compute_60,code=\'compute_60,sm_60\'


embeds binary code compatible with compute capability 3.5 and 5.0 (first and second -gencode options) and PTX and binary code compatible with compute capability 6.0 (third -gencode option).

你可能感兴趣的:(2018-07-27)