when i want to use cufft in hpc_sdk,but the complier con’t find the cufft.h,so how can i finish it?
Add the flag “-cudalib=cufft” and the compiler will implicitly add the include directory where cufft.h is located. It will also implicitly add the CUFFT runtime library when the flag is used on the link line.
Without this flag, you need to add the path to the directory containing the header file. For example:
-I/opt/nv/Linux_x86_64/23.9/math_libs/12.3/include
Though you will need to adjust the path to match your installation and the particular CUDA version you’re using.