There is an example for using Intel MKL to perform FFT in 3D, as posted here. It would be absolutely helpful to know how MKL FFT libraries can be used to compute 3D convolution (in a serial manner) for values of type float or double. Even a simple example would help to a great extend. Thanks in advance.
You can find bundle of examples for 1 and 2-dimension linear convolution operations on single / double precision data into
<mkl_install>\examples\vslc\source directory ( and the same for fortran )