Highly Efficient FFT for Exascale: HeFFTe v2.3
heffte::rocm::cos_pre_pos_processor Struct Reference

Implementation of Cosine Transform pre-post processing methods using CUDA. More...

#include <heffte_backend_rocm.h>

Static Public Member Functions

template<typename precision >
static void pre_forward (hipStream_t, int length, precision const input[], precision fft_signal[])
 Pre-process in the forward transform.
 
template<typename precision >
static void post_forward (hipStream_t, int length, std::complex< precision > const fft_result[], precision result[])
 Post-process in the forward transform.
 
template<typename precision >
static void pre_backward (hipStream_t, int length, precision const input[], std::complex< precision > fft_signal[])
 Pre-process in the inverse transform.
 
template<typename precision >
static void post_backward (hipStream_t, int length, precision const fft_result[], precision result[])
 Post-process in the inverse transform.
 

Detailed Description

Implementation of Cosine Transform pre-post processing methods using CUDA.


The documentation for this struct was generated from the following file: