Computing a number BATCH of one-dimensional DFTs of size NX using cuFFT will typically look like this:

在CUDA上实现DFT算法大概会是这个样子的流程：

#define NX 256
#define BATCH 10
#define RANK 1
...
{
    cufftHandle plan;
    cufftComplex *data;
    ...
    cudaMalloc((void**)&data, sizeof(cufftComplex)*NX*BATCH);
    cufftPlanMany(&plan, RANK, NX, &iembed, istride, idist, 
        &oembed, ostride, odist, CUFFT_C2C, BATCH);
    ...
    cufftExecC2C(plan, data, data, CUFFT_FORWARD);
    cudaDeviceSynchronize();
    ...
    cufftDestroy(plan);
    cudaFree(data);
}

2.1访问cuFFT
   cuFFT和cuFFTW库是免费共享的。它们已经完成了只要我们引用就行，cuFFT can be downloaded from http://developer.nvidia.com/cufft. By selecting Download CUDA Production Release users are all able to install the package containing the CUDA Toolkit, SDK code samples and development drivers. The CUDA Toolkit contains cuFFT and the samples include simplecuFFT。可以在。。。下载安装，simplecuFFT是例程。2.2傅里叶变换类型

• cufftPlan1D() / cufftPlan2D() / cufftPlan3D() - Create a simple plan for a 1D/2D/3D transform respectively.
• 创建一个一维/二维/三维傅里叶变换为的简单计划。
• cufftPlanMany() - Creates a plan supporting batched input and strided data layouts.
• 创建一个计划，支持批量输入和分析数据布局？
• cufftXtMakePlanMany() - Creates a plan supporting batched input and strided data layouts for any supported precision.
• 创建一个计划，支持批量输入和数据布局的任何支持的精度分析
  在以上的三个几哈生成函数中，cufftPlanMany() 

2.3 傅里叶变换类型
    

• cufftexecc2c cufftexecz2z（）/（）-单/双精度复杂变换。
• cufftexecr2c cufftexecd2z（）/（）真正复杂的正变换单/双精度。
• cufftexecc2r cufftexecz2d（）/（）复杂的单/双精度实反变换。

Each of those functions demands different input data layout (see Data Layout for details).

Functions cufftXtExec() and cufftXtExecDescriptor() can perform transforms on any of the supported types.

2.4数据布局

2.5多维过滤和转换
2.6提前进行数据布局
   

2.7. 流傅里叶变换

   每个cuFFT变换可能都与CUDA流有关，

   2.8多GPU同时运行的cuFFT变换

   CUFFT支持使用多达八个GPU和一个CPU连接进行傅里叶变换的计算 。我们定义了一个函数API使用户能够修改咸鱼代码或者写新的代码。

• cufftCreate() - create an empty plan, as in the single GPU case
• 创建一个空的计划在单GPU的情况下
• cufftXtSetGPUs() - define which GPUs are to be used
• 指定哪个GPU被使用
• Optional: cufftEstimate{1d,2d,3d,Many}() - estimate the sizes of the work areas required. These are the same functions used in the single GPU case although the definition of the argument workSize reflects the number of GPUs used.
• 估计所需的工作区域的大小。这些是相同的函数中使用单一GPU案例虽然论证workSize的定义反映了GPU的数量。
• cufftMakePlan{1d,2d,3d,Many}() - create the plan. These are the same functions used in the single GPU case although the definition of the argument workSize reflects the number of GPUs used.
• Optional: cufftGetSize{1d,2d,3d,Many}() - refined estimate of the sizes of the work areas required. These are the same functions used in the single GPU case although the definition of the argument workSize reflects the number of GPUs used.
• Optional: cufftGetSize() - check workspace size. This is the same function used in the single GPU case although the definition of the argument workSize reflects the number of GPUs used.
• Optional: cufftXtSetWorkArea() - do your own workspace allocation.
• cufftXtMalloc() - allocate descriptor and data on the GPUs
• GPU分配内存
• cufftXtMemcpy() - copy data to the GPUs
• 数据拷贝至GPU
• cufftXtExecDescriptorC2C()/cufftXtExecDescriptorZ2Z() - execute the plan
• cufftXtMemcpy() - copy data from the GPUs
• 从GPU拷贝数据
• cufftXtFree() - free any memory allocated with cufftXtMalloc()
• 释放和分配内存
• cufftDestroy() - free cuFFT plan resources

      释放cuFFT计划

大毕设-CUDA-cuFFT库