基于GPU的通用计算，利用显存的cache是提高算法效率的重要方面，该例子是在CUDA环境下如何读取纹理的程序。-based GPU generic terms, the use of video memory cache is to improve the efficiency of the algorithm, The example is the CUDA environment how to read texture procedures.

GPU 实现FFT，包括读取txt文件至内存中实现，实现结果后，将结果以txt文件的格式保存。-After the GPU to realize FFT, including memory read txt file to achieve, achieve results, the results will be saved in txt file format.

根据一个英文文本文件，不区分大小写的统计26个字母的出现频率。假设内存足够大，可以将文件一次性读入。(According to an English text file, the frequency of appearance of 26 letters is not case sensitive. Assuming that the memory is large enough, you can read the file one at a time.)