首页| JavaScript| HTML/CSS| Matlab| PHP| Python| Java| C/C++/VC++| C#| ASP| 其他|
购买积分 购买会员 激活码充值

您现在的位置是:虫虫源码 > C/C++/VC++ > 图书馆,简化了CUDA程序的开发

图书馆,简化了CUDA程序的开发

  • 资源大小:108.30 kB
  • 上传时间:2021-06-30
  • 下载次数:0次
  • 浏览次数:0次
  • 资源积分:1积分
  • 标      签: cuda Academic CPlusPlus GPU

资 源 简 介

General Description DevAr (DeviceArray) is a C++ template library that works as a wrapper for variables allocated in GPU memory. The wrapper dramatically decreases the volume of code and makes program much reliable. Main Features The features of the library are: * Built-in most typical operations with arrays in device memory (fill, clear, resize, etc.), which makes code shorter and more readable. * Checking if operation with CUDA array were successful. If any appropriate function returned anything but cudaSuccess, the exception is thrown. The feature makes programs based on the DevAr template more reliable. How to use it? The example of using DevAr and documentation are available on Downloads page. Example ``` #include "DevAr.h" ... //other headers void kernel (float* data){ ... //do something } int main(){ DevAr devF(256); //array in device memory float* f = new float[256];
VIP VIP
0.344209s