首页 (home) Cuda Programming
文章 (post)
取消 (cancel)

Cuda Programming

##Core Concepts:

  • Host: CPU and its memory
  • Device: GPU and its memory

##Procedures

  • Allocate the memory of host and initialize the data
  • Allocate the memory of device and copy data from host to device
  • Use the kernal functions of CUDA and complete computations on device
  • Copy computed results from device to host
  • Free the memory on both host and device

** using the CUDA kernal functions to execute parallel computing

global declares kernal «<grid, block»> defines the number of executing threads Each thread has a thread ID (threadIdx)


本文由作者按照 CC BY 4.0 进行授权
热门标签 (Trending Tags)

热门标签 (Trending Tags)