0

I have not achieved stage when I will parallel my program with CUDA, but I use C++ instead C and I am worried if my program would not compiled with CUDA!

Somebody can find proof for this deal.

Mat
  • 202,337
  • 40
  • 393
  • 406
Moscow Boy
  • 1,591
  • 3
  • 14
  • 15

1 Answers1

5

This is a common question about CUDA. You can compile device code with nvcc to object file then link in using g++.

For more information see this post, for example.

Community
  • 1
  • 1
geek
  • 1,809
  • 1
  • 12
  • 12