Hadi
Hadi

Reputation: 307

A CUDA parallel program help needed

I am new to CUDA and parallel programming. I want to do a double parallel part in CUDA can someone help me out here. Sorry for posting here but i am a little confused and could not find any nice tutorial or help for it. I want to make a function such that it works parallel in CUDA and then i want to make another function of CUDA that is called in parallel from that function is it possible in kernel function of CUDA ?

EDIT

for example i have two kernel functions my_prog and my prog_1 i am calling them like this

main_function
{
    my_prog<<core,threads>>();
}

my_prog()
{
   //here i want to call my_prog_1 in parallel
   my_prog_1<<core,threads>>();
}

Upvotes: 0

Views: 209

Answers (1)

Robert Crovella
Robert Crovella

Reputation: 152174

If you want to learn how to write CUDA programs, I would suggest taking some of the tutorials that are available.

NVIDIA has a webinar page with a variety of tutorials.

For starters, these two webinars will give you a basic intro to CUDA with enough knowledge to be able to write CUDA programs that run fast:

GPU Computing using CUDA C – An Introduction (2010) An introduction to the basics of GPU computing using CUDA C. Concepts will be illustrated with walkthroughs of code samples. No prior GPU Computing experience required

GPU Computing using CUDA C – Advanced 1 (2010) First level optimization techniques such as global memory optimization, and processor utilization. Concepts will be illustrated using real code examples

It would be 2 hours well spent if you want to learn how to write CUDA programs.

EDIT: What you're describing here is CUDA Dynamic Parallelism. It requires a device with CUDA compute capability of 3.5 (today).

You can find a complete user's guide for it here

Upvotes: 2

Related Questions