The example computes the addtion of two vectors stored in array a and b and put the result in array out. To get things into action, we will looks at vector addition.įollowing is an example of vector addition implemented in C (. The Makefile assumes source code for the project is broken up into two groups, headers (.hpp) and implementation files (.cpp). The CUDA hello world example does nothing, and even if the program is compiled, nothing will show up on screen. This is a very simple C++ Makefile example and associated template, that can be used to get small to medium sized C++ projects up and running quickly and easily. Learn more about bidirectional Unicode characters. To review, open the file in an editor that reveals hidden Unicode characters. You might see following warning when compiling a CUDA program using above command nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning). This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. NVIDIA provides a CUDA compiler called nvcc in the CUDA toolkit to compile CUDA code, typically stored in a file with extension.

Compiling CUDA programsĬompiling a CUDA program is similar to C program. We will discuss about the parameter (1,1) later in this tutorial 02. In CUDA terminology, this is called ' kernel launch'. When a kernel is called, its execution configuration is provided through > syntax, e.g. the main() function in the example, and is also known as ' kernels'.

Such function can be called through host code, e.g. The _global_ specifier indicates a function that runs on device (GPU).

The major difference between C and CUDA implementation is _global_ specifier and > syntax.

Cuda c makefile example