Tutorial by Examples

To get started programming with CUDA, download and install the CUDA Toolkit and developer driver. The toolkit includes nvcc, the NVIDIA CUDA Compiler, and other software necessary to develop CUDA applications. The driver ensures that GPU programs run correctly on CUDA-capable hardware, which you'll ...

This example illustrates how to create a simple program that will sum two int arrays with CUDA. A CUDA program is heterogenous and consist of parts runs both on CPU and GPU. The main parts of a program that utilize CUDA are similar to CPU programs and consist of Memory allocation for data that ...

This simple CUDA program demonstrates how to write a function that will execute on the GPU (aka "device"). The CPU, or "host", creates CUDA threads by calling special functions called "kernels". CUDA programs are C++ programs with additional syntax. To see how it works...

The NVIDIA installation guide ends with running the sample programs to verify your installation of the CUDA Toolkit, but doesn't explicitly state how. First check all the prerequisites. Check the default CUDA directory for the sample programs. If it is not present, it can be downloaded from the offi...