EasyOpenCL

Features:

• A Visual Studio example project with guide on how to specify the include and library folders
• CMake support for Linux and Mac
• No exposed OpenCL or manual memory management calls, just the C++ STL.
• Parametrized kernel types, choose from processing floats, ints or chars on your GPU.
• Variable amount of input vectors
• Human readable OpenCL errors for easy debugging of your program.

Overview: it's this easy!

std::vector<int> input_1 (5,5);
std::vector<int> input_2 { 1, 2, 3, 4, 5 };

try {
    EasyOpenCL<int> framework (SHOW_DEBUG);

    framework.loadKernel("arithmetic.cl");
    framework.addInputBuffer(0, input_1);
    framework.addInputBuffer(1, input_2);
    framework.addOutputBuffer(2);
	
    framework.runKernel();
	
    framework.showOutputBuffer();
}
catch (std::exception& e) {
    std::cerr << "Error: " << e.what() << std::endl;
}

Getting started

Windows

For a more elaborate guide, follow example/VC++2013/README.md

• Clone or download this repository
• Open the example project solution (example/VC++2013/EasyOpenCL.sln) in Visual Studio
• Specify the proper include and library directories using the guide in the README.md
• Build and run the example using Ctrl-F5

Linux and Mac

git clone https://github.com/Gladdy/EasyOpenCL.git
cd EasyOpenCL
mkdir build && cd build
cmake ..
make 
./test

Thanks to:

• Dhruba Bandopadhyay: OpenCL Cookbook: Listing all devices and their critical attributes
• elhigu: cmake-findopencl
• Tom Scogland: OpenCL error checking