Dive into the general-purpose GPU programming
A comparison between OpenCL, CUDA, C++ libraries and compiler toolchains for heterogeneous computing.
The GPGPU ecosystem is rapidly developing thanks to the AI boom. To navigate this complex set of tools, we will compare the languages, libraries and compilers for GPGPU, learn how APIs for computing are different their graphics counterparts like OpenGL and Vulkan and look into caveats of writing concurrent programs in OpenCL & CUDA.
The less common technologies we are going to cover are SyCL, Halide and BLAS libraries.
To wrap things up, we will prepare four different combinations of tools and compare them against each other to make it simpler for you to pick the right set for your next project!