Hands-On Deep Learning with Go

Installing CUDA support for Gorgonia

To make use of CUDA, you need a computer with a GPU made by NVIDIA. Unfortunately, setting up CUDA to work with Gorgonia is a slightly more involved process, as it involves setting up a C compiler environment to work with Go, as well as a C compiler environment that works with CUDA. NVIDIA has kindly ensured that its compiler works with the common toolchain for each platform: Visual Studio on Windows, Clang-LLVM on macOS, and GCC on Linux.

Installing CUDA and ensuring that everything works correctly requires a fair bit of work. We'll look at doing this for Windows and Linux. As Apple has not made a computer featuring an NVIDIA GPU for several years (as of writing this), we will not cover how to do this ...