> But personally I'm excited to have something akin to a scripting language for a GPU. There's a few of these already, depending what you mean by scripting lan

> But personally I'm excited to have something akin to a scripting language for a GPU. There's a few of these already, depending what you mean by scripting language. You have Numba for Python [0], with official backing from nVidia, and turbo.js [1] for JS. Alea for F# might classify, too, thanks to its REPL, and how quick you can get off the ground. [2] [0] https://developer.nvidia.com/how-to-cuda-python https://developer.nvidia.com/how-to-cuda-python [1] https://turbo.js.org/ https://turbo.js.org/ [2] https://developer.nvidia.com/alea-gpu https://developer.nvidia.com/alea-gpu