Comment by neonsunset

Take a look at ILGPU. It's very nice and has been around for a long time! (just no one knows about it, sadly)

Short example: https://github.com/m4rs-mt/ILGPU/blob/master/Samples/SimpleM...

Supports even advanced bits like inline PTX assembly: https://github.com/m4rs-mt/ILGPU/blob/master/Samples/InlineP...