We've written a cuda to amdgpu compiler. It should be possible to persuade someone like Modal that this is wonderful. So far people largely decline to believe t

We've written a cuda to amdgpu compiler. It should be possible to persuade someone like Modal that this is wonderful. So far people largely decline to believe that it can be done at all :) If you've written vaguely reasonable cuda, feed it to our nvcc binary and away you go. It's more annoying with dependencies, e.g. first one builds cutlass with our "nvcc" and so forth. Just occurred to me while writing this that we could start hosting builds of popular libraries ourselves. 3am here so I'll leave you with our somewhat confusing https://spectralcompute.com/ https://spectralcompute.com/ and the more useful direct link to a tarball https://docs.scale-lang.com/stable/manual/how-to-install/#__tabbed_1_3 https://docs.scale-lang.com/stable/manual/how-to-install/#__... I've only just started thinking about how to make it easier to use the stack. Thoughts / feedback very welcome on that!