[libcxx-commits] [libcxx] [RFC] Offloading C++ standard parallel algorithms to GPUs using OpenMP (PR #66465)

[Anton Rydahl via libcxx-commits]

AntonRydahl wrote:

Thanks, Louis! I agree we should add a new GPU backend rather than replace the SIMD backend. That will also make it possible to add offloading support gradually.

I wanted to verify that it was possible to reuse the implementations in `libcxx` and if offloading gave reasonable performance.

https://github.com/llvm/llvm-project/pull/66465