Yeah, I just read this posted by Genxun on the DAIN Discord Server. Despite this, what pauliezmm said about CUDA optimisation is still true. Thanks