We report on our efforts to implement overlap fermions on NVIDIA GPUs using CUDA, commenting on the algorithms used, implemetation details, and the performance of our code.
生成讲解博客的中文版本
生成AI智能校验的中文版本