hello, I want to learn about that how the multi-gpu and multi-mpi matrix inversion is implemented from the article [https://arxiv.org/pdf/2506.03477]。 I have read the install document, which said that gpu implementation source code is switched by cpp, but I do not find the cpp file in the source code
hello, I want to learn about that how the multi-gpu and multi-mpi matrix inversion is implemented from the article [https://arxiv.org/pdf/2506.03477]。 I have read the install document, which said that gpu implementation source code is switched by cpp, but I do not find the cpp file in the source code