|
kernel_enums.hpp
Go to the documentation of this file.
26 This file organizes a growing amount of different mixed precision options for the preconditioners.
__host__ __device__ constexpr bool storeDiagonalAsFloat(MatrixStorageMPScheme scheme) Definition: kernel_enums.hpp:56 bool isValidMatrixStorageMPScheme(int scheme) Definition: kernel_enums.hpp:96 __host__ __device__ constexpr bool usingMixedPrecision(MatrixStorageMPScheme scheme) Definition: kernel_enums.hpp:83 __host__ __device__ constexpr bool storeOffDiagonalAsFloat(MatrixStorageMPScheme scheme) Definition: kernel_enums.hpp:69 A small, fixed‑dimension MiniVector class backed by std::array that can be used in both host and CUDA... Definition: AmgxInterface.hpp:38 @ FLOAT_DIAG_FLOAT_OFFDIAG @ DOUBLE_DIAG_FLOAT_OFFDIAG @ DOUBLE_DIAG_DOUBLE_OFFDIAG MatrixStorageMPScheme makeMatrixStorageMPScheme(int scheme) Definition: kernel_enums.hpp:41 |