|
numerics
|
GPU (CUDA) backend – vector operations. More...
#include "core/vector.hpp"#include "core/parallel/cuda_ops.hpp"#include "../seq/impl.hpp"#include <cmath>Go to the source code of this file.
Namespaces | |
| namespace | num |
| namespace | num::backends |
| namespace | num::backends::gpu |
Functions | |
| void | num::backends::gpu::scale (Vector &v, real alpha) |
| void | num::backends::gpu::axpy (real alpha, const Vector &x, Vector &y) |
| real | num::backends::gpu::dot (const Vector &x, const Vector &y) |
| real | num::backends::gpu::norm (const Vector &x) |
GPU (CUDA) backend – vector operations.
Thin shims that forward to cuda:: kernels (cuda_ops.hpp). Falls back to sequential when NUMERICS_HAS_CUDA is not defined.
Definition in file vector.cpp.