Kokkos Node API and Local Linear Algebra Kernels Version of the Day
KokkosExamples::EmptySparseKernel< Scalar, Node >::bind_scalar< T > Struct Template Reference

Rebind struct, for specifying type information for a different scalar. More...

#include <KokkosExamples_EmptySparseKernelClass.hpp>

List of all members.


Detailed Description

template<class Scalar, class Node>
template<class T>
struct KokkosExamples::EmptySparseKernel< Scalar, Node >::bind_scalar< T >

Rebind struct, for specifying type information for a different scalar.

This typedef lets you tell us where to find sparse kernels for sparse matrices with entries of scalar type T. T may be different than scalar_type.

One point of this typedef is that sometimes you may have noptimized kernels for some scalar types T (such as float or double), but not for other types (such as extended-precision types). Some scalar types T (especially those requiring dynamic memory allocation) might not work correctly or efficiently on certain Kokkos Node types (especially GPU Node types). This typedef lets you provide a "fall-back" implementation of sparse kernels.

Definition at line 153 of file KokkosExamples_EmptySparseKernelClass.hpp.


The documentation for this struct was generated from the following file:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends