My Project
|
This class implements a cusparse-based ilu0-bicgstab solver on GPU. More...
#include <cusparseSolverBackend.hpp>
Public Member Functions | |
cusparseSolverBackend (int linear_solver_verbosity, int maxit, Scalar tolerance, unsigned int deviceID) | |
Construct a cusparseSolver. | |
~cusparseSolverBackend () | |
Destroy a cusparseSolver, and free memory. | |
SolverStatus | solve_system (std::shared_ptr< BlockedMatrix< Scalar > > matrix, Scalar *b, std::shared_ptr< BlockedMatrix< Scalar > > jacMatrix, WellContributions< Scalar > &wellContribs, BdaResult &res) override |
Solve linear system, A*x = b, matrix A must be in blocked-CSR format. | |
void | get_result (Scalar *x) override |
Get resulting vector x after linear solve, also includes post processing if necessary. | |
![]() | |
BdaSolver (int linear_solver_verbosity, int max_it, Scalar tolerance_) | |
Construct a BdaSolver. | |
BdaSolver (int linear_solver_verbosity, int max_it, Scalar tolerance_, unsigned int deviceID_) | |
BdaSolver (int linear_solver_verbosity, int max_it, double tolerance_, unsigned int platformID_, unsigned int deviceID_) | |
virtual | ~BdaSolver ()=default |
Define virtual destructor, so that the derivedclass destructor will be called. | |
Additional Inherited Members | |
![]() | |
int | verbosity = 0 |
int | maxit = 200 |
Scalar | tolerance = 1e-2 |
int | N |
int | Nb |
int | nnz |
int | nnzb |
unsigned int | platformID = 0 |
unsigned int | deviceID = 0 |
bool | initialized = false |
This class implements a cusparse-based ilu0-bicgstab solver on GPU.
Opm::Accelerator::cusparseSolverBackend< Scalar, block_size >::cusparseSolverBackend | ( | int | linear_solver_verbosity, |
int | maxit, | ||
Scalar | tolerance, | ||
unsigned int | deviceID | ||
) |
Construct a cusparseSolver.
[in] | linear_solver_verbosity | verbosity of cusparseSolver |
[in] | maxit | maximum number of iterations for cusparseSolver |
[in] | tolerance | required relative tolerance for cusparseSolver |
[in] | deviceID | the device to be used |
|
overridevirtual |
Get resulting vector x after linear solve, also includes post processing if necessary.
[in,out] | x | resulting x vector, caller must guarantee that x points to a valid array |
Implements Opm::Accelerator::BdaSolver< Scalar, block_size >.
|
overridevirtual |
Solve linear system, A*x = b, matrix A must be in blocked-CSR format.
[in] | matrix | matrix A |
[in] | b | input vector, contains N values |
[in] | jacMatrix | matrix for preconditioner |
[in] | wellContribs | contains all WellContributions, to apply them separately, instead of adding them to matrix A |
[in,out] | res | summary of solver result |
Implements Opm::Accelerator::BdaSolver< Scalar, block_size >.