benchmark_DiracMatrixComputeCUDA.cpp File Reference

This implements micro benchmarking on DiracMatrixComputeCUDA.hpp Currently it also benchmarks the same size matrix's and batch sizes using the Legacy DiracMatrix serially. More...

Include dependency graph for benchmark_DiracMatrixComputeCUDA.cpp:

Go to the source code of this file.

Classes
struct	DiracComputeBenchmarkParameters

Namespaces
	qmcplusplus
	helper functions for EinsplineSetBuilder

Typedefs
template<typename T >
using	OffloadPinnedMatrix = Matrix< T, PinnedDualAllocator< T > >
template<typename T >
using	OffloadPinnedVector = Vector< T, PinnedDualAllocator< T > >

Functions
std::ostream &	operator<< (std::ostream &out, const DiracComputeBenchmarkParameters &dcbmp)
	TEST_CASE ("DiracMatrixComputeCUDA_large_determinants_benchmark_legacy_1024_4", "[wavefunction][fermion][.benchmark]")
	This and other [.benchmark] benchmarks only run if "[benchmark]" is explicitly passed as tag to test. More...
	TEST_CASE ("benchmark_DiracMatrixComputeCUDA_vs_legacy_256_10", "[wavefunction][fermion][benchmark]")
	This test will run by default. More...
	TEST_CASE ("benchmark_DiracMatrixComputeCUDASingle_vs_legacy_256_10", "[wavefunction][fermion][.benchmark]")
	Only runs if [benchmark] tag is passed. More...
	TEST_CASE ("benchmark_DiracMatrixComputeCUDASingle_vs_legacy_1024_4", "[wavefunction][fermion][.benchmark]")
	Only runs if [benchmark] tag is passed. More...

Detailed Description

This implements micro benchmarking on DiracMatrixComputeCUDA.hpp Currently it also benchmarks the same size matrix's and batch sizes using the Legacy DiracMatrix serially.

Definition in file benchmark_DiracMatrixComputeCUDA.cpp.