Cyclops Tensor Framework
parallel arithmetic on multidimensional arrays
|
Go to the source code of this file.
Functions | |
template<typename dtype > | |
Tensor< dtype > | ao_mo_transf_naive (Tensor< dtype > &U, Matrix< dtype > &C) |
naive implementation of AO-MO orbital transformation LIMITATIONS: (1) does not exploit output (syrk-like) symmetry in 3rd and 4th products. that may increase flop cost by 1.5X (2) multiple buffers and no auxiliary blocking to minimize memory usage More... | |
template<typename dtype > | |
Tensor< dtype > | ao_mo_transf_slice (Tensor< dtype > &U, Matrix< dtype > &C) |
AO-MO orbital transformation applied to a slice. More... | |
void | test_ao_mo_transf (int n, int m, int k, MPI_Comm cm=MPI_COMM_WORLD, bool flt_test=true, bool ns_test=true) |
template<typename dtype > | |
void | bench_ao_mo_transf (int n, int m, int k) |
char * | getCmdOption (char **begin, char **end, const std::string &option) |
int | main (int argc, char **argv) |