Cyclops Tensor Framework
parallel arithmetic on multidimensional arrays
|
Performs recursive parallel matrix multiplication using the slice interface to extract blocks. More...
Functions | |
void | recursive_matmul (int n, int m, int k, Tensor<> &A, Tensor<> &B, Tensor<> &C) |
int | test_recursive_matmul (int n, int m, int k, World &dw) |
char * | getCmdOption (char **begin, char **end, const std::string &option) |
int | main (int argc, char **argv) |
Performs recursive parallel matrix multiplication using the slice interface to extract blocks.
char* getCmdOption | ( | char ** | begin, |
char ** | end, | ||
const std::string & | option | ||
) |
Definition at line 128 of file recursive_matmul.cxx.
Referenced by main().
int main | ( | int | argc, |
char ** | argv | ||
) |
Definition at line 138 of file recursive_matmul.cxx.
References getCmdOption(), ctf.core::np(), ctf.core::rank(), and test_recursive_matmul().
Definition at line 12 of file recursive_matmul.cxx.
References CTF::World::comm, NS, ctf.core::rank(), CTF::Tensor< dtype >::slice(), and CTF_int::tensor::wrld.
Referenced by test_recursive_matmul().
int test_recursive_matmul | ( | int | n, |
int | m, | ||
int | k, | ||
World & | dw | ||
) |
Definition at line 75 of file recursive_matmul.cxx.
References CTF::World::comm, CTF::Tensor< dtype >::get_local_data(), CTF::Tensor< dtype >::norm2(), ctf.core::np(), NS, ctf.core::rank(), recursive_matmul(), and CTF::Tensor< dtype >::write().
Referenced by main().