Skip to content

mkl_blas::scal instead of loop in backend rng_rayleigh #1049

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Dec 30, 2021
Merged

mkl_blas::scal instead of loop in backend rng_rayleigh #1049

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
4690fe3
mkl_blas::scal instead of loop in backend rng_rayleigh
LukichevaPolina Nov 19, 2021
3185b12
Using dependent events
LukichevaPolina Nov 23, 2021
ff9ca09
Renameing event returned from mkl_blas::scal
LukichevaPolina Dec 7, 2021
d495379
Merge branch 'master' into random_rayleigh
Alexander-Makaryev Dec 29, 2021
5a21082
Merge branch 'master' of https://github.com/IntelPython/dpnp into ran…
LukichevaPolina Dec 30, 2021
1208118
Exclude data copy in DPNPC_ptr_adapter
LukichevaPolina Dec 30, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 5 additions & 12 deletions dpnp/backend/kernels/dpnp_krnl_random.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -715,22 +715,15 @@ void dpnp_rng_rayleigh_c(void* result, const _DataType scale, const size_t size)
const _DataType a = 0.0;
const _DataType beta = 2.0;

DPNPC_ptr_adapter<_DataType> result1_ptr(result, size, true, true);
DPNPC_ptr_adapter<_DataType> result1_ptr(result, size);
_DataType* result1 = result1_ptr.get_ptr();

mkl_rng::exponential<_DataType> distribution(a, beta);

auto event_out = mkl_rng::generate(distribution, DPNP_RNG_ENGINE, size, result1);
event_out.wait();
event_out = mkl_vm::sqrt(DPNP_QUEUE, size, result1, result1, no_deps, mkl_vm::mode::ha);
event_out.wait();
// with MKL
// event_out = mkl_blas::axpy(DPNP_QUEUE, size, scale, result1, 1, result1, 1);
// event_out.wait();
for (size_t i = 0; i < size; i++)
{
result1[i] *= scale;
}
auto exponential_rng_event = mkl_rng::generate(distribution, DPNP_RNG_ENGINE, size, result1);
auto sqrt_event = mkl_vm::sqrt(DPNP_QUEUE, size, result1, result1, {exponential_rng_event}, mkl_vm::mode::ha);
auto scal_event = mkl_blas::scal(DPNP_QUEUE, size, scale, result1, 1, {sqrt_event});
scal_event.wait();
}

template <typename _DataType>
Expand Down