diff options
| author | Julian Taylor <juliantaylor108@gmail.com> | 2016-09-25 17:16:45 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2016-09-25 17:16:45 +0200 |
| commit | 0887da9386e812d2769744afd6c58a927d6cbb45 (patch) | |
| tree | 65f07b1cf5c4451be4e324e34fe8fd907578b07c /doc | |
| parent | a17e905faa20fcdd1b3f039bceef6efa19671bb3 (diff) | |
| parent | ae32e780d85e8e42402bc1fee2bbdb3ee90426c4 (diff) | |
| download | numpy-0887da9386e812d2769744afd6c58a927d6cbb45.tar.gz | |
Merge pull request #7980 from juliantaylor/avx-runtime
ENH: Add ability to runtime select ufunc loops, add AVX2 integer loops
Diffstat (limited to 'doc')
| -rw-r--r-- | doc/release/1.12.0-notes.rst | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/doc/release/1.12.0-notes.rst b/doc/release/1.12.0-notes.rst index 5ae84a160..8f32f8da9 100644 --- a/doc/release/1.12.0-notes.rst +++ b/doc/release/1.12.0-notes.rst @@ -311,6 +311,10 @@ bugs where such dtypes were not handled properly. In particular, changed ``ndarray.__new__`` to not implicitly convert ``dtype('S0')`` to ``dtype('S1')`` (and likewise for unicode) when creating new arrays. +Integer ufuncs vectorized with AVX2 +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +If the cpu supports it at runtime the basic integer ufuncs now use AVX2 +instructions. This feature is currently only available when compiled with GCC. Changes ======= |