tilelang.cuda.language.reduce_op
================================

.. py:module:: tilelang.cuda.language.reduce_op

.. autoapi-nested-parse::

   CUDA dialect of the reduction operators: the common ops plus CUDA knobs.



Functions
---------

.. autoapisummary::

   tilelang.cuda.language.reduce_op.reduce_max
   tilelang.cuda.language.reduce_op.reduce_min
   tilelang.cuda.language.reduce_op.reduce_absmax


Module Contents
---------------

.. py:function:: reduce_max(buffer, out, dim = -1, clear = True, batch = 1, nan_propagate = False, annotations = None)

   Perform reduce max, with the CUDA NaN-propagation knob.

   Same semantics as the common :func:`tilelang.language.reduce_op.reduce_max`.
   ``nan_propagate`` is meaningful for float16/bfloat16 only: when True the
   reduction lowers to ``__hmax_nan`` so NaNs propagate; when False (default)
   ``__hmax`` returns the non-NaN operand. Targets without these intrinsics
   reject the annotation at compile time.


.. py:function:: reduce_min(buffer, out, dim = -1, clear = True, batch = 1, nan_propagate = False, annotations = None)

   Perform reduce min, with the CUDA NaN-propagation knob.

   See :func:`reduce_max`; this lowers to ``__hmin_nan``/``__hmin``.


.. py:function:: reduce_absmax(buffer, out, dim = -1, clear = True, batch = 1, nan_propagate = False, annotations = None)

   Perform reduce absolute max, with the CUDA NaN-propagation knob.

   See :func:`reduce_max`.