tilelang.tileop.gemm_blockscaled
================================

.. py:module:: tilelang.tileop.gemm_blockscaled

.. autoapi-nested-parse::

   Common block-scaled GEMM tile op.



Submodules
----------

.. toctree::
   :maxdepth: 1

   /autoapi/tilelang/tileop/gemm_blockscaled/gemm_blockscaled_base/index


Classes
-------

.. autoapisummary::

   tilelang.tileop.gemm_blockscaled.GemmBlockScaled


Package Contents
----------------

.. py:class:: GemmBlockScaled

   Bases: :py:obj:`tilelang.tileop.gemm.Gemm`


   Block-scaled GEMM tile op: ``C (+)= (A * SFA) @ (B * SFB)``.

   A ``GemmNode`` subclass on the C++ side, so it shares the dense GEMM's
   operand layouts, warp partition and scheduling, and is lowered through the
   same ``tl.gemm.infer_layout`` / ``tl.gemm.lower`` entry points, which
   dispatch on this Python class. The extra FFI fields are ``sfaRegion``,
   ``sfbRegion`` and ``sfKStart``. The C++ selector returns a block-scaled
   instruction key (``cuda.tcgen05.blockscaled``, ``cuda.mma.blockscaled``)
   that the backend registry maps to an implementation class built on
   ``GemmBlockScaledMixin``, so dense implementation classes never see the
   scale factors.


   .. py:property:: SFARegion


   .. py:property:: SFBRegion


   .. py:property:: sf_k_start


   .. py:property:: is_blockscaled
      :type: bool