tilelang.webgpu.language.kernel
===============================

.. py:module:: tilelang.webgpu.language.kernel

.. autoapi-nested-parse::

   WebGPU dialect of ``T.Kernel``: the common launch plus WebGPU launch annotations.



Functions
---------

.. autoapisummary::

   tilelang.webgpu.language.kernel.Kernel


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

.. py:function:: Kernel(*blocks, threads = None)

   Construct a kernel launch frame for WebGPU: a grid of workgroups.

   Code inside the launch operates at the workgroup level: ``T.Parallel``,
   ``T.copy`` and friends are mapped onto invocations by the compiler.
   ``T.get_thread_binding()`` exposes the invocation index for thread-level
   code. ``threads`` is recorded at trace time and materialized by the WebGPU
   pipeline once the target is known.

   :param \*blocks: Grid extent along each axis (1-3 dimensions). The launch yields one
                    workgroup index per axis.
   :type \*blocks: int | PrimExpr
   :param threads: Invocations per workgroup: a count or up to three per-dimension
                   extents. Defaults to 128 when omitted.
   :type threads: int | list[int] | tuple[int, ...], optional

   .. rubric:: Examples

   .. code-block:: python

       with T.Kernel(T.ceildiv(N, 128), threads=128) as bx:
           ...