IRON: High-level Python API for programming AMD Ryzen AI NPUs.
Provides the primary abstractions for describing NPU designs:
- :class:`Buffer` -- named memory region shared between Workers and the Runtime
- :class:`ObjectFifo` -- synchronized dataflow channel between program components
- :class:`CascadeFlow` -- directed cascade stream connection between two Workers
- :class:`Worker` -- a task running on an AIE compute core
- :class:`Runtime` -- host-side orchestration of data movement and worker execution
- :class:`Program` -- top-level container that compiles a design to MLIR
- :class:`Kernel` / :class:`ExternalFunction` -- pre-compiled or C++ kernel functions
- :class:`WorkerRuntimeBarrier` -- synchronization primitive between workers and runtime
- Tensor utilities (:func:`arange`, :func:`zeros`, :func:`ones`, etc.) for NPU-accessible buffers
- dtype helpers (:func:`str_to_dtype`, :func:`dtype_to_str`)
- :class:`CompilableDesign` / :func:`compileconfig` -- bundle a generator with compile-time config
- :class:`CallableDesign` / :func:`jit` -- JIT-compile and run on the NPU (Triton-style)
- :class:`CompileTime` / :class:`In` / :class:`Out` / :class:`InOut` -- type-annotation markers
- :func:`get_compile_arg` -- dynamic compile-time injection (advanced)
.. note::
IRON designs decorated with ``@iron.jit`` execute inside an implicit
MLIR context (thread-local ``Location`` / ``InsertionPoint``). See
``programming_guide/implicit_mlir_context.md`` for the mental model
and how to read "no active location" errors.
.. note::
``Kernel.__call__`` / ``ExternalFunction.__call__`` silently insert a
``memref.collapse_shape`` when an N-D contiguous memref arg is fed
into a 1-D kernel signature with the same element count and dtype
(typical case: iron L1 ObjectFifo elements are 2-D, ``kernels.X``
helpers declare flat 1-D args). Real shape / dtype mismatches still
surface at MLIR verification time.
