These programming examples are provided so that application programmers can learn how to leverage the IRON design flow with mlir-aie python bindings, and the mlir-aie intermediate representation directly to build applications targeting AI Engines.
Each IRON example has one or more implementations:
<example_name>.py- These designs are generally written using a higher-level version of IRON<example_name>_alt.py- These designs are generally written using a lower-level verion of IRON
They are organized into the following directories:
Basic building blocks to understand the NPU architecture and first steps towards building applications for the NPU in Ryzen™ AI.
Machine learning building blocks, design components, and reference designs for the NPU in Ryzen™ AI.
Computer vision processing pipeline designs for the NPU in Ryzen™ AI.
MLIR-based reference designs expressed in the MLIR-AIE dialect intermediate representation.
Utilty functions leveraged in the programming examples.