aie4ml is an end-to-end compiler that generates optimized AIE firmware automatically, which can be then built and simulated directly using AMD Vitis. It targets the AMD AI Engine (AIE) from model-level frontends and lowers supported operators into AIE graphs and kernels as a standalone AIE project.

• Current hardware targets: AIE-ML and AIE-MLv2 devices.
• Current frontend paths: ONNX for explicit operator graphs, and an optional hls4ml frontend path.

aie4ml currently supports Dense/GEMM, dynamic MatMul, Elementwise Add, quantized LayerNorm, quantized Softmax (approx.), last-two-axis Permute, Split/Slice, Concat, fanout, and fused ReLU across AIE-ML and AIE-MLv2 devices.

See the operator support matrix for coverage, tensor and transport contracts, and current limitations.

• AMD Vitis 2025.2 and a valid AIE tools license.
• Python 3.10+.
• Optional: hls4ml if using the hls4ml frontend integration.

The ONNX path is the recommended route for operator-level compiler development and for models that already express quantized tensors and Q/DQ boundaries explicitly. The hls4ml path is intended for MLP-style pipelines at the moment.

pip install aie4ml

Install hls4ml only if you need the hls4ml frontend/backend integration:

pip install hls4ml

Documentation and usage: https://github.com/dimdano/aie4ml

Tutorial 1: tutorials/tutorial_1.ipynb Tutorial 2: tutorials/tutorial_2.ipynb

General hls4ml concepts: https://fastmachinelearning.org/hls4ml

aie4ml is developed and maintained by Dimitrios Danopoulos.

If aie4ml contributes to your research, please cite the corresponding publications:

@INPROCEEDINGS{11552717,
  author={Danopoulos, Dimitrios and Lupi, Enrico and Sun, Chang and Dittmeier, Sebastian and Kagan, Michael and Loncar, Vladimir and Pierini, Maurizio},
  booktitle={2026 IEEE 34th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM)},
  title={AIE4ML: An End-to-End Framework for Compiling Neural Networks for the Next Generation of AMD AI Engines},
  year={2026},
  volume={},
  number={},
  pages={176-184},
  keywords={Tiles;Modeling;Arrays;Kernel;Memory;Information rates;Throughput;System-on-chip;Loading;Engines;ai engines;hls4ml;aie4ml;versal;acceleration;inference},
  doi={10.1109/FCCM68464.2026.00035}}

@misc{danopoulos2026tamingexponentialfastsoftmax,
      title={Taming the Exponential: A Fast Softmax Surrogate for Integer-Native Edge Inference},
      author={Dimitrios Danopoulos and Enrico Lupi and Michael Kagan and Maurizio Pierini},
      year={2026},
      eprint={2604.02292},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2604.02292},
}