To bring machine learning to the edge, hardware-aware neural architecture search (NAS) has shown to be an effective optimization method across a multitude of use cases. In this presentation, Sebastian Vogel from NXP introduces the key components of hardware-aware NAS. In this context, peculiarities and benefits of hyper-parameter optimization, quantization, and latency estimation of neural networks on edge devices will be discussed. The presentation concludes with an overview of NXP’s enabling technology for AI at the Edge.