Arduino is working hard creating a set of ML enabled boards with varying degrees of HW acceleration. We started a few years ago with Nano 33 BLE Sense which was adopted by Google TensorFlow team as a primary platform and evolved with Portenta,
Nicla Vision and Nicla Sense ME where we implemented various degrees of HW acceleration, primarily focused at low power. Our partnerships with OpenMV and Edge Impulse have also greatly boosted product adoption thanks to the extreme simplification of the learning process.
