Part 2: Computing at the Edge | Flexible Options for Inference Implementation at the Edge