Your question is Deep Learning on Embedded Hardware. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
What are the challenges of deploying deep learning models on embedded hardware with limited compute?
Discuss this as a practical deployment problem, including model compression, numerical precision, runtime compatibility, memory use, latency, power, and accuracy validation. Provide a small implementation that converts, evaluates, and benchmarks a neural network for an embedded-style CPU target.