I would like to load an object detection model into GPU memory on a nano, do inference on about 5 frames, put it to sleep, wake it up using a LAN message, do inference on about 5 frames and so on…
Can Nano be woken by a LAN message?
I read that NX takes 6 seconds to sleep and 3 to wake. Is the nano (or Tx2) similar?
What is the difference between sleep and deep sleep mode?
Approximately how much power is consumed by nano in deep sleep? (I looked at your software power consumption modeling but I didn’t understand it)