Installing 0.11.0 on a Jetson Nano with a USB Coral

[LordNex]

For the longest time I had frigate in a RPi4 8gig with the Coral running DeepStack. I now just use CompreFace running inside Home Assistant as my only detector and it seems to do much better

But now that leaves me with extra hardware and when I went to update from RC2 to the stable version of 11 I ran out of disk space on the paltry 32GB SD card I had in the Pi.

So my idea now is to ditch the Pi (Be reused as a media player or something) and use the Jetson Nano 4gig with the USB Coral TPU attached. And a 128gig Class10 SD card

Should I expect to see an improvement or anything? I figured the TPU would handle all the object detection and the CUDA cores properly enabled could handle the encoding and decoding of the video files. Am I wrong in this thought or is there a better way to do this. In addition to all that, everything is tied together and into Home Assistant through DoubleTake.

Suggestions, comments, etc are welcome please.

[LordNex]

Over a year and a half and nada?