Facing difficulties in setup of Nvidia Jetson Nano

  1. NVIDIA Jetson Nano DK has Ubuntu18 deb (18.04).

  2. We can only use python 3.6.9. Python 3.6 is also very old and not supported by new version of tf and cv2

  3. We can only use 2.6.1 version for Tensorflow

  4. It has 4.5.3 version for opencv

  5. Tense flow requires at least 6 GB of physical Ram but currently we have 4GB of physical Ram in NVIDIA Jetson Nano also we are using openCV which require a huge amount of swap space currently we try to increase it to maximum 6gb but overall the system crashs after reaching the physical memory limit of 4GB. We have also try to optimise tensorflow but as openCV and tenseflow both are working simultaneously there is huge demand of resource so we are getting research error.

  6. Nvidia has stopped giving updates to jetson nano So, it is only getting maintained and newer libraries are not provided.


We have a new Orin Nano that can support JetPack 5 and solve most of your concerns.
It also has an 8GB version to allow users to use more memory.


This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.