Is there a way to upgrade Jetpack from 3.1 to 3.2 without flashing TX2?


  1. Host running Jetpack 3.1 well in Ubuntu 16.04
  2. Target running Jetpack 3.1 flashed by host well.

Is there a “easy” way to upgrade Jetpack from 3.1 to 3.2? (In target only or without flash TX2)

Thanks a lot.

Flash is required for changing L4T release version. If you are worried about saving something, then you could clone first and use the clone on loopback mount on the host at a later date to view the old content. You cannot mix a clone from one release and another release though (mostly that’d just be putting the old release back in with incompatible boot files).

Thanks @linuxdev. Do I need to reinstall Tensorflow after L4T release upgrade? And normally how long in Nvidia’s plan between iterations of Jetpack?

Have a nice night.

Flashing with JetPack completely erases everything from the original install. You would need to once again install anything you’ve added or customized.

I do not know an answer to the time between iterations. However, having a pre-release which has been out for a bit suggests the release has to be in the works already. The technically correct exact answer is “soon-ish” :P.