Can the newest sdk manager use on ubuntu 20.04

Autonomous Machines Jetson & Embedded Systems Jetson TX2
1

Can the newest sdk manager use on ubuntu 20.04? I see sdk manager support Host OS Ubuntu 16.04, 18.04 and 20.04 on x86_64 system
CentOS/RHEL 7.6, 8.0 and 8.2 on x86_64 system

2

Yes, it can, but the JetPack for TX2 can’t work on Ubuntu 20.04. therefore please still use Ubuntu 18.04 as host to do the development work with TX2.

3

The tx2nx don’t need to work on ubuntu20.04. I just want to use sdk manager deploy cuda and download the jetpack without docker. Does it work well?

4

As I mentioned previously, the JetPack for TX2 can’t work well on Ubuntu 20.04.
You still can try, but we are not able to support any issue from there.

5

~_~
If my host is ubuntu 20.04 and how can I use sdk manager to deploy? Because the new PC hardware driver doesn’t have the ubuntu18.04. It’s too trouble for me to find the driver.

6

JetPack/SDK Manager 4.x will always load Ubuntu 18.04 on the Jetson (it loads L4T, which in turn is just Ubuntu with NVIDIA drivers). JetPack/SDKM 5.x will load Ubuntu 20.04 onto the Jetson regardless of which release of Ubuntu is used on the host PC. The actual creation of an image and the tools to do so come from the host PC, and so some issues tend to mean you might need an Ubuntu 18.04 host instead of 20.04. On my system I have 1/3 of the root hard drive reserved for Windows (I don’t like Windows, but sometimes I need it), 1/3 for the current development o/s (Ubuntu 18.04), and 1/3 for a future release (no doubt Ubuntu 20.04). Then I use a separate hard drive to mount on “/home”, and thus if I change release, I just remount home (there might be issues such as converting bookmarks from an old firefox to a new firefox, but mostly this works out).

I won’t guarantee it, but some people have succeeded in “faking” their Ubuntu 20.04 for 18.04. See:
https://forums.developer.nvidia.com/t/sdkmanager-not-supported-on-linux/71742/4

VMs are not supported, but if USB is correctly set up (this is a topic for the VM support, each VM differs), then it can work if you have enough disk space.

1 Like
7

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