EdgeSolutions JetPack FAQ

Table of Contents

Hardware acceleration
- How do i install cuda acceleration?
- How do i use gpu acceleration in my docker container?
How do i install jtop for device monitoring?
- Installation
- Launching jtop
How do i enable MAXN power mode?
How to identify the software version of my EdgeDevice?
How to identify the SOM type being installed in my EdgeDevice?
How do i change the keyboard layout?

Hardware acceleration

How do i install cuda acceleration?

The EdgeDevice will be sent out without any CUDA related software, since it need around 5GB of disk-space. Install all needed Libraries for interacting with the GPU:

$ sudo apt update
$ sudo apt install nvidia-jetpack

You can verify the installed libraries with the command jetson_release.

If the program isn’t installed you have to install the package jetson-stats via pip3:

$ pip3 install jetson-stats

The output should look like the this:

$ jetson_release
Software part of jetson-stats 4.3.1 - (c) 2024, Raffaello Bonghi
Model: PCB Arts - EdgeKit - Jetpack 5.1.1 [L4T 35.3.1]
NV Power Mode[0]: MAXN
Serial Number: [XXX Show with: jetson_release -s XXX]
Hardware:
 - Module: Check with sudo
Platform:
 - Distribution: Ubuntu 20.04 focal
 - Release: 5.10.104-tegra
jtop:
 - Version: 4.3.1
 - Service: Inactive
Libraries:
 - CUDA: 11.4.315
 - cuDNN: 8.6.0.166
 - TensorRT: 8.5.2.2
 - VPI: 2.2.7
 - Vulkan: 1.3.204
 - OpenCV: 4.5.4 - with CUDA: NO

How do i use gpu acceleration in my docker container?

First, make sure you have installed the docker runtime:

$ sudo apt update
$ sudo apt install docker.io

Nor you can run docker containers. With the parameter --runtime nvidia you can use the GPU of the EdgeDevice in the container:

$ docker pull dustynv/pytorch:2.6-r36.4.0-cu128
2.6-r36.4.0-cu128: Pulling from dustynv/pytorch
a186900671ab: Pull complete
[...]
d17362a386cf: Pull complete
Digest: sha256:f0ef5a2c0442809e53e55ce192e7aaa72fec72e8a58f648ff2bf93c899a09b48
Status: Downloaded newer image for dustynv/pytorch:2.6-r36.4.0-cu128
docker.io/dustynv/pytorch:2.6-r36.4.0-cu128
$ docker run --runtime nvidia --rm -it --network host dustynv/pytorch:2.6-r36.4.0-cu128
root@edgekit:/# python3
Python 3.10.12 (main, Jan 17 2025, 14:35:34) [GCC 11.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> torch.cuda.is_available()
True
>>>

How do i install jtop for device monitoring?

jtop is a real-time monitoring tool for Jetson platforms, based on Python and tegra-stats.

Installation

sudo apt update
sudo apt install -y python3-pip
sudo pip3 install jetson-stats

Launching jtop

sudo jtop

This opens a TUI displaying CPU/GPU/RAM usage, different temperatures and the current power mode.

How do i enable MAXN power mode?

To unlock full performance on the Jetson, NVIDIA recommends enabling the MAXN power mode.

Check current power mode

sudo nvpmodel -q

Set MAXN mode

sudo nvpmodel -m 0

Note: On Jetson Orin NX, profile 0 typically corresponds to MAXN.
Alternatively, launch jtop, press P, and select the desired power profile.

Verify changes

sudo nvpmodel -q
sudo jetson_clocks

jetson_clocks maximizes clock frequencies - useful for benchmarking or stress testing.
Changes take effect immediately.

Notes

In MAXN mode, the temperature may rise significantly - ensure proper cooling.
jtop also supports logging and CSV export from the command line.

How to identify the software version of my EdgeDevice?

To identify the software version of your EdgeDevice, run following command:

$ cat /etc/edgeboard_version
Version: 36.4.0_2025-02-25
Flavor: orin_generic

In this case, the EdgeDevice was installed with the L4T Version 36.4.0., built on 2025-02-25.

In case you get an error, that the file does not exist:

$ cat /etc/edgeboard_version
cat: /etc/edgeboard_version: No such file or directory

run the command

$ uname -r
5.15.148-tegra

to get the linux kernel release. When it is 5.10.X you have L4T version 35 installed. In case of 5.15.X you have L4T version 36.4 installed.

How to identify the SOM type being installed in my EdgeDevice?

First, you need to connect the EdgeDevice with a USB cable to a Linux PC. Then you start the EdgeDevice in recovery model. With the command lsusb in the Linux PC, you can find the Device:

$ lsusb
Bus 002 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 001 Device 002: ID 05e3:0608 Genesys Logic, Inc. Hub
Bus 001 Device 022: ID 0955:7423 NVIDIA Corp. T234 [Orin NX 8GB] recovery mode
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

Important is the line of the NVIDIA device, the EdgeDevice. Have a look at the device-id: 0955:7423. The second part tells you the SOM type, in this case a Orin NX with 8GB of RAM.

Just look your device id up in this table:

SOM type	Device ID
Jetson Orin NX 16GB	0955:7323
Jetson Orin NX 8GB	0955:7423
Jetson Orin Nano 8GB	0955:7523
Jetson Orin Nano 4GB	0955:7623
Jetson Nano	0955:7f21
Jetson Xavier NX	0955:7e19

SOM type

Device ID

Jetson Orin NX 16GB

0955:7323

Jetson Orin NX 8GB

0955:7423

Jetson Orin Nano 8GB

0955:7523

Jetson Orin Nano 4GB

0955:7623

Jetson Nano

0955:7f21

Jetson Xavier NX

0955:7e19

How do i change the keyboard layout?

To change the keyboard layout, you need to add the layout to the list of keyboard input sources. After this, a new option on the top right appears to switch between the layouts back and forth.

Open the menu by clicking on the 3x3 icon on the left.
Enter Settings to open the Settings app.
Select the section Keyboard in the list on the left.
Click on the + button to add a new layout.
In the next screen, double click on the three dots in the empty list.
Then click on the entry Other in the list.
After some loading, a list of keyboard layout appear. Select the layout you want to add. You can use the search bar at the bottom.
Add the selected keyboard layout with the Add button.
Now you can switch between the keyboard layouts with the icon on the top right.