Jetson Nano — Ubuntu 22.04 Server Image

A ready-to-run Ubuntu 22.04 (Jammy) server image for the Jetson Nano with CUDA, TensorRT, OpenCV, Python 3.10.12 and automatic fan control pre-installed.

This image is based on NVIDIA JetPack 4.6.6 / L4T R32.7.6 and upgrades the userspace to Ubuntu 22.04.

⚠️ Private Beta — This image is currently in private beta testing. The public download link will be published once everything is verified and working. If you'd like to help test, feel free to send me a DM or open a GitHub Issue.

---

What's Included

Component	Version	Status
Ubuntu	22.04 LTS (Jammy)	✅
L4T	R32.7.6	✅
Kernel	4.9.337-tegra	✅
CUDA	10.2.300	✅
cuDNN	8.2.1	✅
TensorRT	8.2.0.1	✅
OpenCV	4.8.1 (CUDA)	✅
VPI	1.2.3	✅
VisionWorks	1.6.0.501	✅
DeepStream	6.0.1	✅
GStreamer	with nvvidconv, nvv4l2decoder, nvv4l2h264enc	✅
MIPI CSI cameras		✅
Python	3.10.12	✅
jtop	latest	✅
JetFanC	latest	✅

NOT Included

• Desktop environment (maybe in future as separate image)
• Docker (maybe in future as separate image)
• Vulkan drivers are not supported on the Jetson Nano.

---

Requirements

• Jetson Nano (4GB or 2GB)
• SD card — 64GB recommended, 16GB minimum
• Monitor + keyboard or ethernet connection for headless setup via SSH
• optional: USB Stick for swapfile and other high IO files (reduces sd card wear).

---

1. Flash the SD Card

With Balena Etcher (recommended)

Download Balena Etcher, select jetson-nano-ubuntu22-server.img.xz and flash to your SD card. Etcher handles decompression automatically.

With dd

⚠️ Replace /dev/sdX with your actual SD card device. Double check with lsblk — writing to the wrong device will destroy your data.

xz -dc jetson-nano-ubuntu22-server.img.xz | sudo dd of=/dev/sdX bs=4M status=progress
sync

---

2. First Boot

Insert the SD card into your Jetson Nano and power it on.

On first boot the partition will automatically expand to fill your SD card — this takes about 30 seconds. The board will then be ready to use.

Login credentials:

User	Password
jetson	jetson

🔒 Change your password after first login with passwd

---

3. Connect to Network

Ethernet

Just plug in a cable, it connects automatically.

WiFi

# list available networks
nmcli device wifi list

# connect to your network (will prompt for password)
sudo nmcli device wifi connect "YourSSID" --ask

Headless via SSH

ssh jetson@nano1.local

Or use the IP address shown on the monitor.

---

4. Verify System

# Jetson system monitor (CPU, GPU, RAM, temperature, fan)
sudo jtop

# Fan control service status
sudo systemctl status jetfanc

# Check CUDA
nvcc --version

# Run full SDK validation
sudo ~/test-sdk.sh

Note: if nvcc is not found interactively, add CUDA to your PATH permanently:

sudo tee /etc/profile.d/cuda.sh >/dev/null <<'EOF'
export PATH=/usr/local/cuda/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
EOF
source /etc/profile.d/cuda.sh

---

Included Services

Service	Description
jetfanc	Automatic PWM fan control
jtop	Jetson stats monitor
ssh	Remote access
NetworkManager	WiFi and ethernet management
expand-rootfs	Auto-expand partition on first boot (runs once)

---

Known Limitations

• CUDA 10.2 only — newer CUDA versions are not supported on Jetson Nano
• OpenCV 4.5.4 is CPU only for now — a CUDA-enabled build is in progress
• JetPack 4.6.6 base — some newer libraries targeting JetPack 5.x will not be compatible
• Python bindings for TensorRT and VPI are not functional — C++ APIs work correctly
• No desktop environment — this is a server image
• Docker not included - work in progress

---

Tested On

• Jetson Nano 4GB (B01)

---

Credits

This image builds upon work from the community:

• xronos-inc — Jetson Nano Ubuntu 22.04 upgrade method https://github.com/xronos-inc/jetson-nano-ubuntu-22.04

• Qengineering — Jetson Ubuntu & CUDA guidance https://qengineering.eu/install-ubuntu-20.04-on-jetson-nano.html

This image includes:

• jtop (jetson_stats) licensed under AGPL-3.0 Source: https://github.com/rbonghi/jetson_stats License: AGPL-3.0

• JetFanC licensed under MIT https://github.com/mischa-robots/JetFanC

---

Author

Michael "Mischa" Schaefer

---

⚠️ License & Legal Notice

This image contains proprietary NVIDIA software components from Linux for Tegra (L4T), including CUDA, cuDNN, TensorRT and related drivers.

These components are redistributed unmodified under the original NVIDIA Tegra Software License Agreement.

• No NVIDIA binaries have been modified.
• Some sample and demo files were removed to reduce image size.
• NVIDIA components remain property of NVIDIA Corporation.
• This project is not affiliated with or endorsed by NVIDIA.

The full NVIDIA license text is included in this repository as:

• NVIDIA-L4T-License.txt

Or on the official NVIDIA website:

• Tegra_Software_License_Agreement-Tegra-Linux

Your use of this image must comply with the NVIDIA Tegra Software License Agreement.