NVIDIA Jetson T4000 and NVIDIA JetPack 7.1 now available: Accelerate AI Inference for Edge and Robotics

Robotics & Edge Computing Jetson Systems Jetson Thor
1

NVIDIA Jetson T4000 is now available to place orders

NVIDIA Jetson T4000 is now available to place orders. The NVIDIA Jetson T4000, brings high-performance AI and real-time reasoning to a wider range of robotics and edge AI applications. Optimized for tighter power and thermal envelopes, T4000 delivers up to 1200 FP4 TFLOPs of AI compute and 64 GB of memory, providing an ideal balance of performance, efficiency, and scalability.

Find more details and technical documents for the Jetson T4000 here.

JetPack 7.1/Jetson Linux 38.4 for NVIDIA Jetson Thor is now live

We are pleased to announce the production release of JetPack 7.1. JetPack 7.1 is the next release in the JetPack 7 series, and brings the support for NVIDIA Jetson T4000 module and new software features in addition to support for existing Jetson AGX Thor Developer Kit and the Thor based T5000 module. With JetPack 7 series, Jetson software aligns with the Server Base System Architecture (SBSA), positioning Jetson Thor alongside industry-standard ARM server design. JetPack 7.1 packages Jetson Linux 38.4 with Linux Kernel 6.8 and Ubuntu 24.04 LTS based root file system.

What’s new in JetPack 7.1

  • Support for Jetson T4000 module

  • Support for Video Codec SDK

  • Support for TensorRT EdgeLLM

BLOG POST: Learn more about the key features in this release in the supporting blog post here.

You can install JetPack 7.1 on your Jetson AGX Thor Developer Kit with any of the methods below

  • ISO Image: You can download the JetPack 7.1 ISO image from JetPack 7.1 Downloads and Notes page and use Balena Etcher to prepare the installer USB drive. Please follow the instructions in the User Guide on flashing the Developer Kit with JetPack 7.1.

  • SDK Manager: You can do a fresh install of JetPack 7.1 using SDK Manager.

  • Manual Flashing: If you prefer to install using the command line, you can flash Jetson device from a linux host by following steps here.

Once Jetson Linux is flashed, you can install the compute stack using:

  • SDK Manager (using linux host) or

  • By running “sudo apt update” followed by “sudo apt install nvidia-jetpack” on Jetson

Important Notes:

1. The manual flashing instructions have slightly changed because of Thor leveraging SBSA architecture. Please follow the manual flashing instructions carefully

2. If you are flashing JetPack 7.1 from an ISO image on a developer kit that was previously flashed with JetPack 7.0, please follow the guide here.

Containers

Note: Since Jetson Thor is based on the SBSA stack, the “-igpu” container tag is no longer required when running on Thor.

JetPack 7.1/Jetson Linux 38.4 resources

1 Like
Failed to fetch source through source_sync
3

Is there a way to upgrade an existing Thor? I have a lot of stuff on my drive that I would prefer not to have to copy off.

4

edit /etc/apt/sources.list.d/nvidia-l4t-apt-source.list change 38.2 to 38.4

deb https://repo.download.nvidia.com/jetson/common r38.4 main
deb https://repo.download.nvidia.com/jetson/som r38.4 main
deb https://repo.download.nvidia.com/jetson/ffmpeg r38.4 main

Then:
sudo apt update
sudo apt upgrade

2 Likes
5

Is the cross-compilation container available along with jetpack 7.1 ?

The most recent one at JetPack Cross Compilation container | NVIDIA NGC is still for jetpack 6.1 .

6

Thanks, @whitesscott . That worked. At first I thought it didn’t because I went to my Thor in person and tried to run Firefox, which kept crashing. Then after looking at the syslog I saw the system didn’t have enough memory because I was already running GPT-OSS-120b! I stopped it an all was well for Firefox. But I went back to GPT-OSS because why not?

1 Like
7

Cool.

Are there plans to publish TensorRT Edge LLM as a container?
If possible I prefer to use containers for repeatability.