Jetpack 5.0 deep dive webinar was just launched in May.10th, 2022. Check out NVIDIA on-demand webinar here. The webinar covered:
- A deep dive into JetPack 5.0
- Migration from JetPack 4 to JetPack 5
- Jetson Software roadmap for 2022
Let take a look at key takeaways from the webinar and current Jetpack 5.0 release infomation. Noted Jetpack 5.0 Production release for Jetson AGX Orin 32GB production module will be available in July as NVIDIA Jetson software roadmap listed.
In April, NVIDIA launched JetPack 5.0 Developer Preview. This is a development release with a full compute stack update including CUDA 11.4, as well as Jetson Linux 34.1 with Linux Kernel 5.10, an Ubuntu 20.04 based root file system, a UEFI based bootloader, and OP-TEE as Trusted Execution Environment.
The release supports Jetson AGX Orin Developer Kit, Jetson AGX Xavier series and Jetson Xavier NX series developer kits and production modules.
Explore Seeed Edge AI Devices powered by NVIDIA Production Module:
- reComputer with NVIDIA Jetson Nano
- reComputer with NVIDIA Jetson Xavier NX
- reserver with NVIDIA Jetson AGX Orin
JetPack 5.0 Developer Preview includes:
- Support for Jetson AGX Orin Developer Kit
- Support to emulate the performance of production Jetson Orin NX series and Jetson AGX Orin series modules with Jetson AGX Orin Developer Kit
- LTS Kernel 5.10
- Ubuntu 20.04 based reference file system
- OP-TEE Trusted Execution Environment
- UEFI bootloader
- Support for NVSCI: Provides utilities for streaming data packets between different applications and for interprocess communication (IPC)
- New Jetson Power GUI: Tool for monitoring power and thermal status
- New NVUtils Multimedia buffer management and buffer transformation API
- NVIDIA Container Runtime no longer mounts user level libraries like CUDA, cuDNN and TensorRT from the host. These will instead need to be installed inside the container.
JetPack 5.0 Developer Preview is not for production use and has known issues documented in the Jetson Linux release notes.
Features not included yet but will support Jetpack 5.0 Production Release
- Support Jetson AGX Xavier Industrial Module
- Secure Boot and disk encryption
- Image-based over-the-air updates tools
- Rootfs A/B
- OP-TEE on Jetson AGX Orin (Jetson Xavier NX and Jetson AGX Xavier series support it)
- Jetson Xavier NX SD card updating the Xavier NX SPI components
- Flashing an SD card image for a Jetson Xavier NX Developer Kit out of the box (you can flash an SD card image after you have flashed the developer kit with SDK Manager)
JetPack 5.0 Developer Preview containers at NGC
Containers can now run directly on Jetson Linux BSP
- Jetson Linux Base container
- CUDA Runtime container
- TensorRT Runtime container
- TensorFlow container
- PyTorch container
- Machine Learning container
Install Jetpack component using apt
List of Debian packages: https://docs.nvidia.com/jetson/jetpack/install-jetpack/index.html
Sample dockerfiles: https://github.com/dusty-nv/jetson-containers
Migration from Jetpack 4 to Jetpack 5
- Migrating from Cboot to UEFI
- Migrating from Trusty to OP-TEE
- Camera Driver Migration from Kernel 4.9 to Kernel 5.10
- Migration from NvBuf Utils to Nvutils
Emulate ORIN NX at Jetson AGX Orin
Now you can get started with development for Jetson AGX Orin or Jetson Orin NX on Jetson AGX Orin Developer Kit!
Emulate Performance of following Orin production modules:
• Jetson AGX Orin 32GB
• Jetson Orin NX 16GB
• Jetson Orin NX 8GB
Just flash Jetson AGX Orin Developer Kit with appropriate flashing config:
• jetson-agx-orin-devkit-as-jao-40w.conf: For Jetson AGX Orin 32GB emulation
• jetson-agx-orin-devkit-as-nx-16gb.conf: For Jetson Orin NX 16GB emulation
• jetson-agx-orin-devkit-as-nx-8gb.conf: For Jetson Orin NX 8GB emulation
Example: To emulate Jetson Orin NX 16GB performance on Jetson AGX Orin Developer Kit:
sudo ./flash.sh jetson-agx-orin-devkit-as-nx-16gb mmcblk0p1
VPI Webinar showcasing PyTroch and Python interoperability: June 10th
Create a minimal BSP and JetPack
About Jetpack SDK
NVIDIA JetPack SDK is the most comprehensive solution for building end-to-end accelerated AI applications. All Jetson modules and developer kits are supported by JetPack SDK.
JetPack SDK includes the Jetson Linux Driver Package (L4T) with Linux operating system and CUDA-X accelerated libraries and APIs for Deep Learning, Computer Vision, Accelerated Computing and Multimedia. It also includes samples, documentation, and developer tools for both host computer and developer kit, and supports higher level SDKs such as DeepStream for streaming video analytics and Isaac for robotics.
NVIDIA JetPack includes NVIDIA Container Runtime with Docker integration, enabling GPU accelerated containerized applications on Jetson platform. Jetpack also brings support for NVIDIA Triton™ Inference Server to simplify the deployment of AI models at scale.