目录
0. Introduction of Jetson Developer kits
Taking Performance to the Edge
A Jetson AGX Xavier For Any Application
2. NVIDIA SDK manager and Jetpack
1. Install DRIVE with SDK Manager
Step 1: Set Up the Development Environment
Step 2: Review Components and Accept Licenses
2.1. Recommended Recovery Steps
1.3.1. Install JetPack Components on Jetson Linux
1.4. List of JetPack Debian Packages
0. Introduction of Jetson Developer kits
NVIDIA® Jetson™ 采用节能高效和小巧精致的外形设计，可为边缘提供加速 AI 性能。这些 Jetson 模组与 NVIDIA JetPack™ SDK 一起，为您开启了在各行各业开发和部署创新产品的大门。
Jetson 系列模组均使用相同的 NVIDIA CUDA-X™ 软件，并支持容器化和编排等云原生技术，以便构建、部署和管理边缘 AI。
客户可以借助 Jetson 加速所有现代 AI 网络，轻松推出新功能，并将同一软件用于不同的产品和应用。
1. 硬件对比
Jetson 模组系列
Jetson AGX Orin 系列
借助功能强大的 AI 计算机，为节能高效的自主机器带来新一代产品。它拥有高达 275 TOPS 的算力，性能是上一代产品的 8 倍，适用于多个并发 AI 推理管道，此外还可以通过高速接口连接多个传感器，因此是制造、物流、零售和医疗健康领域应用的理想解决方案。
JETSON AGX ORIN 64GB | JETSON AGX ORIN 32GB
Jetson Orin NX 系列
在极小的 Jetson 外形规格中，体验专为节能高效的自主机器打造的强大 AI 计算机。与 NVIDIA Jetson Xavier™ NX 相比，它能够提供高达 5 倍的性能和 2 倍的 CUDA® 核心数，以及多个传感器的高速接口支持。Jetson Orin NX 拥有高达 100 TOPS 的算力，适用于多个并发 AI 推理管道，并且外形小巧，性能出色。
JETSON ORIN NX 16GB | JETSON ORIN NX 8GB
Jetson AGX Xavier 系列
Jetson AGX Xavier 系列模组可在边缘提供更高级别的计算密度、能效和 AI 推理能力。Jetson AGX Xavier 附带预设为 10W、15W 和 30W 的可配置功率模式，Jetson AGX Xavier 工业级则附带预设为 20W 和 40W 的可配置功率模式。这些功率模式在运行时可切换，并可根据特定需求进行定制。这些模组可将能效提升至 Jetson TX2 的 10 倍，并将性能提升至 Jetson TX2 的 20 倍。
JETSON AGX XAVIER 64GB | JETSON AGX XAVIER | JETSON AGX XAVIER 工业级
Jetson Xavier NX 系列
外形小巧的 Jetson Xavier NX 模组可为边缘提供高达 21 TOPS 的加速 AI 计算。它能并行运行多个现代神经网络，处理来自多个高分辨率传感器的数据，满足完整 AI 系统的需求。Jetson Xavier NX 是一款支持量产的产品，支持所有热门 AI 框架。
JETSON XAVIER NX 16GB | JETSON XAVIER NX
Jetson TX2 系列
Jetson TX2 模组系列的性能高达 Jetson Nano 的 2.5 倍。这些模组可将节能高效的嵌入式 AI 计算引入一系列用例，包括采用更精简外形的 Jetson TX2 NX 的大众市场产品，以及采用坚固耐用 Jetson TX2i 的专用工业环境。
JETSON TX2 NX | JETSON TX2 4GB | JETSON TX2 | JETSON TX2i
Jetson Nano
Jetson Nano 是一款外形小巧、性能强大的计算机，专用于嵌入式 AI 系统和 IoT，可在低功耗平台中实现现代 AI 的强大功能。借助 NVIDIA Jetpack SDK 和完整的桌面 Linux 环境快速入门，并开始探索嵌入式产品的新领域。
* Jetson Nano 模组和 Jetson Xavier NX 模组是 Jetson Nano 开发者套件的一部分，Jetson Xavier NX 开发者套件配备一个插槽，可使用 microSD 卡（而非 eMMC）作为系统存储设备。
** Jetson AGX Orin 系列和 Jetson Orin NX 系列的虚拟通道相关摄像头信息并非最终信息，可能会发生变化。
请参阅此处的 NVIDIA 在线软件文档，获取有关受支持功能的信息。
2. 应用场景
智能设备 OEM 和 AI 应用开发者使用 Jetson 研发制造、物流、零售、服务、农业、智慧城市、医疗健康和生命科学等领域的突破性产品。Embedded Computing Systems for Product Development | NVIDIASee how companies in a wide variety of industries are transforming their business with Jetson embedded systems. https://www.nvidia.com/en-us/autonomous-machines/embedded-systems/product-development/#success-stories
AGX Xavier
1. Introduction of AGX Xavier
AGX Xavier
Taking Performance to the Edge
As the world’s first computer designed specifically for autonomous machines, Jetson AGX Xavier has the performance to handle the visual odometry, sensor fusion, localization and mapping, obstacle detection, and path-planning algorithms that are critical to next-generation robots. Get GPU workstation-class performance with up to 32 TOPS of peak compute and 750Gbps of high-speed I/O in a compact form factor.
A Jetson AGX Xavier For Any Application
The Jetson AGX Xavier series of modules delivers up to 32 TOPS of AI performance and NVIDIA’s rich set of AI tools and workflows, letting developers train and deploy neural networks quickly.
1.1 AGX Xavier硬件组成
外观
IO接口
- PCIe X16（x8 PCIe Gen4 / x8 SLVS-EC）
- （2x）USB3.1（可选显示端口）（可选电源）
- （16x）MIPI CSI-2通道（摄像头）
- RJ45（千兆以太网）
- M.2键M（NVMe）
- M.2键E（PCIe x1 + USB 2.0 + UART（用于Wi-Fi / LTE）/ I2S / PCM）
- eSATAp + USB3.0 Type A（SATA通过PCIe x1桥（PD + 2.5英寸SATA的数据）+ USB 3.0）
- uSD / UFS卡插槽（SD / UFS）
- 40针接头连接器（UART + SPI + CAN + I2C + I2S + DMIC + GPIO）
- 高清音频头（高清音频）
- HDMI Type A（HDMI 2.0）
- 9.0-20VDC电源桶式插孔
1.2 驱动软件
NVIDIA Jetson AGX Xavier是NVIDIA Jetson TX2的升级版本，它的性能和效率是NVIDIA Jetson TX2的20倍以上。
新的Jetson AGX Xavier模块使以AI为动力的自动机成为可能，其运行功率低至10W，最多可提供32个TOP。作为全球领先的AI计算平台的一部分，它受益于NVIDIA丰富的AI工具和工作流，使开发人员能够快速训练和部署神经网络。NVIDIA JetPack SDK支持Jetson AGX Xavier，它可以通过减少开发工作量和费用来帮助您节省大笔费用。
借助NVIDIA Jetson AGX Xavier开发者套件，您可以轻松地创建和部署端到端AI机器人应用程序，以用于制造、交付、零售、农业等。
该套件受NVIDIA JetPack和DeepStream SDK、CUDA®、cuDNN和TensorRT软件库的支持，提供了开箱立即上手所需的所有工具。由于它采用了新的NVIDIA Xavier处理器，使它的性能和效率是NVIDIA Jetson TX2的20倍以上，而能源效率也高出10倍。
https://developer.nvidia.com/embedded/jetpack
产品特性¶
- 人工智能性能: 32 TOPs
- 超过NVIDIA Jetson TX2的性能20倍，能源效率提高10倍
- 受NVIDIA JetPack和DeepStream SDK、CUDA®、cuDNN和TensorRT软件库的支持
JetPack¶
NVIDIA JetPack SDK是用于构建AI应用程序的最全面的解决方案。它捆绑了所有Jetson平台软件，包括TensorRT、cuDNN、CUDA工具包、VisionWorks、GStreamer和OpenCV。所有这些软件均基于带有LTS Linux内核的L4T构建。
JetPack为您的Jetson开发人员工具包提供了最新的OS映像、库和应用程序接口、示例、文档以及开发人员工具。
JetPack 4.3¶
JetPack 4.3支持所有Jetson模块的最新生产版本，包括Jetson AGX Xavier系列，Jetson TX2系列，Jetson TX1和Jetson Nano。
主要功能包括新版TensorRT和cuDNN，将AI推理性能提高25％，以及针对所有JetPack组件的新的基于Debian软件包的安装机制。它全面支持Jetson AGX Xavier系列的DLA INT8，进一步提高了DLA的性能和效率，从内部支持DLA、CSI和编码。
主要特点¶
JetPack包括操作系统映像、库和应用程序接口、开发人员工具、示例和文档。
有关以下JetPack组件和功能的更多详细信息，请参阅JetPack发行说明。
操作系统（OS）
NVIDIA L4T提供了启动程序、Linux内核、必要固件、NVIDIA驱动程序、示例文件系统等。
JetPack 4.3 特点：
- 支持通过Debian软件包存档安装JetPack组件
- JetPack组件通过NVIDIA托管的公共APT服务器以Debian软件包的形式提供。使JetPack组件的安装更加容易，并升级到将来的版本。
- NVIDIA容器运行时支持内部的DLA、CSI和编码
- 在所有Jetson产品上：容器内支持摄像机串行接口（CSI）和NVENC。
- 仅在Jetson AGX Xavier系列上：容器内现在支持NVIDIA深度学习加速器（DLA）引擎。
- 支持DeepStream 4.0.2
- 支持ISAAC SDK版本2019.3
TensorRT
TensorRT是用于图像分类、分割和对象检测神经网络的高性能深度学习推理运行。它加快了深度学习推理的速度，并减少了卷积和反卷积神经网络的运行时内存占用量。
JetPack 4.3 特点：
- 新图层、运算符和调整大小的操作
- 新样品
- 新的优化
cuDNN
CUDA深度神经网络库为深度学习框架提供了高性能的原语。它包括对卷积，激活函数和张量转换的支持。
JetPack 4.3 特点：
- 分组卷积的性能增强
- 增强的深度和分组卷积
CUDA
CUDA工具包为构建GPU加速应用程序的C和C++开发人员提供了全面的开发环境。该工具包包括用于NVIDIA GPU的编译器、数学库以及用于调试和优化应用程序性能的工具。
多媒体应用程序接口（API）
Jetson多媒体API软件包提供了用于灵活的应用程序开发的低级API。
摄像头应用程序API：libargus为摄像头应用程序提供了低级帧同步API，具有每帧摄像头参数控制，多个（包括同步）摄像头支持和EGL流输出。需要ISP的RAW输出CSI摄像机可以与libargus或GStreamer插件一起使用。 无论哪种情况，都使用V4L2媒体控制器传感器驱动程序API。
传感器驱动程序API：V4L2 API支持视频解码、编码、格式转换和缩放功能。 用于编码的V4L2开辟了许多功能，例如比特率控制、质量预设、低延迟编码、时间折衷、运动矢量映射等。
计算机视觉
VisionWorks是用于计算机视觉（CV）和图像处理的软件开发包。
OpenCV是用于计算机视觉，图像处理和机器学习的领先开源库，现在具有GPU加速功能以实现实时操作。
视觉编程接口（VPI）是一个提供在可编程视觉加速器（PVA），GPU和CPU上实现的计算机视觉/图像处理算法的软件库。
JetPack 4.3 特点：
- 从JetPack 4.3开始，VPI可用作开发人员预览版
Note
- PVA仅支持适用于Jetson AGX Xavier系列
- 此版本未对GPU和CPU的实现进行性能优化。未来的版本将带来性能优化的GPU和CPU实施
开发者工具
JetPack包含以下开发人员工具。有些直接在Jetson系统上使用，有些则在连接到Jetson系统的Linux主机上运行。
CUDA工具包为C和C++开发人员提供了一个全面的开发环境，允许开发人员使用CUDA库构建高性能GPU加速的应用程序。该工具包包括Nsight Eclipse Edition，调试和性能分析工具（包括Nsight Compute），以及用于交叉编译应用程序的工具链。
NVIDIA Nsight Systems是一种低开销的全系统性能分析工具，可为开发人员提供分析和优化软件性能所需的见解。
NVIDIA Nsight Graphics是用于调试和分析图形应用程序的独立应用程序。
NVIDIA Nsight Compute是用于交互式CUDA内核配置文件的，它在Linux主机上运行。支持Jetson AGX Xavier。
Note
对于Jetson AGX Xavier，不推荐使用Visual Profiler。 鼓励开发人员改用Nsight Systems和Nsight Compute。
对于旧版本的JetPack，请访问JetPack存档。
以下示例说明了JetPack组件的使用。这些都包含在参考文件系统中，并且可以在开发人员工具包上进行编译。
JetPack组件 示例所处参考文件系统上的位置 TensorRT /usr/src/tensorrt/samples/ cuDNN /usr/src/cudnn_samples_/ CUDA /usr/local/cuda-/samples/ MM API /usr/src/tegra_multimedia_api VisionWorks /usr/share/visionworks/sources/samples/
/usr/share/visionworks-tracking/sources/samples/
/usr/share/visionworks-sfm/sources/samples/OpenCV /usr/share/OpenCV/samples/ VPI /opt/nvidia/vpi/vpi-/samples/
2. NVIDIA SDK manager and Jetpack
https://developer.nvidia.com/embedded/jetpack
https://developer.nvidia.com/embedded/jetson-linux-r351
2.1 安装Ubuntu20.04LTS
1. 设备清单
- Linxu主机×1 （装有linux ubuntu16.04/18.04的电脑，存储容量充足：/home至少50g ，SDK manager下载和安装需要）
- NVIDIA AGX XAVIER ×1、供电、USB-C数据线
- 显示器×1
- 键鼠
2. 接线
主要分为三部分:
a: 连接显示器及电源线
连线后，启动盒子，显示器会出现开机画面，并进入ubuntu18.04系统
b:将盒子与主机相连，使用自带的USB-typec线(选用USB3.0接口，typec端连在盒子靠近开关键一侧)连好线后，保持盒子通电但是关机状态。然后同时按住盒子上左边和中间的按钮两秒钟，再同时松开，typec口旁边的小灯亮起。
成功之后，在主机终端输入lsusb命令，如果出现NVIDIA CROP字样，说明连接成功。此时连接盒子的显示器是关闭状态。
c:将主机与盒子置于同一个局域网下
3. 刷机
在Host主机上安装SDK managerJetPack SDK | NVIDIA Developer
SDK安装及使用过程如下
1. Install DRIVE with SDK Manager
This section is intended to help you use the NVIDIA SDK Manager GUI to configure your development environment successfully.
Step 1: Set Up the Development Environment
From the Step 01 Development Environment window, select the following:
From the Product Category panel, select the DRIVE development environment.
From the Hardware Configuration panel, select the host machine and target hardware.
When a DRIVE target is connected, SDK Manager will auto-select DRIVE from the supported product category.
From the Target Operating System panel, select the desired operating system, such as Linux or QNX. Notice that the target operating systems available can change, depending on the selected options in the other panels.
If relevant, select any Additional SDKs that you want to install.
An ellipsis (...) in the bottom right corner of a category box indicates that more than one option is available. Click on the ellipsis to show a drop-down menu of available options.
Note:
Your display may differ from the one shown here. The information in this screen is populated from your NVIDIA user account access and permissions. If you don't see your product category in the available selections, verify that your NVIDIA account is registered to the required programs.
Click Continue to proceed to the next step.
Note:
If you are setting up a DRIVE OS QNX environment, you will see a pop-up where the path to the QNX toolchain should be entered before proceeding
Step 2: Review Components and Accept Licenses
From Step 02 Details and License, you can expand the host components and target components panels to review the components that will be installed on your system.
To review the licenses, click on the license agreements hyperlink at the bottom of the page.
Enable the checkbox to accept the terms and conditions of the license agreements.
If you want SDK Manager to download all setup files to a location other than the default path, go to the Download & Install Options located at the bottom of the screen, then select the path you want to use.
For more information about the Download & Install Options, refer to Offline Install.
Select Continue to proceed to the next step.
Step 3: Installation
Before the installation begins, SDK Manager prompts you to enter your sudo password.
The display shows the progress of the download and installation of the software.
Select Pause / Resume to toggle the download and installation process.
At the top, you can toggle between the Details and Terminal tabs. The Terminal tab displays detailed information about the download and installation, with any errors highlighted.
On the Terminal tab, you can use the Filter text field to filter and search for specific information.
SDK Manager opens a dialog when it is ready to flash your target device. A prompt provides instructions for preparing your device to get it ready for flashing.
Note:
The instructions in the flashing dialog vary based on your host and target environment settings.
Note:
Pop-up Windows on the Linux Host During Target Flashing
When flashing the DRIVE AGX platform, different windows may pop up on the host. This can be seen on all DRIVE Software and DRIVE OS releases. These pop-up windows are harmless and do not affect flashing of the unit.
When flashing the DRIVE AGX Developer Kit, some data (e.g., user data, ssh keys, etc.) may be stored in a persistent partition that is not overwritten during flash. This allows you to keep your work data during re-flash. While this is default behavior, in some cases, you may need to wipe the persistent partition during flash. If so, enable the Force wipe of user logins in persistent partition option.
Note:
We care about the safety and security of your data; therefore, SDK Manager prompts you to change your password before enabling the network on DRIVE AGX Systems.
SDK Manager will now complete the installation of the software libraries. Skipping this step will not install any SDK components on your target hardware, and will keep a clean operating system on your device.
Step 4: Finalize Setup
From Step 04 Summary Finalization, there is a summary of the components that were installed, along with any warnings or errors that were encountered.
The Export Debug Logs link creates a ZIP file of all log files created during installation. This ZIP file is located in the same folder path where the SDK Manager installer downloaded all components.
Alternatively, click the menu icon in the top right corner of the window ("⋮"), and choose Export Debug Logs from the drop-down menu in the top-right corner.
Consult the Error Messages for information about any errors you may encounter.
Click Finish and Exit to complete the installation.
2. Repair and Uninstall
To update or uninstall an SDK on your system, launch SDK Manager again.
On Step 1, under the installation step numbers, click the Repair/Uninstall hyperlink.
The Manage NVIDIA SDKs screen shows what has been installed on your system. You can select whether to repair a broken installation, update an existing SDK, or uninstall an SDK.
2.1. Recommended Recovery Steps
There are many causes of various installation errors. Below is a checklist of common installation issues, which may help you recover from a broken installation.
Review the summary table to identify which component failed.
If the error is related to an environmental issue, such as a broken apt repository or missing prerequisite, try to fix it manually, then click the Retry Failed Items button.
Retrying the installation is also available in two other ways:
Finally, try to uninstall and reinstall the relevant SDK.
4. 进入sdkmanager刷机
Download and Run SDK Manager :: NVIDIA SDK Manager Documentation
01- Ubuntu host: install the Debian package.
Ubuntu 16.04, 18.04, 20.04, or 22.04:
sudo apt install ./sdkmanager_[version]-[build#]_amd64.deb 02- 启动sdkmanger
sdkmanager03- 登陆NVIDIA Developer
04-进入图形化界面
将盒子与主机相连，使用自带的USB-typec线(选用USB3.0接口，typec端连在盒子靠近开关键一侧)连好线后，保持盒子通电但是关机状态。然后同时按住盒子上左边和中间的按钮两秒钟，再同时松开，typec口旁边的小灯亮起。
成功之后，在主机终端输入lsusb命令，如果出现NVIDIA CROP字样，说明连接成功。此时连接盒子的显示器是关闭状态。
2.2 安装Jetpack
install Jetpack 5.02 and update(src. -ports, l4t.list, upgrade\install)
(CUDA11.4\cuDNN\TensorRT..)
更换源：Ubuntu 20.04系统下更改apt源为阿里源 - 知乎
执行如下命令：
sudo vim /etc/apt/sources.list
保存后更新
sudo apt-get update报错：
E: Failed to fetch http://mirrors.aliyun.com/ubuntu/dists/focal/main/binary-arm64/Packages 404 Not Found [IP: 117.185.132.27 80]
E: Failed to fetch http://mirrors.aliyun.com/ubuntu/dists/focal-security/main/binary-arm64/Packages 404 Not Found [IP: 117.185.132.27 80]
E: Failed to fetch http://mirrors.aliyun.com/ubuntu/dists/focal-updates/main/binary-arm64/Packages 404 Not Found [IP: 117.185.132.27 80]
E: Failed to fetch http://mirrors.aliyun.com/ubuntu/dists/focal-backports/main/binary-arm64/Packages 404 Not Found [IP: 117.185.132.27 80]
E: Failed to fetch http://mirrors.aliyun.com/ubuntu/dists/focal-proposed/main/binary-arm64/Packages 404 Not Found [IP: 117.185.132.27 80]
E: Some index files failed to download. They have been ignored, or old ones used instead
注意：更新时会出现错误，aarch64版本Ubuntu apt-get更换阿里源后更新失败，这是由于ARM架构软件源与linux X86源不同，所以apt-get update所下载的索引出错无法解析； 原文地址：Failed to fetch ubuntu bionic arm64 packages
根据文章中的回复，修改 /etc/apt/source.list，在链接最后加上-ports
sudo vi /etc/apt/sources.list
安装步骤：
01- 更新源
sudo apt update02- 修改etc/apt/sources.list.d/nvidia-l4t-apt-source.list
sudo apt install nvidia-jetpack出错 E: Unable to locate package nvidia-jetpack
打开路径/etc/apt/sources.list.d
下面应该有一个文件叫：nvidia-l4t-apt-source.list
如果没有的话就创建这个文件，然后写如这两句话
Please edit your nvidia-l4t-apt-source.list to r35.1:
-
deb https://repo.download.nvidia.com/jetson/common r35.1 main
-
deb https://repo.download.nvidia.com/jetson/t234 r35.1 main
之后执行安装指令
-
sudo apt update
-
sudo apt dist-upgrade
-
sudo reboot
-
sudo apt install nvidia-jetpack
1. How to Install JetPack
Depending on your Jetson device, there are multiple ways to install JetPack.
1.1. SD Card Image
For NVIDIA Jetson Xavier NX developer kit users, the simplest JetPack installation method is to follow the steps at the Getting Started web page to download and write an image to your microSD card, then use it to boot the developer kit.
Note:
If you have not previously run a JetPack 5.x release on your Jetson Xavier NX Developer kit, you must first update its QSPI before using this JetPack 5.0.2 SD Card image. You can download an updated QSPI image from JetPack 5.0.2 page, and then follow the QSPI update instructions from the Jetson Linux Developer Guide.
1.2. NVIDIA SDK Manager
NVIDIA SDK Manager supports JetPack installation on these Jetson products:
NVIDIA Jetson AGX Orin Developer Kit
NVIDIA Jetson AGX Xavier Developer Kit
NVIDIA Jetson Xavier NX Developer Kit
NVIDIA Jetson AGX Orin 32GB module on a Jetson AGX Orin Developer Kit carrier board
NVIDIA Jetson AGX Xavier series modules on a Jetson AGX Xavier Developer Kit carrier board
NVIDIA Jetson Xavier NX modules on a Jetson Xavier NX Developer Kit carrier board
A Linux host computer running Ubuntu Linux x64 version 20.04 or 18.04 is required to run SDK Manager. Detailed instructions can be found here:
1.3. Package Management Tool
NVIDIA offers JetPack components as Debian packages. The Debian package management server can be accessed at Index.
1.3.1. Install JetPack Components on Jetson Linux
Assuming your Jetson developer kit has been flashed with and is running L4T 35.1, the following commands will install all other JetPack components that correspond to your version of Jetson Linux L4T:
sudo apt updatesudo apt install nvidia-jetpackTo view individual Debian packages which are part of nvidia-jetpack metapackage, enter the command:
sudo apt show nvidia-jetpackRefer to the NVIDIA Jetson Linux Developer Guide for details about L4T specific Debian packages.
If disk space is limited, use these commands:
sudo apt updateapt depends nvidia-jetpack | awk '{print $2}' | xargs -I {} sudo apt install -y {}1.3.2. Upgrade JetPack
To upgrade from JetPack 5.0/5.0.1 Developer Preview, first edit etc/apt/sources.list.d/nvidia-l4t-apt-source.list to point to the 35.1 repo (just change the version to r35.1 in both lines). Next, use the following commands, then physically reboot the system:
sudo apt updatesudo apt dist-upgradesudo apt install --fix-broken -o Dpkg::Options::="--force-overwrite"The last line is required because in the JetPack 5.0.2 release, the cuda-nvprof-11-4 package was renamed.
1.4. List of JetPack Debian Packages
The following is a list of Debian update packages for JetPack components for Jetson devices as of the time of JetPack 5.0.2. Please refer to the Jetson Linux Developer Guide for the list of Jetson Linux Debian packages.
Component Group Packages CUDA Toolkit for L4T
cuda-11-4
cuda-cccl-11-4
cuda-command-line-tools-11-4
cuda-compiler-11-4
cuda-cudart-11-4
cuda-cudart-dev-11-4
cuda-cuobjdump-11-4
cuda-cupti-11-4
cuda-cupti-dev-11-4
cuda-cuxxfilt-11-4
cuda-documentation-11-4
cuda-driver-dev-11-4
cuda-gdb-11-4
cuda-libraries-11-4
cuda-libraries-dev-11-4
cuda-nvcc-11-4
cuda-nvdisasm-11-4
cuda-nvml-dev-11-4
cuda-nvprune-11-4
cuda-nvrtc-11-4
cuda-nvrtc-dev-11-4
cuda-nvtx-11-4
cuda-profiler-api-11-4
cuda-runtime-11-4
cuda-samples-11-4
cuda-sanitizer-11-4
cuda-toolkit-11-4
cuda-toolkit-11-4-config-common
cuda-toolkit-11-config-common
cuda-toolkit-config-common
cuda-tools-11-4
cuda-visual-tools-11-4
libcublas-11-4
libcublas-dev-11-4
libcudla-11-4
libcudla-dev-11-4
libcufft-11-4
libcufft-dev-11-4
libcurand-11-4
libcurand-dev-11-4
libcusolver-11-4
libcusolver-dev-11-4
libcusparse-11-4
libcusparse-dev-11-4
libnpp-11-4
libnpp-dev-11-4
cuDNN
libcudnn8
libcudnn8-dev
libcudnn8-samples TensorRT
graphsurgeon-tf
libnvinfer-bin
libnvinfer-dev
libnvinfer-plugin-dev
libnvinfer-plugin8
libnvinfer-samples
libnvinfer8
libnvonnxparsers-dev
libnvonnxparsers8
libnvparsers-dev
libnvparsers8
python3-libnvinfer
python3-libnvinfer-dev
tensorrt
uff-converter-tf
OpenCV
libopencv-dev
libopencv
libopencv-python
libopencv-samples
opencv-licenses
VPI
libnvvpi2
python3.8-vpi2
python3.9-vpi2
vpi2-demos
vpi2-dev
vpi2-samples
NVIDIA container runtime with Docker integration
libnvidia-container-tools
libnvidia-container0
libnvidia-container1
nvidia-container-runtime
nvidia-container-toolkit
nvidia-docker2
Multimedia API
nvidia-l4t-jetson-multimedia-api
NVSCI
nvidia-lt-nvsci
The following is a list of meta-packages that are available to easily install on Jetson. At a higher level, the nvidia-jetpack meta-package includes nvidia-jetpack-runtime meta-package and nvidia-jetpack-dev meta-package. nvidia-jetpack-runtime includes runtime only parts of JetPack components and does not include samples, documentation, etc. Meanwhile, the nvidia-jetpack-dev meta-package includes everything required for development.
You can install either the higher level meta-packages using apt install, OR install individual component meta-packages depending on your requirements. These meta-packages can be installed either on top of Jetson Linux, or in a container running on Jetson Linux.
2.3 更新/升级 Jetpack5.02
3. rootfs eMMC to SSD
按照上述流程刷机之后就可以启动了，但是启动的是在eMMC上所以需要进行rootOnNVMe的流程，以将 rootfs 指向安装在 /dev/nvme0（M.2 Key M 插槽）上的 SSDhttps://github.com/jetsonhacks/rootOnNVMe
https://github.com/jetsonhacks/rootOnNVMe/issues/21
https://github.com/jetsonhacks/rootOnNVMe/issues/21
3.1 安装SSD
3.2 拷贝eMMC系统至SSD
SSD 128GB nvmen1p1 partrial/ copy sysytem fielshttps://github.com/jetsonhacks/rootOnNVMe
注意：This procedure should be done on a fresh install of the SD card using JetPack 4.3+. Install the SSD into the M.2 Key M slot of the Jetson, and format it gpt, ext4, and setup a partition (p1). The AGX Xavier uses eMMC, the Xavier NX uses a SD card in the boot sequence.
01- Next, copy the rootfs of the eMMC/SD card to the SSD
$ ./copy-rootfs-ssd.sh02- Then, setup the service. This will copy the .service file to the correct location, and install a startup script to set the rootfs to the SSD.
$ ./setup-service.sh03- After setting up the service, reboot for the changes to take effect.
Boot Notes
These script changes the rootfs to the SSD after the kernel image is loaded from the eMMC/SD card. For the Xavier NX, you will still need to have the SD card installed for booting. As of this writing, the default configuration of the Jetson NX does not allow direct booting from the NVMe.
重启之后查看系统资源管理器Jtop，查看磁盘容量：
4. ROS /Realsense Lib install
ROS 安装教程参考
https://blog.csdn.net/qq_44339029/article/details/120579608
Realsense lib编译安装教程
【精华】ROS学习（二）：Realsense ROS驱动安装_LeeZhao@的博客-CSDN博客Realsense ROS驱动安装
评论（0）
您还未登录，请登录后发表或查看评论