Skip to main content

AGX

Flash AGX​

Using sdkmanager GUI​

  • Current JetPack Version: 4.3
  • To install older JetPack Versions via the sdkmanager run in cmd:
    $ sdkmanager --archivedversions

Using Docker​

First install the dependencies:

sudo apt install qemu-kvm libvirt-daemon-system libvirt-clients bridge-utils
sudo apt install qemu-user-static
sudo apt install binutils
sudo apt install qemu

Then, run

docker run -it --rm -v /home/$USER/Nvidia/:/home/$USER/Nvidia/ --privileged -v /dev/bus/usb:/dev/bus/usb/ sdkmanager --cli install --logintype devzone --product Jetson --host --targetos Linux --version 4.6.2 --target JETSON_AGX_XAVIER_TARGETS --additionalsdk 'DeepStream 6.0.1' --select 'Jetson OS' --select 'Jetson SDK Components' --flash all

Steps:

  • Choose "3. Manual Setup - Jetson AGX Xavier"
  • OEM-Configuration:
    • Choose "Pre-config"
  • Set username and password
  • Select Storag Device: EMMC (default)
  • Connect to the AGX and login to Ubuntu on AGX.
  • Install the SDK components after logging in.
  • Use IPv4 and the AGX's ssh ip address (you can find it in http://192.168.0.1/info/Login.html).

Troubleshooting:

TODO: The following error occurs while flashing the AGX (via the docker method), but the installation will finish successfully. Fix this error.

error: In file included from ../3rdparty/eigen/Eigen/Core:276:0, │
│ Exit ││error: from ../3rdparty/eigen/Eigen/Dense:1, │
│ ││error: from src/NVX/Render/CUDA-OpenGL/BaseRender3DImpl.cpp:31: │
│ ││error: ../3rdparty/eigen/Eigen/src/Core/Functors.h:932:28: warning: ‘template<class _Operation> class std::binder2nd’ is deprecated │
│ ││[-Wdeprecated-declarations] │
│ ││error: struct functor_traits<std::binder2nd<T> > │
│ ││error: ^~~~~~~~~ │
│ ││error: In file included from /usr/include/c++/7/bits/stl_function.h:1127:0, │
│ ││error: from /usr/include/c++/7/string:48, │
│ ││error: from /usr/include/c++/7/bits/locale_classes.h:40, │
│ ││error: from /usr/include/c++/7/bits/ios_base.h:41, │
│ ││error: from /usr/include/c++/7/ios:42, │
│ ││error: from /usr/include/c++/7/istream:38, │
│ ││error: from /usr/include/c++/7/sstream:38, │
│ ││error: from /usr/include/c++/7/complex:45, │
│ ││error: from ../3rdparty/eigen/Eigen/Core:28, │
│ ││error: from ../3rdparty/eigen/Eigen/Dense:1, │
│ ││error: from src/NVX/Render/CUDA-OpenGL/BaseRender3DImpl.cpp:31: │
│ ││error: /usr/include/c++/7/backward/binders.h:143:11: note: declared here │
│ ││error: class binder2nd │
│ ││error: ^~~~~~~~~ │
│ ││error: In file included from ../3rdparty/eigen/Eigen/Core:276:0, │
│ ││error: from ../3rdparty/eigen/Eigen/Dense:1, │
│ ││error: from src/NVX/Render/CUDA-OpenGL/BaseRender3DImpl.cpp:31:
update-initramfs: Generating /boot/initrd.img-4.9.253-tegra
WARNING: missing /lib/modules/4.9.253-tegra
Ensure all necessary drivers are built into the linux image!
depmod: ERROR: could not open directory /lib/modules/4.9.253-tegra: No such file or directory
depmod: FATAL: could not search modules: No such file or directory
Warning: couldn't identify filesystem type for fsck hook, ignoring.
I: The initramfs will attempt to resume from /dev/zram3
I: (UUID=8c3be5ed-efaf-4e54-ac76-1dd34c94785a)
I: Set the RESUME variable to override this.
depmod: WARNING: could not open /var/tmp/mkinitramfs_7QrmPG/lib/modules/4.9.253-tegra/modules.order: No such file or directory
depmod: WARNING: could not open /var/tmp/mkinitramfs_7QrmPG/lib/modules/4.9.253-tegra/modules.builtin: No such file or directory
/sbin/ldconfig.real: Warning: ignoring configuration file that cannot be opened: /etc/ld.so.conf.d/aarch64-linux-gnu_EGL.conf: No such file or directory
/sbin/ldconfig.real: Warning: ignoring configuration file that cannot be opened: /etc/ld.so.conf.d/aarch64-linux-gnu_GL.conf: No such file or directory
Processing triggers for gnome-icon-theme (3.12.0-3) ...

Install other Dependencies​

  • numpy: sudo apt install python3-numpy, numpy 1.18.5 (do not use pip install! Some newer numpy versions do not work with ROS eloquent!)
  • h5py, hdf5-related stuff: will be installed during Tensorflow installation (see below)
  • Tensorflow: read the doc: jetson installation

SSH​

Prerequisite: You must be connected to vpn-i11-studenten.

DeviceIPSSH Command
AGX-LAN192.168.0.116$ ssh -p 8116 galaxis@137.226.8.100
AGX-WiFi192.168.0.139$ ssh -p 8139 galaxis@137.226.8.100

Performance Modes​

Can be used to switch to another power mode. (e.g. for compiling)

NVPMODEL​

# Power Modes
$ sudo nvpmodel -m <mode ID> # pick mode ID from table below
# Fan Modes
$ sudo nvpmodel -d <mode> # modes are: quiet, cool
Power Mode Table​

https://docs.nvidia.com/jetson/l4t/index.html#page/Tegra%20Linux%20Driver%20Package%20Development%20Guide/power_management_jetson_xavier.html

NVPModel clock configuration for Jetson AGX Xavier 16GB and 32GB

Property

Mode

MAXN

10W

15W

30W

30W

30W

30W

15W *

Power budget

n/a

10W

15W

30W

30W

30W

30W

15W

mode ID

0

1

2

3

4

5

6

7

Online CPU

8

2

4

8

6

4

2

4

CPU maximal frequency (MHz)

2265.6

1200

1200

1200

1450

1780

2100

2188

GPU TPC

4

2

4

4

4

4

4

4

GPU maximal frequency (MHz)

1377

520

670

900

900

900

900

670

DLA cores

2

2

2

2

2

2

2

2

DLA maximal frequency (MHz)

1395.2

550

750

1050

1050

1050

1050

115.2

PVA cores

2

0

1

1

1

1

1

1

PVA maximal frequency (MHz)

1088

0

550

760

760

760

760

115.2

CVNAS maximal frequency (MHz)

1356.8

601.6

716.8

1011.2

1011.2

1011.2

1011.2

115.2

Memory maximal frequency (MHz)

2133

1066

1333

1600

1600

1600

1600

1333

SOC clocks maximal frequency (MHz)

All modes

adsp 300

ape 150

axi_cbb 408

bpmp 896

bpmp_apb 408

display 800

display_hub 400

csi 400

host1x 408

isp 1190.4

nvdec 1190.4

nvenc 1075.2

nvjpg 716.8

pex 500

rce 819.2

sce 729.6

se 1036.8

tsec 1036.8

vi 998.4

vic 1036.8

  •  The default mode is 15W (mode ID 7).

Default mode is intended to improve desktop application performance. PVA and DLA are not in use and run at minimal frequency.

Jetson Clocks​

Hich performance mode.

$ sudo jetson_clocks

Wifi​

Realtek Module​