Zum Inhalt springen

WSAL23035 – JETSON-TX2-NX-DEV-KIT/en

Aus JW Tech Wiki

Documentation notice: This technical documentation was independently rewritten from an archived supplier source. Check technical details and commands against the product version and manufacturer documentation before use.

Summary

A concise technical introduction to the Waveshare Jetson TX2 NX development kit: module specs, interfaces, system programming and basic commands.

Related products at JW Tech

Technical overview

  • The kit includes a Jetson TX2 NX module with 40-pin GPIO headers, a Micro-USB data port, a gigabit Ethernet port (10/100/1000Base-T), four USB 3.0 ports, HDMI and DisplayPort connectors, a DC power input, two MIPI CSI camera ports, and a fan header.
  • Module specifications: GPU with 256 CUDA cores (Pascal architecture); CPU combining dual-core Denver 2 and quad-core Cortex-A57; 4 GB LPDDR4 (128-bit, 51.2 GB/s); 16 GB eMMC 5.1 plus M.2 NVMe SSD interface.
  • Video encoding supports up to 3×4K30 or 4×1080p60 (H.265) and 1×4K60 or 3×4K30 (H.264); decoding up to 4×4K30 or 7×1080p60.
  • Camera support handles up to 5 physical cameras (up to 12 via virtual channels) through a 12-channel MIPI CSI-2 with D-PHY 1.2 (up to 30 Gbps).
  • System programming requires an Ubuntu 18.04 host using the SDK Manager; the kit has no SD card slot, so flashing is done via eMMC/NVMe.
  • To program, the module is put into recovery mode by shorting the FC REC and GND pins, then connected to the host via a USB data cable and the SDK Manager is launched.
  • After flashing, the system can be moved to boot from an SSD: format with mkfs.ext4, edit /boot/extlinux/extlinux.conf, copy the system with cp -ax, reboot, and verify with df -h.
  • Basic Linux commands such as sudo, ls, chmod, mkdir, cp, mv, rm, reboot, shutdown, pwd, head, tail, df, tar, apt, plus network commands ifconfig and hostname are documented.
  • For AI development, environments for PIP, CUDA, TensorFlow, PyTorch and YOLOv4 are provided; additionally a “Hello AI World” example uses DetectNet for real-time camera detection.
  • Hardware control covers GPIO (turning on an LED) and I2C (i2cdetect, i2cdump, i2cset, i2cget) with examples for register access.

Commands and examples – verify before use:  

sudo apt install ./sdkmanager_[version]-[build#]_amd64.deb
ls /dev/nvme*
sudo mkfs.ext4 /dev/nvme0n1
sudo vi /boot/extlinux/extlinux.conf
sudo mount /dev/nvme0n1 /mnt
sudo cp -ax / /mnt
sudo umount /mnt/
sudo reboot
df -h
sudo apt-get update
sudo apt-get install -y i2c-tools
sudo apt-get install -y python3-smbus
apt-cache policy i2c-tools
sudo i2cdetect -y -r -a 0
sudo i2cdump -y 0 0x68
sudo i2cset -y 0 0x68 0x90 0x55
sudo i2cget -y 0 0x68 0x90

Notes and version constraints

  • The source is untrusted supplier material; all statements should be verified before deployment.
  • Version-sensitive steps: The exact SDK Manager .deb filename must be replaced with the actually downloaded name.
  • Recovery mode requires precise shorting of the FC REC/GND pins; incorrect handling may damage the module.
  • The prescribed host OS is Ubuntu 18.04; other versions are not confirmed.
  • When booting from SSD, the label must be changed correctly (LABEL NX instead of primary); errors here will prevent booting.
  • INSUFFICIENT SPACE: If the eMMC system lacks space for SDK installation, switch to SSD boot.
  • The source mentions both “Jetson Nano” and “TX2 NX”; the correct product designation should be verified before publishing.

Source and revision record

Abgerufen von „https://wiki.jw-tech.eu/index.php?title=WSAL23035_–_JETSON-TX2-NX-DEV-KIT/en&oldid=2390“