Jetson SUB Mini PC V2 (Blue) with Jetson Xavier NX, Aluminum Case+Fan, 128GB SSD, WiFi, JetPack 5.0.2

Overview

Jetson SUB Mini PC V2 is a hand-size edge AI mini PC built with the NVIDIA Jetson Xavier NX production module (up to 21 TOPS INT8). It integrates the reComputer J202 carrier board offering rich I/Os (4x USB 3.1, HDMI 2.0, M.2 Key E for WiFi/Bluetooth, M.2 Key M for SSD, RTC, CAN, Raspberry Pi 40-pin GPIO, and more). The system includes an aluminum case, a cooling fan with heatsink, antennas, 128GB M.2 NVMe SSD, WiFi, and a pre-installed JetPack System (JetPack 5.0.2). Designed for the NVIDIA Jetson ecosystem, Jetson SUB Mini PC V2 simplifies AI application development and deployment at the edge.

Key Features

• Fast AI performance: On-device processing up to 21 TOPS AI performance with low power and low latency.
• Hand-size edge AI device: Compact 130mm x120mm x 50mm enclosure including Jetson Xavier NX module, heatsink with cooling fan, case, and power adapter.
• Expandable I/O: USB 3.1 ports (4x), Gigabit Ethernet, HDMI 2.0 and DP, M.2 Key E for WiFi/Bluetooth, M.2 Key M for SSD, RTC, CAN, Raspberry Pi GPIO 40-pin, Control & UART header, MIPI-CSI camera connectors (2x), Fan header, DC power jack.
• Accelerate time-to-market: Pre-installed JetPack 5.0.2 and Linux OS BSP; supports Jetson software and leading AI frameworks and platforms.

Specifications

Comparison among Jetson SUB Mini PCs - Blue

What’s Included

Applications

Ideal for complex AI tasks such as image recognition, object detection, pose estimation, semantic segmentation, video processing, and NLP. Suitable for deployment across retail, robotics, healthcare, agriculture, disaster relief, and more.

Developer Tools

Pre-installed JetPack for fast development and edge AI integration

Jetson software stack begins with NVIDIA JetPack SDK providing a full development environment with CUDA‑X accelerated libraries and NVIDIA technologies. JetPack includes the Jetson Linux Driver Package, drivers, flashing utilities, sample filesystem, toolchains, security features, and OTA update capabilities.

Computer Vision and embedded machine learning

• NVIDIA DeepStream SDK for streaming analytics and multi-sensor video/image understanding on Jetson.
• NVIDIA TAO Toolkit (TensorFlow and PyTorch based) to accelerate model training.
• alwaysAI: build, train, and deploy computer vision applications at the edge on reComputer; see the wiki.
• Edge Impulse for embedded ML at the edge.
• Roboflow tools for dataset-to-model pipelines; see documentation for deploying to Jetson.
• YOLOv5 by Ultralytics for few-shot object detection; see step-by-step wiki tutorials.
• Deci: optimize models on Jetson Nano and Xavier NX; see the webinar.

Speech AI

• NVIDIA Riva GPU-accelerated SDK for building real-time Speech AI applications.

Remote Fleet Management

Enable secure OTA and remote device management with Allxon. Unlock 90 days free trial with code H4U-NMW-CPK.

Robot and ROS Development

• NVIDIA Isaac ROS GEMs accelerate ROS development on NVIDIA hardware. Learn more at NVIDIA Developer Tools.
• Cogniteam Nimbus cloud platform supports NVIDIA Jetson and ISAAC SDK/GEMs out-of-the-box; see the webinar.

Notes

• The top acrylic cover is removable for easy access to add a HAT or Cape on the SBC inside the case without re-screwing the whole enclosure.
• This product includes a power adapter but not a power cord. Please purchase the appropriate cord: EU version or US version.
• Before use, open a terminal and execute:
  1. cd /home/nvidia
  2. sudo./ssd extern.sh
  to make sure the roof expands to take the unallocated 110GB space.

Documents

• NVIDIA Jetson Devices and carrier boards comparision.pdf
• reComputer J202 datasheet.pdf
• For a broader overview, visit the NVIDIA Jetson Series Catalog.

Certification

Details

The Oz4200 single board computer features a range of interfaces, including UART, CAN, MIPI-CSI camera connectors, 40-pin expansion headers for GPIO, I2C, and UART, as well as USB3.1 ports, POE gigabit Ethernet, DP, HDMI, LED light, and DC power jack.

Jetson SUB Mini PC with M.2 KEY E and M, RTC Socket, USB ports, power input, and seed studio branding.

NVIDIA Jetson Software: TAO Applications (Train, Adapt, Optimize) for Omnivisual AI Systems. Includes NVIDIA NGC Synthetic Pre-Trained Ecosystem Data Models for Al, Software, Hardware, Machine Vision Cameras, and System Services.

Jetson Nano Xavier NX, 4 x M2.5 screws and standoffs for mounting. Includes: 4 x M2.5 * 6mm screws, 1 x XM2.5*12+3.2mm standoff, and 1 x 4xM2.5*22+3.2mm standoff.