Unitree Go2 AI Docking Station
AI compute upgrade for the Unitree Go2 quadruped robot. NVIDIA Jetson Orin module with ROS 2 support, USB 3.0, 100Gb Ethernet, M.2 NVMe. Available in 40 TOPS or 100 TOPS configurations.
Free Shipping, fulfilment within 1 Week
Choose options
Unitree Go2 AI Docking Station
product.price.sale_price
$3,000.00 AUD
Product Description
The Unitree Go2 Docking Station is an AI compute upgrade module designed specifically for the Unitree Go2 quadruped robot, delivering dedicated NVIDIA Jetson processing power for advanced edge intelligence, real-time perception, and deep learning workloads on-robot.
Built on a custom Unitree carrier board, the docking station integrates seamlessly with Go2 EDU platforms. Multiple high-speed interfaces — USB 3.0, USB-C, Gigabit and 100Gb Ethernet, plus M.2 NVMe storage — provide ample bandwidth for sensor fusion, cameras, LiDAR, and external accelerators. The development computer runs Ubuntu 20.04/22.04 with ROS 2 support, NVIDIA JetPack SDK, CUDA, and TensorRT.
Available in two configurations: Orin Nano 8GB (40 TOPS) for educational and entry-level development, and Orin NX 16GB (100 TOPS) for advanced research, SLAM, vision-language models, and reinforcement learning.
Key Specifications
- Orin Nano 8GB: 40 TOPS, 6-core ARM Cortex-A78AE @ 1.5 GHz, 1024-core Ampere GPU, 32 Tensor Cores, 8 GB LPDDR5
- Orin NX 16GB: 100 TOPS, 8-core ARM Cortex-A78AE @ 2.0 GHz, 1024-core Ampere GPU @ 918 MHz, 32 Tensor Cores, 16 GB LPDDR5
- Connectivity: USB 3.0 Type-A, USB 2.0 Type-C, USB 3.0 Type-C, 2× Gigabit Ethernet (RJ45), 100Gb Ethernet, M8 air plug
- Power Input: 16–60 V DC
- Compatible with: Unitree Go2 EDU
- Software: Ubuntu 20.04 / 22.04, ROS 2, NVIDIA JetPack SDK, CUDA, TensorRT, PyTorch, TensorFlow
- Custom Unitree carrier board (use only Unitree-supplied images — do NOT flash third-party BSP)