Unitree - Go2 Docking Station Computing Module 100 TOPS
- Designed for Unitree Go2 Docking Station – Dedicated AI computing module engineered specifically for seamless integration with the Unitree Go2 docking station.
- 100 TOPS AI Computing Performance – Delivers up to 100 TOPS of edge AI processing power for advanced robotics, computer vision, deep learning, and autonomous applications.
- Wide DC Power Input – Supports a 16V–60V DC input range, providing flexible compatibility with a variety of power systems.
- High-Speed USB Connectivity – Equipped with 1 × USB 3.0 Type-A and 2 × USB 3.0 Type-C ports for connecting cameras, sensors, storage devices, and other peripherals.
- Dual Gigabit Ethernet Ports – Features 2 × RJ45 Gigabit Ethernet ports for high-speed networking, data communication, and system integration.
- Industrial Expansion Interfaces – Includes 1 × GH1.25-4PIN Fast Ethernet interface and 1 × M8 aviation connector for secure industrial-grade connectivity.
- Optimized for AI & Edge Computing – Ideal for autonomous navigation, real-time AI inference, object detection, SLAM, robotics research, and industrial automation.
- Compact & Industrial-Grade Design – Built with durable components to ensure reliable operation in robotics labs, industrial environments, and commercial deployments.
- Easy Integration with Go2 Systems – Designed for straightforward installation into compatible Unitree Go2 docking stations and robotic workflows.
- Go2 Compatible Only – Specifically designed for the Unitree Go2 platform. Please verify compatibility before purchase if using custom hardware or other robot models.
- Trusted dealerDirect manufacturer partnership
- Pan-India deliveryInsured shipping, all states
- Post-sale supportTraining + service included
What makes it work
Unitree Go2 Docking Station (100 TOPS)
The Unitree Go2 Docking Station (100 TOPS) is a genuine NVIDIA Orin NX AI computing module built for advanced research, SLAM, and vision-language workloads beyond the entry-level 40 TOPS tier.
NVIDIA Orin NX 16GB
Multi-Port Industrial I/O
Go2-Only Fit
Genuine Unitree Part
Full Specifications
| Specification | Details |
|---|---|
| AI Performance | 100 TOPS |
| Compute Module | NVIDIA Jetson Orin NX, 16GB LPDDR5 memory |
| Processor | 8-core ARM Cortex-A78AE @ 2.0GHz |
| GPU | 1024-core NVIDIA Ampere architecture GPU @ 918MHz, with 32 Tensor Cores |
| Power Input | 16V-60V DC, wide input range |
| USB Ports | 1× USB 3.0 Type-A, 2× USB 3.0 Type-C |
| Network | 2× Gigabit Ethernet (RJ45), 1× Fast Ethernet (GH1.25-4PIN) |
| Industrial Connector | 1× M8 aviation plug interface |
| Software Environment | Ubuntu 20.04/22.04, ROS 2, NVIDIA JetPack SDK, CUDA, TensorRT |
| xBoom SKU | xboom-144843 |
| Specification | Details |
|---|---|
| Compatible Platform | Unitree Go2 — xBoom’s own listing recommends confirming compatibility before purchase if using custom hardware or another robot model |
| Sibling Configuration | A lower-cost 40 TOPS configuration (NVIDIA Orin Nano, 8GB) is also available, sold separately, better suited to educational and entry-level development |
| Go2-W Note | A separate Go2-W-specific 100 TOPS docking station module exists as its own distinct product from other resellers. This listing is for the standard Go2 |
| Part of a Larger Ecosystem | One module in Unitree’s official Go2 accessory family, alongside the Charging Pile, 3D Navigation Radar (Mid-360 or Hesai XT16), Depth Camera (D435i), and Servo Arm |
| Typical Use Case | Advanced research, SLAM, vision-language models, and reinforcement learning workloads beyond what the 40 TOPS tier comfortably handles |
When 40 TOPS Isn’t Enough
Unitree Go2 Docking Station: Real Headroom for Advanced Workloads
The Unitree Go2 Docking Station (100 TOPS) steps up from the entry-level Orin Nano to a full NVIDIA Jetson Orin NX — 16GB of LPDDR5 memory, an 8-core ARM Cortex-A78AE processor, and a 1024-core Ampere GPU with 32 Tensor Cores. That’s real headroom for advanced SLAM, vision-language models, and reinforcement-learning workloads that the 40 TOPS tier isn’t built to carry.
The connectivity is identical to the 40 TOPS module — the same USB, Ethernet, and industrial M8 connector options, and the same Ubuntu, ROS 2, and JetPack development environment. So the step up is purely about how much compute headroom a project actually needs, not a different way of working.
100 TOPS AI Performance
More than double the entry-level tier’s compute for demanding workloads.
16GB Orin NX Memory
Enough headroom for larger models and heavier sensor-fusion pipelines.
32 Tensor Cores
Dedicated silicon built specifically to accelerate deep learning workloads.
Same Dev Environment as 40 TOPS
Ubuntu, ROS 2, and JetPack stay consistent across both compute tiers.
See Every Detail
One Module, Two Compute Tiers
The 40 TOPS and 100 TOPS configurations share the same external module design and the same place in Unitree’s official Go2 accessory family — the difference is entirely in the compute silicon inside, not the housing. This is Unitree’s own accessories catalog, showing “The High Computing Power Module” alongside the Charging Pile, 3D Navigation Radar, depth camera, and servo arm.
Built for Advanced Robotics Development
The Unitree Go2 Docking Station (100 TOPS) is engineered for teams whose workloads genuinely need more than entry-level compute.
🧠
100 TOPS NVIDIA Orin NX
Real headroom for advanced SLAM, vision-language, and RL workloads.
🎮
16GB LPDDR5 Memory
Enough capacity for larger models and heavier sensor-fusion pipelines.
⚙️
1024-Core Ampere GPU
32 dedicated Tensor Cores built specifically for deep learning acceleration.
🔌
Wide 16-60V DC Input
Flexible integration with a range of compatible power sources.
🔗
Rich Multi-Port I/O
USB 3.0, dual Gigabit Ethernet, and an industrial M8 connector in one module.
🤖
Ubuntu + ROS 2 + JetPack Ready
The same proven development environment as the 40 TOPS configuration.
Built for Teams Pushing Past Entry-Level Compute
The Unitree Go2 Docking Station (100 TOPS) suits advanced research and development workloads that genuinely outgrow the 40 TOPS tier.
Advanced SLAM & Mapping
Real compute headroom for dense, high-precision mapping workloads.
Vision-Language Model Research
Enough memory and compute to run larger multimodal models onboard.
Reinforcement Learning Projects
The compute headroom demanding RL training and inference needs.
Real-Time Computer Vision at Scale
Handle multiple camera streams and heavier vision models at once.
Advanced Robotics R&D
A genuine, standard configuration keeps advanced research reproducible.
Industrial Automation Integration
Industrial connectors support integration into larger automated systems.
See the Orin NX vs. Orin Nano Compared
A dedicated explainer on this exact compute tier, an install walkthrough, and the Go2 platform it upgrades.
Frequently Asked Questions
What computing hardware is actually inside this module?
How is this different from the 40 TOPS version?
What software does it come with?
Does this work with the Unitree Go2-W?
What ports and connectors does it have?
Is pricing listed online?
The full sheet
Related products
-
·LND·00
Unitree Go2 Air- Robot Dog (Without Controller)Original price was: ₹292,373.₹249,153Current price is: ₹249,153. (Excl. GST)Add to cart -
·LND·01
-
·XBM·02
-
·LND·03
There are no reviews yet.