Unitree AGX Module for B2/B2-W
Genuine Unitree NVIDIA Jetson AGX module for B2/B2-W — single AGX Orin delivering 275 TOPS of CUDA compute, mounting externally on the robot's rear for AI, vision and autonomy workloads.
- Unitree B2
- Unitree B2-W
About the Unitree AGX Module for B2/B2-W
The Unitree AGX module for B2/B2-W is the external AI compute upgrade for the B2 industrial quadruped and the wheeled B2-W: one or two NVIDIA Jetson AGX Orin computing board assemblies in a single module that mounts on the robot's rear and turns a CPU-only machine into a CUDA-equipped autonomous platform. It comes in two versions — the 275 TOPS module built around a single Jetson AGX Orin, and the 550 TOPS module that pairs two of them for the heaviest perception stacks.
Inside every B2, an Intel Core i5 runs the platform and an Intel Core i7 is reserved for your code — but neither is a GPU. When a stack outgrows CPU compute — multi-camera perception, learned navigation, VLA-class models, real-time sensor fusion — the Unitree AGX module for B2/B2-W adds the missing layer without opening the robot: it mounts on the rear exterior, the internal computers keep their roles, and it comes back off as easily when a robot changes jobs. It buys its place on a fleet the way a spare battery does — by being the difference between a workload that fits and one that doesn't.
Choose the 275 TOPS version for demanding-but-standard autonomy and machine-vision workloads, or the 550 TOPS dual version when several large models must run at once. Both versions fit the legged B2 and the wheeled B2-W, and the Unitree AGX module for B2/B2-W arrives fully assembled — one module, one mount, no integration project.
What the Unitree AGX Module for B2/B2-W Adds to a Robot That Already Has an i7
Every B2 and B2-W leaves the factory with two computers: a Core i5 running the robot's own platform functions and a Core i7 reserved for user development. That pair covers a great deal, but it is CPU compute — there is no CUDA GPU in the standard machine. The Unitree AGX module for B2/B2-W closes exactly that gap: one or two NVIDIA Jetson AGX Orin assemblies, mounted on the robot's rear, adding 275 or 550 TOPS of GPU-accelerated AI performance for camera-heavy perception, deep-learning inference and multi-sensor fusion.
Because the module lives outside the body, nothing inside is disturbed: the internal computers keep their roles, and the GPU layer is added on top. When a robot is reassigned, or the module needs to move to another unit in the fleet, an external mount makes that a hardware swap rather than a teardown.
275 TOPS or 550 TOPS: Which Version Should You Choose?
Both versions of the Unitree AGX module for B2/B2-W mount the same way and fit the same robots — the difference is how much CUDA compute is onboard. The 275 TOPS module carries a single Jetson AGX Orin assembly; the 550 TOPS module carries two, doubling the GPU budget for stacks that run several large models at once.
- Choose the 275 TOPS module for demanding-but-standard autonomy: multi-camera perception, SLAM, learned navigation and edge inference all fit comfortably on a single AGX Orin.
- Choose the 550 TOPS dual module when the robot must run heavyweight pipelines concurrently — VLA-class models, real-time fusion across several cameras and lidars, or perception shared with a second heavy application.
- Size for the model you plan to run in twelve months, not the one you run today — the module swaps between robots, but an undersized module is the wrong saving.
The AGX Module Is Not the Orin NX Option
The B2 family offers several compute paths, and they are not interchangeable. The Intel Core i7 computing board is the internal x86 user-development computer. The NVIDIA Jetson Orin computing board is the internal GPU option — 100 TOPS inside the body, with its own page. The Unitree AGX module for B2/B2-W is the external path: 275 or 550 TOPS of AGX Orin compute on the robot's rear, for workloads beyond what an internal slot can hold. The internal boards and the AGX module are complements, not competitors — heavy fleets often run both.
One External Module Family for the B2 and the B2-W
The legged B2 and the wheeled B2-W share one compute architecture, and the Unitree AGX module for B2/B2-W fits both: the module mounts on the rear exterior of either robot, with no internal disassembly. A mixed fleet of B2 and B2-W units can share the same module pool, and a module bought for one robot moves to the other when the workload does.
The Unitree AGX module for B2/B2-W is not a cross-model part, though. Unitree's other platforms — the Go2, G1 and R1 families — take their own compute options, and the humanoid lines carry AGX modules under their own model numbers. This page's module is the B2/B2-W version.
Specifications
| 275 TOPS | 550 TOPS | |
|---|---|---|
| Module | 1 × NVIDIA Jetson AGX Orin | 2 × NVIDIA Jetson AGX Orin (dual assembly) |
| AI performance | 275 TOPS | 550 TOPS (2 × 275 TOPS) |
| CPU | 12-core Arm Cortex-A78AE (NVIDIA module spec) | 2 × 12-core Arm Cortex-A78AE (NVIDIA module spec) |
| GPU | 2048-core NVIDIA Ampere with 64 Tensor Cores (NVIDIA module spec) | 2 × 2048-core NVIDIA Ampere with 64 Tensor Cores (NVIDIA module spec) |
| Memory | 64GB 256-bit LPDDR5 (NVIDIA module spec) | 2 × 64GB 256-bit LPDDR5 — 128GB total (NVIDIA module spec) |
| Mounting | External — mounts on the robot's rear | External — mounts on the robot's rear |
| Fits | Unitree B2 · Unitree B2-W | Unitree B2 · Unitree B2-W |
| In the box | 1 × Unitree AGX Module for B2/B2-W — 275 TOPS (single AGX Orin assembly) | 1 × Unitree AGX Module for B2/B2-W — 550 TOPS (dual: 2 × Jetson AGX Orin computing board assemblies) |
Shipping, Quality & Module Care
The module contains no lithium cells, so it ships as standard express freight rather than battery-class cargo — worldwide via DHL or FedEx, with the transit time confirmed in your quote.
Every Unitree AGX module for B2/B2-W is 100% quality-inspected before it ships and packed as protected electronic cargo, so it arrives in perfect working order. From there it is sensitive electronics like any other: dry hands, gentle with the connectors, and keep the module dry once it is on the robot.
Last updated: October 2026.
Frequently Asked Questions
Can't find what you're looking for? Contact our
