RDK X5 AI Board 4 GB 10 TOPS for ROS and AI D-Robotics

The D-Robotics RDK X5 is a powerful AI and edge computing board, specifically designed for robotics, ROS development, and AI applications. It is based on the Rockchip RK3588 octa-core processor with an integrated 10 TOPS Neural Processing Unit (NPU) and combines high computing power, energy efficiency, and extensive interfaces in a compact form factor. With 4 GB LPDDR4 RAM and microSD storage support, the board provides sufficient performance for ROS 2, computer vision projects, and autonomous systems that need to execute AI models locally. The RDK X5 runs on Ubuntu 22.04 LTS and supports ROS 1 / ROS 2 right from the start – no manual installation necessary. Unlike a Raspberry Pi 5, the focus here is not on hobby GPIO applications but on edge AI computing power. The integrated NPU runs neural networks directly on the device, reducing latency, improving privacy, and enabling real-time responses. The board offers Gigabit Ethernet with PoE, Wi-Fi 6, Bluetooth 5.4, HDMI output (1080p60), two MIPI-CSI camera inputs, MIPI-DSI display interface, CAN-FD bus, four USB 3.0 ports, and a comprehensive GPIO header with 28 pins. Thus, the RDK X5 is ideal for research, education, autonomous robots, image processing, and edge AI gateways.

Why RDK X5 instead of Raspberry Pi 5?

10 TOPS AI performance instead of 0.5 TOPS
Ubuntu 22.04 + ROS 2 Humble readily usable
Power-over-Ethernet, Wi-Fi 6 & Bluetooth 5.4 integrated
Ideal for AI applications and robotics projects
Designed for research, teaching, and industrial use

Recommended Accessories:

5V 3000mA USB Power Supply AC/DC Adapter

Features:

Octa-Core Cortex-A55 CPU (1.5 GHz) with 4 GB LPDDR4 RAM for high computing power
10-TOPS NPU and 32-GFLOPS GPU for AI acceleration
Stable wireless connection via Wi-Fi 6 and Bluetooth 5.4
Four USB 3.0 ports, HDMI, MIPI-CSI/DSI interfaces, and 28 GPIOs for versatile expansions
Dual MIPI-CSI inputs for stereo vision and precise image recognition
CAN FD and Gigabit Ethernet with PoE for easy integration into industrial networks
Flash-Connect paradigm: one USB-C cable for programming, data transfer, and debugging
Operating System: Ubuntu 22.04 LTS + ROS 2 Humble (compatible with ROS 1)

Application Examples:

A deep learning line follower that uses onboard AI to autonomously detect and follow lanes.
Visual-inertial odometry for precise position estimation in indoor robots.
Skeleton and keypoint detection for gesture or person tracking.
Object recognition and gripping applications with robotic arms.
Cartographic SLAM solutions (e.g., Lidar-SLAM) for autonomous navigation.

Documentation:

Quick Start Wiki

Technical Specifications:

Manufacturer: DFRobot / D-Robotics
Manufacturer SKU: DFR1238
Processor: RK3588 (8 × Cortex-A55 @ 1.5 GHz)
GPU: 32 GFLOPS
AI BPU: 10 TOPS
RAM: 4 GB LPDDR4
Storage: microSD Card Slot
Network: 1 × GbE with PoE, Wi-Fi 6, Bluetooth 5.4
Interfaces: HDMI, 4 × USB 3.0, USB-C Device, 2 × MIPI CSI, 1 × MIPI DSI, CAN-FD, GPIO 28 Pin
Operating System: Ubuntu 22.04 LTS + ROS 2 Humble
Supply Voltage: 5 V DC / 5 A
Dimensions: 90 × 57 × 19 mm
Weight: 127g