Radxa AICore Dx-M1 is a high-performance edge Ai acceleration computing module based on Deepx's Dx-M1 module. It features high energy efficiency, high-precision computing, and ease of use, specifically optimized for edge computing scenarios and machine vision.

The Radxa AICore Dx-M1 module is currently compatible with various single-board computers (SBCs), such as Radxa Rock 5B/5B+ and other motherboards.

Features

• High Energy Efficiency Design
  Features an energy-efficient design with excellent inference performance and low power consumption, achieving 25 Tops computing power with only 3-5W power consumption.
  
  
• High-Precision Computing
  Utilizes IQ8 (Intelligent Quantization Integer 8) technology, maintaining the same precision level as Gpu's FP32.
  
  
• Ease of Use
  • Ai Frameworks: Supports Onnx, PyTorch, TensorFlow.
  • System Support: Supports Ubuntu, Debian.
  • Standardized Interface: Adopts standard M.2 2280 M Key form factor, compatible with industrial and embedded devices.

Specifications