USB 3.0 computer-on-module AI-SOC

WiFiPCI Expressgigabit Ethernet

Product overview
The Edge AI SoM computing module delivers scalable AI acceleration and multimedia processing for security, industrial automation and smart IoT at the edge. Built for real-time video analytics, predictive maintenance and autonomous decision-making, it combines a multicore CPU with a dedicated NPU in a compact, low-power System-on-Module form factor.

Core advantages

• Faster development & lower hardware cost: up to 69% reduction in development cycles and ~30% lower hardware expenditure; first vision prototype in ~30 minutes with preconfigured images.
• Plug-and-play prototyping: UVC (USB Video Class) support allows immediate recognition as a standard USB camera without additional drivers, simplifying integration and testing.
• Flexible integration: wide OS and toolchain support (Linux/Android/RTOS) and common I/O make it adaptable to existing industrial and security stacks.

Technical specifications
Processor: Multicore CPU + dedicated NPU (AI accelerator)
AI performance: 2.0–10+ TOPS (scalable by model)
Video: 4K@60fps hardware codec (H.265/H.264)
Memory/storage: LPDDR4x 2/4/8 GB; eMMC 16/32 GB
Vision input: Multi-lane MIPI-CSI (dual/triple camera support)
Connectivity: Gigabit Ethernet, WiFi 6, Bluetooth 5.2, optional 5G/LTE
I/O: PCIe, USB 3.0, UART, I2C, GPIO, CAN Bus
OS: Linux (Ubuntu/Debian), Android, specialized RTOS

Typical applications

• Real-time video analytics for security and perimeter monitoring
• Predictive maintenance and condition monitoring in industrial automation
• AI-enabled IoT gateways and edge vision for smart cities