Multi-modal Multi-agentic Emboddied Chips, AI, 6G Communication Integrated Solutions 

This project aims to develop the world’s first AI-Native Perception Networking Chip integrating computer vision and communication into a unified SoC architecture. By embedding dynamic scheduling, ROI-based perception, and perception-aware networking protocols directly in hardware, the system enables ultra-low-latency, secure, and efficient perception sharing among autonomous vehicles, drones, and industrial robots. The architecture reduces communication delay by an order of magnitude, supporting real-time collaboration in ports, airports, and smart factories. Positioned as the “Qualcomm + NVIDIA” of perception networking, it establishes the core infrastructure for next-generation multi-agent intelligence.