Industry's First Deterministic Network-Based AI Agent Platform AXP Launched in Nanjing
Zixingshan Laboratory and partners officially launched AXP, the industry's first deterministic network-based AI agent interconnection platform, at the 10th Future Network Development Conference on October 16. The platform supports millions of registered agent accesses and tens of thousands of simultaneous online operations. This launch marks a critical technical breakthrough in China's AI internet infrastructure, providing robust cross-domain collaboration and scalable application support for artificial intelligence. By leveraging deterministic networking, it significantly enhances the reliability and execution speed of distributed multi-agent systems. AXP achieves deep integration of computing and networking, unified heterogeneous agent protocol translation, second-level fault perception switching, and sub-hundred-millisecond semantic addressing responses. Utilizing the National Future Network Experiment Facility (CENI), it supports over 4,000 kilometers of cross-domain high-precision agent addressing, reducing end-to-end task execution latency by over 20%.
## BACKGROUND
Deterministic networking provides guaranteed bandwidth and low delay within a network domain, making it crucial for industrial-grade applications requiring high reliability. The National Future Network Experiment Facility (CENI) is a major national sci-tech infrastructure project in China designed to provide an open, efficient, and low-cost experimental verification environment for future network architectures.