TLDR: Cisco Systems has launched its Silicon One P200 chip and 8223 router, designed to revolutionize AI infrastructure by connecting geographically dispersed AI data centers into a single, unified supercomputer. This innovation offers 51.2 Tbps performance, a 65% reduction in power consumption, and extends connectivity up to 1,000 kilometers, addressing critical challenges in AI scalability and power efficiency.
San Jose, CA – October 8, 2025 – Cisco Systems (NASDAQ: CSCO) today announced a significant advancement in artificial intelligence (AI) infrastructure with the introduction of its groundbreaking Silicon One P200 chip and the accompanying Cisco 8223 router. This powerful new combination is engineered to seamlessly connect AI data centers that are geographically dispersed, enabling them to function as a single, unified supercomputer. The move is poised to redefine the architecture of AI infrastructure, marking a pivotal moment for the rapidly expanding AI industry.
The new technology directly addresses critical challenges in the scalability, power efficiency, and immense computational demands of next-generation AI workloads. As AI models continue to grow exponentially in size and complexity, the ability to distribute training and inference across multiple data centers becomes paramount, particularly as companies seek locations with abundant and affordable power. Cisco’s solution aims to overcome the limitations of traditional networking, promising unprecedented levels of performance and efficiency for hyperscalers and enterprises building their AI foundations.
At the heart of this innovation is the Cisco Silicon One P200, a cutting-edge deep-buffer routing chip. It delivers an astounding 51.2 Terabits per second (Tbps) of routing performance. Remarkably, this single chip consolidates functionality that previously required 92 separate chips, resulting in a significant 65% reduction in power consumption compared to existing comparable routers. This efficiency is crucial for the energy-intensive nature of AI infrastructure, where power consumption has become a primary constraint on growth.
Powering the new Cisco 8223 routing system, the P200 enables this 3-rack-unit (3RU) fixed Ethernet router to provide 51.2 Tbps of capacity with 64 ports of 800G connectivity. The 8223 is capable of processing over 20 billion packets per second and performing over 430 billion lookups per second. A key differentiator is its support for coherent optics, which allows for long-distance data center interconnect (DCI) and metro applications, extending connectivity up to 1,000 kilometers.
Also Read:
- ZTE Launches Advanced Co-Sight Super Agent Globally, Driving AI Ecosystem Expansion
- Simplismart and Yotta Forge Alliance for Advanced AI Infrastructure in India
Cisco views the launch of the Silicon One P200 chip and the 8223 router as a foundational leap in AI history. By directly tackling the most pressing networking challenges of the AI era—specifically, connecting massive, distributed AI data centers with unparalleled performance, power efficiency, and security—Cisco has positioned itself as a critical enabler of future AI innovation. This development is considered not merely an incremental improvement but a strategic architectural shift that will empower the next generation of AI models and applications.
