In a major step toward redefining AI connectivity, Cisco Systems has introduced its new P200 networking chip, designed to link AI data centres spread across the globe. The chip forms the heart of a new routing device aimed at enabling data centres to collaborate seamlessly, acting as one giant AI-powered computer.
The move positions Cisco in direct competition with Broadcom, a major player in the networking chip market. Early adopters of Cisco’s P200 chip include cloud giants Microsoft and Alibaba, reflecting strong industry confidence in the new technology.
Powering AI at Scale
Modern AI models demand massive computing resources. Companies like Nvidia are connecting tens of thousands of GPUs within single data centres to train large AI models. Cisco’s innovation extends this concept further — by linking multiple data centres located miles apart to operate as a unified system.
Martin Lund, Executive Vice President of Cisco’s Common Hardware Group, told Reuters, “The training job is so large, I need multiple data centres to connect together — and they can be 1,000 miles apart.”
This approach not only enhances AI scalability but also helps address the energy constraints of modern computing. AI data centres consume immense electricity, driving companies to set them up in power-rich regions such as Texas and Louisiana.
A Leap in Efficiency and Design
Cisco claims the P200 chip replaces 92 separate chips with just one, drastically improving performance and reducing energy consumption. The router built around this chip reportedly uses 65% less power than existing comparable solutions — a crucial advantage as energy costs soar and sustainability becomes a top priority in tech infrastructure.
One of the toughest challenges in linking far-flung data centres is ensuring seamless data synchronization without loss. Cisco’s decades of work in data buffering technology provides a critical edge here, enabling stable and high-speed data transfer across networks.
Industry Impact and Outlook
According to Microsoft’s Azure Networking VP Dave Maltz, the scale of cloud and AI operations today demands “faster networks with more buffering to absorb bursts of data.” Cisco’s P200 innovation directly addresses this need, offering both speed and efficiency.
While Cisco has not disclosed the investment amount or sales projections, the P200 chip signals the company’s strategic entry into the AI infrastructure era, combining decades of networking expertise with next-gen computing needs.
With AI reshaping global industries, Cisco’s P200 could become a pivotal bridge between AI computing power, sustainability, and global scalability.
