Cisco Unified Edge platform focuses on real-time AI at the edge

Cisco has introduced the Cisco Unified Edge, a cutting-edge integrated edge computing platform tailored to handle real-time AI inferencing and agentic workloads directly at the data source. The unveiling took place at the Partner Summit in San Diego, where Cisco positioned the system as a crucial component for large-scale AI implementations within enterprises.

The announcement underscores the increasing importance of edge compute architectures in managing the exponential growth of AI workloads, particularly in diverse sectors such as industrial, retail, healthcare, and telecommunications.

Empowering Data Center-Scale AI at the Edge

With the evolution of AI applications from centralized model training to instantaneous inference, organizations are realizing that traditional data center setups are struggling to keep pace. Cisco notes that AI agents can generate network traffic up to 25 times higher than standard chatbot workloads, emphasizing the necessity to bring computing and decision-making closer to the data source.

“The current infrastructure is inadequate to support AI at scale,” stated Jeetu Patel, President and Chief Product Officer at Cisco. “As AI agents and experiences become more prevalent, they will naturally gravitate towards where customer interactions and decisions occur – be it the branch office, retail outlet, factory floor, stadium, and beyond. This is where computing needs to be situated. With our Unified Edge, we are simplifying the process of enabling AI in real-world scenarios with adaptable, secure systems that are easy to deploy, manage, and scale as demand increases.”

The Unified Edge platform integrates computing, networking, storage, and security into a single modular system. It accommodates both CPU and GPU configurations and seamlessly interfaces with Cisco’s SD-WAN, Intersight cloud-based management, and monitoring platforms like Splunk and ThousandEyes. The system facilitates zero-touch deployment and enterprise-wide operations without the need for on-site specialists.

Enhanced Security and Flexibility for Distributed AI

Cisco has embedded security features throughout the system, employing a zero-trust architecture, tamper-proof hardware elements, and consistent policy enforcement from device to application. This becomes crucial as the attack surface expands with the migration of AI workloads closer to the physical environment.

A diverse partner ecosystem, including Intel, Verizon, Rockwell Automation, and World Wide Technology, played a pivotal role in shaping the platform. Intel emphasized the significance of its Xeon 6 SoC as a cornerstone for high-throughput, low-latency edge applications, while Verizon and Rockwell highlighted the escalating competitive necessity for real-time decision-making in manufacturing and service operations.

The Cisco Unified Edge platforms are currently available for ordering, with Cisco aiming for general availability by the end of the year.

Cisco