Cisco CEO Chuck Robbins on Building Infrastructure for the AI Era

Cisco is positioning itself as a key provider of secure, scalable, and high-performance infrastructure tailored for the AI era.

The company focuses on powering AI workloads across data centers, campuses, branches, edges, and interconnects, emphasizing that “there’s no secure AI without Cisco.”

This involves integrating networking, security, observability, and AI-native capabilities deeply into its offerings.

Core Strategies and Focus Areas

Cisco’s approach revolves around several pillars:

• AI-Native Networking and Data Centers — Building robust, lossless Ethernet fabrics optimized for AI/ML workloads, supporting massive scale with low latency, high throughput, and sustainability. They promote Ethernet as the dominant backbone for AI due to its scalability, cost-effectiveness, and ability to handle thousands of GPUs.
• Mass-Scale AI Infrastructure — Solutions for pre-training, large-scale inference, distributed edge clusters, and interconnecting them all.
• Security Fusion — Embedding security directly into the network for end-to-end protection of AI infrastructure, workloads, and usage, including guardrails, threat intelligence, and visibility (e.g., Cisco AI Defense).
• Observability and Operations — Real-time insights across the AI stack, powered by tools like Splunk and ThousandEyes, for optimized performance and digital resilience.
• AI-Enabled Simplification — Native AI in products for automation, predictive analytics, and generative experiences (e.g., Cisco AI Assistant, AI Canvas for AgenticOps).
• Edge and Distributed AI — Bringing compute, networking, and security closer to data sources for low-latency inference.

Key Products and Innovations

Cisco has rolled out hardware and software specifically designed for AI demands:

• Cisco Silicon One — A unified, programmable architecture (e.g., P200 chip) powering efficient, high-bandwidth routing and switching, with significant power savings (up to 65% less than prior generations in some configurations).
• Nexus Series Switches — Including next-gen models like Nexus 9300 with embedded DPUs for efficient traffic steering, and high-performance options like the 9100 series (51.2 Tbps bandwidth) for AI data centers.
• Nexus Hyperfabric AI — A cloud-managed, full-stack solution (compliant with NVIDIA architectures) that simplifies deployment of on-premises AI infrastructure, allowing “bring your own” compute/GPU/storage while delivering high-performance fabrics via Cisco 6000 Series Switches.
• AI PODs and Secure AI Factory — Pre-validated, integrated systems (often with NVIDIA) for rapid AI deployment.
• Optical Interconnects and High-Speed Optics — Leveraging silicon photonics for reliable, high-performance connections within and between data centers.
• Unified Management — Tools like Nexus Dashboard for overseeing fabrics, plus AI-driven NetOps.

Partnerships

Cisco collaborates closely with leaders to accelerate AI adoption:

• NVIDIA — Long-standing partnership for Secure AI Factory, reference architectures, Spectrum-X Ethernet integration, and plug-and-play solutions.
• Others include VAST Data for storage fabrics, ServiceNow for simplified adoption, and ecosystem partners for edge/compute.

Broader Initiatives

Cisco emphasizes responsible AI, with an AI Readiness Index highlighting gaps (e.g., only a minority of organizations have fully AI-capable infrastructure as of recent reports). They’ve launched investments like a $1B global AI fund, agentic AI support, and sovereign AI projects. CEO Chuck Robbins has highlighted Cisco’s role in building the foundational infrastructure—secure networks fused with AI—that connects the world and powers the economy.

Overall, Cisco is transforming from traditional networking to a full-stack AI enabler, helping enterprises (from hyperscalers to branches) overcome challenges like exponential traffic growth, power demands, and security risks while simplifying operations for the agentic AI future.