Converge Digest

Cisco Doubles AI Infrastructure Target as Networking Sales Surge

Cisco closed fiscal 2025 with accelerating AI infrastructure momentum, booking more than $2 billion in AI orders from webscale customers—over twice its original target—driven by hyperscaler deployments of large language model (LLM) training clusters. In Q4 alone, AI orders exceeded $800 million, marking the fourth consecutive quarter of triple-digit year-over-year order growth from four of the top six webscale customers. The majority of these AI orders (over two-thirds) were for systems, with the remainder in optics, reflecting demand for Cisco’s Series 8K, Silicon One, and optical platforms integrated into high-speed AI back-end networks.

Enterprise AI demand also grew sharply, with mid-teens growth in data center switching orders for the sixth straight quarter, fueled by private AI cloud deployments. Roughly half of Q4 enterprise AI bookings were for DC switching, with the rest for AI compute. Cisco’s Nexus switches are now fully integrated with NVIDIA’s Spectrum-X architecture, a key differentiator for AI cluster performance. The company also introduced AI Canvas, a generative UI powered by its Deep Network Model, and rolled out Agentic AI capabilities across the product portfolio, including AI-driven operational tools and security enhancements.

Cisco expects AI to drive a fundamental network architecture shift. It projects that Agentic AI workloads—where AI agents interact continuously with applications and end-users—will create sustained high network traffic, unlike intermittent chatbot usage. This will require upgraded switching, routing, wireless, and embedded security to ensure performance and protection. The company sees rising demand for sovereign AI and “Neocloud” architectures in FY 2026, as well as continued adoption of its new C9000 Smart Switches and advanced firewalls for AI-era data centers.

• FY 2025 AI orders from webscale customers: $2B+ (2× original target)

• Q4 AI orders: $800M+, 4th consecutive quarter of triple-digit growth

• Majority of AI orders in systems; remainder in optics

• Enterprise AI demand: mid-teens DC switching growth for 6 consecutive quarters

• Cisco-NVIDIA Spectrum-X integration for AI clusters

• AI Canvas and Agentic AI capabilities embedded across product portfolio

• Network upgrades driven by continuous AI agent traffic models

• FY 2026 AI growth drivers: sovereign AI, Neocloud, Smart Switch adoption

“We delivered a strong close to fiscal 2025, driven by our accelerated innovation and solid execution,” said Chuck Robbins, chair and CEO of Cisco. “The AI infrastructure orders we received from webscale customers in fiscal 2025 were more than double our original target, indicating a massive opportunity ahead as we lead the required architectural shift and build the critical infrastructure needed for the AI era.”

🌐 Why it Matters

Cisco’s AI results point to a sustained hyperscaler investment cycle in networking and optical platforms for AI workloads, as well as rising enterprise demand for private AI infrastructure. By tightly integrating with NVIDIA’s AI networking stack and embedding AI into operational tools, Cisco is positioning itself against both traditional networking rivals and emerging AI-focused infrastructure startups. The shift to Agentic AI could create a persistent, high-volume network traffic environment, driving recurring hardware and software refresh opportunities across global data centers.

Some more observations from Cisco’s Investor Call

AI Orders and Customer Segments

Enterprise and Emerging AI Opportunities

Three Pillars of Cisco AI

  1. AI Training Infrastructure – Cisco 8000 Series, Silicon One, optics, and optical systems for hyperscalers, sovereign, and Neocloud providers.
  2. AI Inference & Enterprise Clouds – NVIDIA partnership to accelerate and de-risk enterprise AI deployments.
  3. AI Network Connectivity – Modernizing, securing, and automating networks to prepare for pervasive Agentic AI workloads.

Key Product and Innovation Highlights

Security Positioning in AI Era

Exit mobile version