Cisco has expanded its partnership with NVIDIA to accelerate AI infrastructure across neocloud, enterprise and telecom markets, announcing a new generation of data-centre switches and a unified architecture for AI workloads.
At NVIDIA’s GTC event in Washington, Cisco introduced the N9100 series switch – the first NVIDIA partner-developed data-centre switch based on NVIDIA Spectrum-X Ethernet switch silicon.
The launch makes Cisco the first vendor to offer a NVIDIA Cloud Partner (NCP)-compliant reference architecture, designed for neocloud and sovereign cloud customers.
“We’re at the beginning of the largest data centre build-out in history,” said Jeetu Patel, President and Chief Product Officer, Cisco.
“The infrastructure that will power the agentic AI applications and innovation of the future requires new architectures designed to overcome today’s constraints in power, computing, and network performance.
“Together, Cisco and NVIDIA are leading the way in defining the technologies that will power these AI-ready data centres in all their varieties, from emerging neoclouds, to global service providers, to enterprises, and beyond.”
Secure AI Factory
Cisco also announced updates to its Secure AI Factory with NVIDIA, a platform that integrates compute, networking, security and observability into a single AI infrastructure framework.
Since its debut at GTC in March 2025, the platform has provided enterprises with a comprehensive architecture for AI workloads. Key features include:
- Security and Observability: Cisco AI Defense integrates with NVIDIA NeMo Guardrails for AI cybersecurity. It can help security and AI teams protect models and applications and limit sensitive data leaving the data centre. Splunk Observability Cloud provides monitoring of performance, quality, security, and cost, while Splunk Enterprise Security extends visibility to AI workloads.
- Core AI Infrastructure: Cisco Isovalent is validated for inference workloads on AI PODs. Nexus Hyperfabric AI with the cloud-managed G200 Silicon One switch delivers 800G Ethernet. UCS 880A M8 and UCS X-Series modular servers with NVIDIA GPUs support a range of AI workloads, including generative AI fine-tuning and inference.
- Ecosystem Expansion: NVIDIA Run:ai software enables AI workload orchestration. Nutanix Kubernetes Platform (NKP) and Nutanix Unified Storage (NUS), along with Nutanix Enterprise AI (NAI), support containerised deployment and simplify operations.
- Government Readiness: Cisco aligns with NVIDIA AI Factory for Government, providing a full-stack reference design for regulated environments.
“NVIDIA Spectrum-X Ethernet delivers the performance of accelerated networking for Ethernet,” said Gilad Shainer, SVP of Networking at NVIDIA.
“Working with Cisco’s Cloud Reference Architectures and NVIDIA Cloud Partner design principles, customers can choose to deploy Spectrum-X Ethernet using the newest Cisco N9100 series or Cisco Silicon One based switches to build open, high-performance AI networks.”
AI-native wireless stack sets stage for 6G
In the telecom sector, Cisco, NVIDIA and industry partners revealed the first AI-native wireless stack – a key step towards 6G.
As AI extends beyond smartphones into connected vehicles, robotics and wearable devices, the partners aim to help telecom providers handle billions of simultaneous connections with improved speed and efficiency.
The new AI-RAN stack combines Cisco’s user-plane function and 5G core software with the NVIDIA AI Aerial platform.
This enables providers to integrate AI into their networks starting with advanced 5G services, while establishing the groundwork for next-generation 6G networks.
According to Cisco, this will allow telecom providers to “infuse AI into their mobile networks, starting with 5G advanced services and establishing the groundwork for 6G.”
Implications for IT Leaders
For IT leaders, Cisco and NVIDIA’s latest moves could reshape how AI infrastructure is deployed.
The N9100 switches, paired with Nexus Dashboard, aim to simplify operations, offering a single view across complex AI workloads.
Support for both NX-OS and SONiC gives teams the flexibility to plug AI networking into existing cloud or enterprise environments without major overhauls.
Security and observability are also built in.
The Secure AI Factory with NVIDIA provides tools to monitor AI workloads, protect sensitive data, and meet regulatory requirements, while the AI-native wireless stack for 5G and future 6G networks sets the stage for billions of connected devices and next-generation AI applications.
Taken together with high-performance AI PODs and orchestrated GPU workloads, the platform could help IT teams scale efficiently and manage demanding AI tasks.
Interoperable hardware and software from NVIDIA, Nutanix, and other partners further eases deployment and ongoing operations, offering a more integrated approach to building AI-ready networks.
Building an AI-ready communications backbone
Cisco’s announcements highlight a strategic approach to unify compute, networking and AI under a single operational framework.
From data-centre performance to enterprise collaboration and telecom innovation, the partnership aims to streamline how organisations build, manage and secure AI-driven infrastructure.
As AI becomes embedded across collaboration, cloud and connectivity, Cisco and NVIDIA are positioning networks themselves as the foundation for intelligent systems.
