Cisco and NVIDIA Launch AI Factory to Transform Enterprise AI and Security

Cisco has introduced an AI factory architecture developed in collaboration with NVIDIA, with security as a central focus. Building on their expanded partnership announced last month, the two companies have quickly delivered validated reference architectures. They are jointly working on the Cisco Secure AI Factory with NVIDIA to streamline the deployment, management, and security of AI infrastructure for enterprises at any scale.

“AI can unlock groundbreaking opportunities for the enterprise,” said Chuck Robbins, Chair and CEO, Cisco. “To achieve this, the integration of networking and security is essential. Cisco and NVIDIA's trusted, innovative solutions empower our customers to harness AI's full potential simply and securely.”

“AI factories are transforming every industry, and security must be built into every layer to protect data, applications and infrastructure,” said Jensen Huang, Founder and CEO, NVIDIA. “Together, NVIDIA and Cisco are creating the blueprint for secure AI—giving enterprises the foundation they need to confidently scale AI while safeguarding their most valuable assets.”

Building and deploying AI applications demands high-performance, scalable infrastructure and a comprehensive AI software toolchain. Securing this setup calls for a new kind of architecture—one that integrates security across all layers of the AI stack and can adapt automatically as the infrastructure evolves. Cisco and NVIDIA’s collaboration on the NVIDIA Spectrum-XTM Ethernet networking platform forms the basis of the Cisco Secure AI Factory. Cisco is incorporating solutions such as Cisco Hypershield to safeguard AI workloads, and Cisco AI Defense to secure the development, deployment, and use of AI models and applications. This partnership aims to give customers the flexibility to tailor infrastructure to their AI needs while maintaining operational simplicity and security.

Building a Secure AI Factory

AI factories—data centers specifically built to support AI workloads—are becoming more modular, scalable, and agile. However, organizations need to consider more than just compute power, as these environments also introduce complex security challenges. Cisco’s recent State of AI Security report examines dozens of AI-specific threat vectors and over 700 AI-related regulations to highlight trends in this rapidly evolving landscape. Organizations that take a strategic approach to both AI infrastructure and security are better positioned to scale efficiently, adapt quickly, and realize business value sooner.

The Cisco Secure AI Factory with NVIDIA will draw on the companies’ combined capabilities in flexible AI networking and full-stack technology, leveraging a joint architecture to support enterprise AI needs. The solution brings together technologies from Cisco, NVIDIA, and key ecosystem partners in a secure AI factory design, including:

• Compute: Cisco UCS AI servers built on NVIDIA HGX and MGX platforms for accelerated performance

• Networking: Cisco Nexus Hyperfabric AI and Nexus solutions powered by Silicon One and NVIDIA Spectrum-X Ethernet

• Storage: High-performance storage provided by certified partners such as Pure Storage, Hitachi Vantara, NetApp, and VAST Data

• Software: NVIDIA AI Enterprise platform to support streamlined development and deployment of production-grade agentic AI workloads

The Cisco Secure AI Factory with NVIDIA includes security at all layers:

• Securing the Infrastructure: The Cisco Hybrid Mesh Firewall offers unified security management with consistent policies applied across various enforcement points, including network switches, traditional firewalls, and workload agents. This integrated approach provides broad and consistent protection, combining deep packet inspection with extensive infrastructure coverage to detect, block, and contain threats. Looking ahead, Cisco Hypershield—part of the Hybrid Mesh Firewall—will integrate with NVIDIA BlueField-3 DPUs to bring pervasive, zero-trust security enforcement to every AI node.

• Securing the Workload: Cisco Hypershield helps prevent lateral movement by adversaries and mitigates vulnerabilities proactively without requiring patches, all through a unified management interface. It monitors and controls process execution, file access, and network activity to provide deep visibility and precise runtime enforcement within AI workloads. Future updates will enhance this protection by integrating with NVIDIA BlueField-3’s DOCA AppShield, enabling real-time threat detection in AI-specific virtual machines and containers.

• Securing the AI Application: Cisco AI Defense Cisco AI Defense provides security and AI teams with tools to address safety risks—such as off-policy actions or toxic behavior—and security concerns like prompt injection and data privacy throughout the AI development lifecycle. It integrates into existing CI/CD pipelines, enabling automated vulnerability testing and consistent runtime protection across multiple models and applications. The solution also supports alignment with AI security standards including NIST, MITRE ATLAS, and OWASP LLM Top 10. Future updates will include integration with NVIDIA AI Enterprise to further streamline AI security workflows.

Cisco and NVIDIA bring complementary expertise in AI infrastructure, and by combining their insights, they offer flexible deployment models supported by validated reference architectures. The Secure AI Factory is designed to deliver scalable, high-performance infrastructure for enterprises, with security integrated at every stage to support customers throughout their AI journey.

Cisco Secure AI Factory with NVIDIA will have flexible deployment options, including:

• Ready-to-Deploy: By leveraging Cisco Nexus Hyperfabric AI, Cisco’s security solutions, and NVIDIA technology, customers can implement a vertically integrated AI solution that streamlines and automates the secure AI factory lifecycle—from design and deployment to ongoing monitoring.

• Build-your-Own: With modular and customizable components from Cisco, NVIDIA, and their storage ecosystem partners, customers can integrate existing infrastructure and create solutions tailored to their specific environments.

"In today’s fast-moving market, businesses need more than just technology—they need end-to-end solutions that address their most pressing challenges. I see Cisco and NVIDIA combining their strengths to deliver integrated solutions that I believe will drive innovation, simplify deployment, and streamline operations,” said Patrick Moorhead, Founder, CEO, and Chief Analyst, Moor Insights & Strategy. “AI isn't easy but the combination of the two could be an ‘easy button’ for AI infrastructure. By making AI infrastructure easier to adopt, and manage, they could empower enterprises to accelerate digital transformation and achieve their strategic goals with more confidence."

Cisco and NVIDIA: The Journey to a Validated and Unified Architecture

Speed is essential to keep up with the growing demand for AI infrastructure, and Cisco and NVIDIA have made notable progress since announcing their collaboration in February 2025. Cisco has introduced new reference architectures with deployment options for Cisco Nexus Hyperfabric AI or Cisco Nexus 9000 Series Switches, validated against the NVIDIA Enterprise Reference Architecture for HGX H200 and Spectrum-X.

Availability

Solutions built on the Cisco Secure AI Factory with NVIDIA architecture are expected to be available for purchase by the end of 2025, while many of the individual technology components included in the architecture are already available.

Additional Resources

• Executive Blog Post: Embracing the AI Era: Cisco Secure AI Factory with NVIDIA, Jeetu Patel, Cisco’s Executive Vice President and Chief Product Officer.

• For more information on the Cisco Secure AI Factory with NVIDIA, click here.