AI is reshaping industries, necessitating new data center infrastructure that can handle high-performance, secure, and adaptable workloads. Traditional networks struggle to manage the demands of AI, particularly the intense GPU-to-GPU traffic required for AI training. Organizations now seek a unified operating model that integrates on-premises and cloud environments, alongside automation across systems. Cisco's Nexus operating model, paired with the Nexus Dashboard and Nexus Hyperfabric, aims to streamline operations and provide consistent management across diverse networks. Cisco is expanding its Nexus portfolio with the new N9100 Series Switches, which incorporate NVIDIA Spectrum-X Ethernet switch silicon. This series offers a choice between Cisco’s reliable NX-OS and the open SONiC operating system, catering to different customer needs. These switches align with Cisco's Cloud Reference Architecture, which emphasizes flexibility and programmability, ensuring scalability for AI infrastructure. The Nexus Hyperfabric AI is now available, simplifying the deployment of full-stack AI clusters. An NVIDIA Cloud Partner compliant reference architecture will be available for the N9100 Series, targeting neocloud and sovereign cloud customers. This architecture combines NVIDIA's AI technology with Cisco's reliability, facilitating easier management. Companies like BlueSky Compute are already planning to use these switches in their AI training clusters, aiming to enhance their capabilities rapidly. The collaboration between Cisco and NVIDIA is focused on creating a secure, manageable, and scalable network foundation to support AI's growth in enterprises.