Cisco and Nvidia Expand Secure AI Factory with Supermicro Servers and Liquid Cooling

Cisco is set to integrate Supermicro servers into the "Secure AI Factory" solution it jointly operates with Nvidia, marking a significant expansion of the companies' collaborative AI infrastructure offering. According to Techzine's report on August 25, the first shipment of the newly added systems is scheduled for October, signaling a rapid move to market for the enhanced platform.

The updated system leverages a high-density GPU configuration that allows enterprises to run AI clusters and non-AI workloads concurrently within the same environment. This hybrid capability is increasingly critical as organizations seek to maximize infrastructure utilization while scaling artificial intelligence operations. The integrated solution combines Cisco's liquid-cooled AI networking system with Supermicro's liquid-cooled servers, delivering rack-and-fabric liquid cooling designed to handle the thermal demands of next-generation AI hardware.

The combined offering supports training models with up to 1 trillion parameters and is engineered to manage high-throughput inference workloads on platforms such as Nvidia Vera Rubin NVL72 and HGX Rubin NVL8. Cisco will deliver Nvidia Cloud Partner (NCP)-certified solutions to neocloud and sovereign cloud customers, with Cisco Silicon One-based switches managing the front-end network and Spectrum-X-based switches handling the back-end. The Cisco Nexus One platform connects the two layers, providing a unified architecture for AI-scale deployments.

Cisco emphasized that it is the only Nvidia technology partner applying its own switches and network operating system together to NCP-certified solutions, a differentiation that underscores the depth of its integration with Nvidia's AI ecosystem. Additionally, Cisco is introducing Cisco Validated Infrastructure Services (CVIS) tailored to Nvidia Infrastructure Services, offering customers a validated path to deployment and operational efficiency.

Jeetu Patel, Cisco's president and chief product officer, framed the announcement within the broader industry momentum, stating, "This is the moment when the largest data center buildout in history is starting." He added, "Every company is joining the race to scale AI, but speed only matters when it is backed by data control, management of token costs, and tangible return on investment."

The expansion comes at a time when hyperscale and enterprise data center operators are increasingly prioritizing liquid cooling and high-density compute to support AI workloads. By combining Supermicro's server expertise with Cisco's networking and Nvidia's GPU platforms, the Secure AI Factory aims to offer a turnkey solution that reduces deployment complexity while addressing the power and cooling constraints facing modern data centers.