Cisco and Nvidia Partner to Bring Rack-Scale AI Infrastructure to Enterprises

Cisco and Nvidia are collaborating to bring rack-scale AI infrastructure to the enterprise, neocloud, and sovereign cloud markets. The partnership seeks to lower technical barriers for organizations that lack the massive scale of cloud giants but require high-performance computing to run generative AI workloads.

As part of the expansion of the "Cisco Secure AI Factory with NVIDIA," the two companies are integrating rack-scale power and networking. To deliver the necessary high-density compute systems within this architecture, Cisco is partnering with Supermicro. The resulting infrastructure supports a variety of high-performance environments, including both liquid-cooled and air-cooled GPU systems, providing a streamlined path for companies to deploy AI capabilities on-premises or in hybrid cloud configurations.

The Enterprise AI Gap

Most enterprise companies struggle to implement generative AI because the underlying physical requirements are prohibitively complex. The power, cooling, and networking specifications required for modern AI clusters have traditionally been the domain of hyperscalers—the massive cloud providers with the capital and expertise to build specialized data centers. For the average enterprise, the gap between wanting to run a large language model and having the physical infrastructure to support it has remained a significant bottleneck.

Lowering the Barrier to Entry

By providing pre-integrated, rack-scale solutions, Cisco and Nvidia are effectively democratizing the AI data center. This shift allows organizations to bypass the fragmented process of sourcing and integrating separate compute, power, and networking components. When these elements are delivered as a cohesive unit, the time to deployment is reduced and the risk of configuration errors—which can be catastrophic in high-density GPU environments—is minimized. This enables a broader range of companies to maintain control over their data and workloads without relying entirely on public cloud providers.

The Path Forward

Industry observers will now be watching how quickly these integrated racks can be deployed across diverse enterprise environments. While the partnership with Supermicro addresses compute density, the long-term success of the initiative depends on how easily these systems can be integrated into existing legacy data centers that may not have been built for the extreme power and cooling demands of Nvidia's latest GPUs. Further details on the specific scaling limits and pricing for these "AI factories" remain to be seen.