Cisco Partners with Supermicro for Rack-Scale AI Infrastructure
The networking giant expands its Secure AI Factory architecture to include high-density compute systems targeting neocloud and sovereign cloud deployments.
Cisco is expanding its AI infrastructure offerings through a partnership with Supermicro that adds rack-scale computing capabilities to its Secure AI Factory architecture. The collaboration brings high-density, liquid- and air-cooled server systems into Cisco's portfolio, creating what the company positions as end-to-end infrastructure for organizations deploying large-scale AI workloads.
The expanded architecture targets enterprise, neocloud, and sovereign cloud customers seeking validated, full-stack solutions that combine compute and networking in pre-integrated configurations. According to details first reported by Cisco, the systems will support workloads ranging from trillion-parameter model training to edge inference applications.
NVIDIA Cloud Partner compliance
The infrastructure meets NVIDIA Cloud Partner (NCP) requirements, combining Supermicro compute with Cisco's networking portfolio. The architecture uses Cisco Silicon One-based switches for front-end networking and NVIDIA Spectrum-X switches for back-end connectivity, unified through Cisco Nexus One management. Cisco notes it is the only NVIDIA technology partner using its own networking switches and operating system in an NCP-compliant solution.
The systems support NVIDIA's latest platforms, including Vera Rubin NVL72 and HGX Rubin NVL8, with options for rack-to-fabric liquid cooling alongside traditional air-cooled configurations.
Why it matters
As organizations race to deploy AI capabilities, infrastructure complexity has become a significant barrier. Pre-validated, full-stack architectures reduce deployment risk and time-to-production—critical factors when GPU availability remains constrained and power density requirements push traditional data center designs to their limits. For neocloud and sovereign cloud providers, NCP compliance provides a standardized foundation for building AI-as-a-service offerings while maintaining control over data residency and security.
Deployment and support model
Cisco is introducing Cisco Validated Infrastructure Services (CVIS), aligned with NVIDIA Infrastructure Services, to certify that deployed systems match reference architecture specifications. The company is building a dedicated large-scale AI lab to develop testing tools and validation software for these services.
The architecture integrates with NVIDIA AI Enterprise software and what Cisco calls "AgenticOps" through Cisco Cloud Control, allowing customers to correlate job performance with compute, network interface card, optics, and network metrics for end-to-end observability.
"AI factories are revenue-generating infrastructure, where compute produces intelligence, and intelligence drives revenue," said Justin Boitano, vice president of Enterprise AI at NVIDIA, in the announcement. "By expanding the Cisco Secure AI Factory with NVIDIA, Cisco is providing enterprises and neoclouds full-stack infrastructure that can get into production faster and generate more value from every watt."
Cisco will begin offering Supermicro compute solutions as part of the Secure AI Factory portfolio in October 2026, according to the announcement. The company positions the combined supply chain capabilities of both partners as a risk mitigation strategy amid ongoing GPU and memory supply constraints.
These details were first reported by Cisco in a press release issued August 25, 2026.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call