Cisco Secure AI Factory with NVIDIA Expands to Supermicro Rack-Scale Systems

Cisco this week announced that its Secure AI Factory with NVIDIA is expanding to rack-scale infrastructure built around Supermicro liquid-cooled and air-cooled systems. NVIDIA Cloud Partner Reference Architecture compliance anchors the design, joined by Cisco Validated Infrastructure Services, Cisco AI networking, and unified operations.

Cisco says it will begin offering the Supermicro systems in October through its enterprise sales and channel ecosystem, with Cisco support and lifecycle services wrapped around the hardware. Cisco Secure AI Factory with NVIDIA Expands to Supermicro Rack-Scale Systems Cisco plans to pair its NCP Reference Architecture, networking, validation services, and management software with liquid-cooled and air-cooled Supermicro rack-scale systems. Cisco’s reference design splits the network between two switch families. Frontend fabric runs on Cisco N9300 Series switches with Cisco Silicon One, while the backend fabric uses Cisco N9100 Series switches built around NVIDIA Spectrum-X silicon. Cisco positions that pairing for clusters spanning roughly 1,000 to more than 100,000 GPUs, mapping it onto NVL72 builds for Vera Rubin and Grace Blackwell, HGX NVL8 systems for Rubin and B300, and MGX PCIe GPU platforms. CVIS is where Cisco takes ownership of the build. A CVIS Toolkit handles repeatable provisioning and automated validation, specialist-led delivery hands over a validated cluster with a performance and compliance report, and dedicated CVIS performance clusters support software testing and tool development. Cisco splits its validated designs by scale. An Enterprise Reference Architecture targets AI server clusters under roughly 1,000 (1024?) GPUs on Cisco Silicon One N9300 Series switching, while a Cloud Reference Architecture extends from about 1,000 to more than 100,000 GPUs with an N9300 frontend and an N9100 Series backend carrying NVIDIA Spectrum-X silicon. The Supermicro systems are focused on the AI clusters beyond Cisco Unified Edge and Cisco AI POD offerings. Supermicro’s AI clusters are aimed at model training and trillion-parameter workloads rather than edge inferencing or optimization. Cisco says it will begin offering the Supermicro systems in October through its enterprise sales and channel partner ecosystem. Cisco support and lifecycle services will be included, along with the company’s Ethernet silicon, optics, security, and observability portfolio. On a call this week, Patrick asked about who would service the physical boxes. It sounds like if a SSD or GPU needs to be replaced in one of the Supermicro servers, it would be Supermicro doing the physical service.