Supermicro Ships NVIDIA Vera Rubin NVL72 Racks
Analysis based on 14 articles · First reported Sep 23, 2026 · Last updated Sep 23, 2026
The announcement reinforces Supermicro's position as a first-to-market integrator of NVIDIA's next-generation AI platforms, which could support server revenue growth and margin expansion tied to liquid-cooled AI infrastructure demand. It also strengthens NVIDIA's ecosystem lock-in for the Vera Rubin platform, a positive read-through for AI data center capex and related semiconductor and cooling supply chains.
Supermicro announced on September 23, 2026 that it is now shipping NVIDIA Vera Rubin NVL72 racks integrated with its Data Center Building Block Solutions (DCBBS) and its direct liquid cooling stack (DLC-2). The racks pair 72 NVIDIA Rubin GPUs and 36 NVIDIA Vera CPUs in a single liquid-cooled rack operating as one machine, with eighteen 1U compute trays connected through nine sixth-generation NVIDIA NVLink switch trays delivering 216 TB/s of scale-up bandwidth, 20.7 TB of HBM4 memory, and up to 54 TB of LPDDR5X per rack. Supermicro's DCBBS Blueprint defines a balanced bill-of-materials from 5 MW to gigawatt scale, with one Scalable Unit providing 1,152 NVIDIA Rubin GPUs and 331 TB of HBM4 memory across 16 compute racks. The offering includes in-row cooling distribution units rated at 1.8MW per CDU with N+1 redundancy, optional rear door heat exchangers, and networking integration based on NVIDIA reference architecture. CEO Charles Liang said the company has spent years building the liquid-cooling stack, manufacturing capacity, and deployment teams for this moment, and that customers can now order a production-ready Scalable Unit with end-to-end integration. Supermicro manages the project across site survey, design, integration, testing, delivery, deployment, and ongoing support.
Set up alerts, explore entity relationships, search across thousands of events, and build custom intelligence feeds.
Open Dashboard