Supermicro Begins Shipping NVIDIA Vera Rubin NVL72 Racks with Full…

Event summary

Supermicro is now shipping NVIDIA Vera Rubin NVL72 racks integrated with its Data Center Building Block Solutions (DCBBS) and direct liquid cooling stack (DLC-2).
Each NVL72 rack contains 72 NVIDIA Rubin GPUs and 36 NVIDIA Vera CPUs, with 20.7 TB of HBM4 memory and up to 54 TB of LPDDR5X.
Supermicro's DCBBS Blueprint defines a balanced bill-of-materials for deployments from 5 MW to gigawatt scale, with 1,152 GPUs and 331 TB of HBM4 per Scalable Unit.
The company offers end-to-end integration, from site survey to ongoing support, for NVIDIA Vera Rubin NVL72 deployments.

The big picture

Supermicro's move to ship NVIDIA Vera Rubin NVL72 racks underscores the growing importance of liquid cooling in high-performance AI infrastructure. The partnership with NVIDIA highlights the strategic shift towards more efficient, high-throughput AI systems. As AI workloads become more demanding, the ability to provide end-to-end cooling solutions could become a key differentiator in the market.

What we're watching

Deployment Pace: How quickly Supermicro can scale production and deployment of NVIDIA Vera Rubin NVL72 racks to meet market demand.
Competitive Positioning: Whether Supermicro's end-to-end liquid cooling solutions can differentiate it from competitors in the AI hardware space.
Customer Adoption: The pace at which enterprises and cloud providers adopt the NVIDIA Vera Rubin NVL72 racks for AI workloads.