Super Micro Computer Ships Nvidia Vera Rubin Racks With Liquid Cooling for AI Workloads

Super Micro Computer Begins Vera Rubin Server Racks Shipments

Super Micro Computer (SMCI) announced it has begun shipping Nvidia Vera Rubin NVL72 racks integrated with its Data Center Building Block Solutions and direct liquid cooling technology1

1

. The company's shares rose 1.7% following the announcement, reflecting investor confidence in its expanding AI hardware infrastructure capabilities. The San Jose-based Total IT solutions portfolio provider has positioned itself to address the growing demand for scalable AI infrastructure deployments across data centers worldwide.

Technical Specifications of Nvidia Vera Rubin Platform

Each Nvidia Vera Rubin NVL72 rack contains 72 NVIDIA Rubin GPUs and 36 Vera CPUs housed in a single liquid-cooled rack configuration1

1

. The system features 18 compute trays that collectively deliver 20.7 TB of HBM4 memory and up to 54 TB of LPDDR5X, providing substantial AI compute resources for demanding AI workloads. The Nvidia Vera Rubin platform was specifically designed to work with direct liquid cooling technology to achieve efficiency levels that remain unattainable with air cooling alone. This architectural approach addresses the thermal challenges posed by high-density GPU deployments in modern data centers.

Data Center Building Block Solutions Architecture

Super Micro Computer's DCBBS Blueprint defines a comprehensive bill-of-materials spanning from 5 MW to gigawatt scale implementations1

1

. Each Scalable Unit incorporates 1,152 GPUs and 331 TB of HBM4 memory, enabling organizations to scale their AI infrastructure according to specific requirements. The modular building block solutions approach allows customers to optimize configurations for exact workload and application needs by selecting from a comprehensive family of systems. This flexibility extends across form factors, processors, memory, storage, networking, power, and cooling solutions.

In-House Manufacturing and Liquid Cooling Components

Super Micro Computer manufactures liquid cooling components in-house at facilities located in the United States, Taiwan, and the Netherlands2

2

. The company produces cold plates, manifolds, hose kits, rack power shelves, cooling distribution units, and cooling towers internally. The in-row cooling distribution units are rated at 1.8MW per CDU and deployed with N+1 redundancy to ensure reliability1

1

. This vertical integration enables tighter quality control and faster deployment timelines for customers implementing large-scale AI infrastructure.

Comprehensive Integration and Deployment Services

Super Micro Computer provides end-to-end integration services that encompass site survey, project design, integration, testing, on-site deployment, and ongoing support1

1

. The company tests and validates every rack with the full liquid cooling stack before deployment, minimizing risks associated with complex AI hardware infrastructure installations. This comprehensive service model positions SMCI as a turnkey provider for organizations seeking to rapidly deploy AI compute resources without developing internal expertise in liquid cooling systems or site design complexities. The integration of Vera Rubin server racks into the company's portfolio strengthens its competitive position in the accelerating market for AI workloads infrastructure.