Supermicro Unveils NVIDIA Vera Rubin NVL4 Blueprint for HPC and AI Infrastructure

Supermicro has introduced a new Data Center Building Block Solutions (DCBBS) Blueprint for high performance computing based on the NVIDIA Vera Rubin NVL4 platform, announced in a press release. The solution was presented at ISC 2026 and extends Supermicro’s existing DCBBS Blueprints for NVIDIA Vera Rubin NVL72 and HGX Rubin NVL8 to scientific computing applications.

The complete solution includes up to 1,152 NVIDIA Rubin GPUs and 576 NVIDIA Vera CPUs arranged in liquid cooled racks. Each 3.2 MW Scalable Unit supports 362 kW per rack through Direct Liquid Cooling technology, including in row cooling distribution units and copper cold plates. The system is designed to scale from small research clusters to installations of one gigawatt.

Supermicro’s DCBBS Blueprint covers all phases of deployment, from site surveys and design proposals to manufacturing, system testing, and on site integration. Each compute rack includes eight power shelves supplying 72 kW each, with dedicated management switches for remote control and maintenance.

Networking across the solution uses NVIDIA Quantum X800 InfiniBand fabric to support distributed scientific and AI workloads. According to Supermicro, configurations based on the NVIDIA GB200 NVL4 are also available for immediate deployment. The company will showcase the new infrastructure at ISC 2026 in Hamburg.