Supermicro Delivers NVIDIA Vera Rubin NVL4 End-to-End DCBBS Blueprint With Native FP64 Performance for Converged HPC and AI Infrastructure
HPCwire, Monday, June 22nd, 2026
Supermicro launched an end-to-end DCBBS blueprint for NVIDIA Vera Rubin NVL4, targeting converged FP64 HPC and AI.
Supermicro introduced its Data Center Building Block Solutions (DCBBS) Blueprint for HPC built on the NVIDIA Vera Rubin NVL4 platform, announced at ISC 2026.
The full solution scales to 1,152 NVIDIA Rubin GPUs and 576 Vera CPUs in liquid-cooled racks, with a 3.2MW Scalable Unit delivering performance for AI workloads and FP64 simulation.
Each unit comprises 8 liquid-cooled 52U compute racks housing 288 nodes, using DLC-2 direct liquid cooling and NVIDIA Quantum-X800 InfiniBand fabric.
The blueprint pairs traditional FP64 double-precision simulation with accelerated AI computing.
Supermicro DCBBS experts deliver compute, networking, cooling, power, and site infrastructure to speed time-to-online for research and supercomputing centers.