The complete solution comprises up to 1,152 NVIDIA Rubin GPUs and 576 NVIDIA Vera CPUs in liquid-cooled racks.
Super Micro Computer, a total solution provider for AI, enterprise, storage, and 5G/edge, introduces the Data Center Building Block Solutions® (DCBBS) Blueprint for HPC based on the NVIDIA Vera Rubin NVL4 platform, announced during ISC 2026. Following Supermicro’s DCBBS Blueprints for NVIDIA Vera Rubin NVL72 and NVIDIA HGX Rubin NVL8, which were introduced during Computex, the Blueprint for HPC and AI applies the same end-to-end methodology to scientific computing. The Blueprint is based on Supermicro’s DCBBS, which includes the necessary compute, networking, advanced liquid cooling, power distribution, and site infrastructure and is delivered by a team of Supermicro DCBBS experts, to accelerate time to commissioning for research institutions and supercomputing centers.
“Scientific discoveries have always been driven by the tools available to researchers, and AI has now become an essential part of the research process,” said Charles Liang, President and CEO of Supermicro. “The institutions that accelerate infrastructure rollout will lead the next generation of breakthroughs. With our DCBBS Blueprints for NVIDIA Vera Rubin NVL4, research organizations can confidently deploy HPC and AI infrastructure at any scale, knowing that it is supported by Supermicro’s proven experience in building some of the world’s largest liquid-cooled clusters.”
Also Read: Power Integrations Unveils Slim PSU for NVIDIA AI Data Center
Researchers are increasingly relying on a converged approach to computing, combining traditional double-precision FP64 simulations with accelerated computing and AI methods, drastically shortening the time to new discoveries in climate research, drug development, materials science, and the energy sector. The NVIDIA Vera Rubin NVL4 platform is designed for this convergence, and the DCBBS Blueprint for HPC defines the steps for a successful implementation, supported by Supermicro’s proven track record in building the world’s largest liquid-cooled supercomputing clusters with more than 100,000 GPUs.
The Blueprint covers the entire end-to-end process Supermicro has deployed to realize large-scale liquid-cooled projects in record time: During site surveys, Supermicro experts assess loading dock access, data hall dimensions and available clearance, floor load, and existing power and cooling infrastructure to develop a tailored design proposal for each project. System integration begins well before delivery, with racking, stacking, cabling, and system-level (L10) and cluster-level (L11) testing at Supermicro’s global manufacturing facilities. White-glove delivery and on-site integration include rack placement, power and cooling connections, network cabling, commissioning, and on-site system validation, with ongoing support options, including on-site response times of just 4 hours for business-critical uptime.
A scalable AI and HPC solution to modernize the computing infrastructure for scientific research
The Supermicro DCBBS Blueprint for HPC and AI based on the NVIDIA Vera Rubin NVL4 Scalable Unit includes the following, which can be scaled up to deploy clusters of any size, from 3.2 MW to 1 GW:
- 8 liquid-cooled compute racks in custom 750 mm wide 52U enclosures, each equipped with 36 NVIDIA Vera Rubin NVL4 nodes within a power budget of 362 kW, accounting for a total of 288 nodes, up to 1,152 NVIDIA Rubin GPUs and 576 NVIDIA Vera CPUs per Scalable Unit.
- Advanced Direct Liquid Cooling (DLC-2) technology stack, including 3 in-row cooling distribution units (up to 1.8 MW per unit) per Scalable Unit in a 2+1 redundant configuration, direct-to-chip copper cold plates and vertically mounted cooling distribution manifolds, with Supermicro SMC PG25-A refrigerant developed for exceptional chemical and thermal stability.
- NVIDIA Quantum-X800 InfiniBand compute fabric via dedicated network switch racks, with fully liquid-cooled options available, ensures scalable, high-bandwidth interconnection for distributed scientific and AI workloads.
- Rack Power Supply and Management: 8 power shelves of 72 kW per compute rack provide busbar power for the 362 kW power budget, with two ToR management switches per rack for out-of-band management.
Configurations for HPC and AI based on the NVIDIA GB200 NVL4 platform are also available for immediate deployment.
Supermicro DCBBS delivers a complete, modular AI infrastructure built from validated components and subsystems, enabling flexible deployment from individual servers and networks to full rack- and data center-level solutions, including software and services. Supermicro remains a sector leader with an extensive portfolio of AI infrastructure solutions, enabling organizations worldwide to deploy scalable, efficient, and environmentally friendly AI data centers.
Source: PRNewswire
