Supermicro Computer, Inc., a global leader in comprehensive IT solutions for AI/ML, HPC, Cloud, Storage, and 5G/Edge, has announced a strategic collaboration with NVIDIA to enhance manufacturing capabilities and advance cooling technology. This partnership aims to rapidly deliver liquid-cooled AI infrastructure solutions for NVIDIA’s next-generation Vera Rubin and Rubin platforms.
Through this collaboration, Supermicro gains the fastest time-to-market advantage for the NVIDIA Vera Rubin NVL72 and NVIDIA HGX Rubin NVL8 solutions. Leveraging Supermicro’s Data Center Building Block Solutions (DCBBS), the company accelerates AI infrastructure adoption by utilizing a modular, building block-based design. This approach simplifies production, offers diverse configuration options, and enables rapid deployment for enterprises and hyperscalers alike.
AI Authority Trend: Supermicro Expands NVIDIA Blackwell Portfolio with New Liquid-Cooled HGX B300 Solutions
“Our long-standing partnership with NVIDIA and Supermicro’s agile building block solutions enable us to bring the industry’s best AI platform to market faster than anyone else,” said Charles Liang, president and CEO of Supermicro. “With our expanded manufacturing capabilities and industry-leading liquid cooling technology, we will enable hyperscalers and enterprise customers to quickly and efficiently deploy Vera Rubin and Rubin platform-based AI infrastructure at scale.”
Supermicro supports a variety of advanced AI infrastructure solutions for NVIDIA’s next-generation platforms. The NVIDIA Vera Rubin NVL72 Supercluster integrates 72 NVIDIA Rubin GPUs and 36 NVIDIA Vera CPUs into a single rack-scale solution. It ensures high-speed GPU-to-GPU and CPU-to-GPU communication via NVIDIA NVLink 6 interconnects, Connect X-9 Super NICs, and BlueField-4 DPUs. The system achieves massive cluster scalability through NVIDIA Quantum-X800 InfiniBand and Spectrum-X Ethernet, delivering up to 3.6 exaflops of AI compute performance, 1.4 petaflops of HBM4 memory bandwidth, and 75 TB of high-speed memory. Its liquid cooling technology minimizes energy and water consumption while maintaining high-density efficiency.
AI Authority Trend: Supermicro Launches Air-Cooled AI Solution with AMD Instinct MI355X
The 2U liquid-cooled NVIDIA HGX Rubin NVL8 solution is a compact, 8-GPU system optimized for AI and HPC workloads. It provides up to 400 petaflops of compute performance, 176 TB/s of memory bandwidth, and high-speed networking via 1600Gb/s NVIDIA Connect-9 Super NICs. Its x86-based configuration supports Intel Xeon or AMD EPYC processors, while Supermicro’s advanced DLC technology maximizes energy efficiency and rack consolidation.
Additionally, NVIDIA’s next-gen technologies including NVLink 6, Vera CPUs, Transformer Engine 3rd Gen, Confidential Computing, and the RAS Engine 2nd Gen enhance compute, memory, security, and reliability. The platform also integrates NVIDIA Spectrum-X Ethernet photonics networking, providing up to 102.4Tb/s switching performance with unmatched energy efficiency and uptime.
By investing in expanded manufacturing and an end-to-end liquid-cooling stack, Supermicro is accelerating production, rigorous validation, and deployment of high-density AI infrastructure. Coupled with the DCBBS modular architecture, these efforts allow enterprises to scale rapidly, achieve superior performance, and gain a competitive edge in the AI-driven era.
AI Authority Trend: Supermicro Launches Fully Integrated AI Factory Solutions Powered by NVIDIA Blackwell
To share your insights, please write to us at info@intentamplify.com
