Summary:
- Vertiv and NVIDIA collaborate on a new reference architecture for AI infrastructure design.
- The architecture supports high rack power densities and is customizable for bespoke data center designs.
- Vertiv’s solutions enable faster scaling, optimized AI-focused data centers, and increased energy efficiency.
Article:
Revolutionizing AI Infrastructure Design with Vertiv and NVIDIA Collaboration
Vertiv and NVIDIA have joined forces to introduce a cutting-edge reference architecture that is set to transform the landscape of AI infrastructure design. This collaboration brings forth a new era of integrated cooling and power strategies, moving away from traditional buildouts to streamline the planning and buildout process for data centers.
One of the key highlights of this partnership is the availability of Vertiv’s solutions as SimReady™ 3D assets in the NVIDIA Omniverse Blueprint. This integration allows for real-time collaboration across IT, facilities, and engineering teams, enabling customers to test, optimize, and validate designs before any physical components are built.
Moreover, Vertiv’s reference architecture for the NVIDIA GB300 NVL72 platform is designed to support rack densities of up to 142kW, setting a new benchmark for accelerated infrastructure tailored for AI factories at scale. This design not only delivers 1.5 times more AI performance but also enables up to 50% faster on-site builds and requires 30% less physical space.
Additionally, the architecture is liquid cooling-ready, offering thermal flexibility and supporting both air- and hybrid-cooled configurations. This feature contributes to up to 70% better annual energy efficiency, achieved through higher water temperatures.
Key Benefits of Vertiv’s Reference Architecture for NVIDIA GB300 NVL72:
- Simulation to Deployment in One Path: Bridging physical and digital infrastructure for real-time collaboration.
- Built for the Age of AI Reasoning: Setting a new standard for accelerated infrastructure engineered for AI factories at scale.
- Accelerated Performance, Scale, and Speed: Enabling faster scaling with unprecedented agility and energy efficiency.
- Thermal Flexibility Meets Efficiency Gains: Achieving up to 70% better annual energy efficiency through liquid cooling readiness.
- Design, Deploy, Operate – Globally: Backed by Vertiv’s global network of field service engineers for unmatched scalability and reliability.
With Vertiv and NVIDIA’s innovative collaboration, customers can now unlock new levels of performance, scalability, and energy efficiency as they pave the way for the future of AI infrastructure. This comprehensive reference design and SimReady™ 3D assets offer a glimpse into the next generation of data center design, tailored to meet the evolving demands of AI workloads and high-performance computing applications.
