International. Schneider Electric's newly launched systems address energy and sustainability challenges driven by the high demand for AI systems.
First, the company announced a new reference design for data centers, co-developed with NVIDIA, that will support high-density AI clusters and liquid cooling up to 132 kW per rack. Optimized for NVIDIA's GB200 NVL72 and Blackwell chips, the design streamlines planning and deployment with tested and validated architectures, addressing the unique challenges of utilizing liquid cooling at scale.
In addition, Schneider Electric introduced its new Galaxy VXL uninterruptible power supply (UPS), designed for AI, data centers, and large-scale electrical workloads. The Galaxy VXL UPS offers 52% space savings compared to the industry average, and with a power density of up to 1042 kW/m², this 1.25 MW scalable modular UPS is designed to deliver power in a smaller, high-density space.
Both innovations are part of Schneider Electric's end-to-end solutions for AI-ready data centers, which focus on three key areas: developing an energy strategy for the AI era, deploying advanced infrastructure, and sustainability consulting.
"The energy and environmental impact of AI is growing at an unprecedented rate, and it is paramount that we bend the energy curve downward by finding new ways to decarbonize data centers and digital infrastructure," said Pankaj Sharma, Executive Vice President of Data Centers and Networks at Schneider Electric. "At Schneider Electric we are committed to pushing the boundaries, setting new standards, and shaping the future of AI, while protecting the environment. This requires a strategic approach from the network to the chip, to the cooler, and beyond."
NVIDIA Partnership
Schneider Electric's latest data center reference design has been co-developed with NVIDIA for the implementation of liquid cooling in hyperscale, colocation, and enterprise data center environments.
Such a reference design includes options for liquid-to-liquid (CDU) and direct-to-chip liquid cooling distribution units, and shares complete mechanical and electrical plans for more energy-efficient and sustainable operations.
This solution can be customized to meet specific AI workload requirements, while helping users leverage the most sustainable and energy-efficient infrastructure designs for high-density applications.
"Building the future of accelerated computing and AI requires speed and a strong foundation," said Jensen Huang, founder and CEO of NVIDIA. "Our work with Schneider Electric enables customers to design the world's technological advancements on a stable and resilient infrastructure. Together, we are creating AI data centers built specifically for accelerated computing, supporting complex architectures essential to delivering digital intelligence to all businesses and industries."
