Two major industrial companies have just unveiled a blueprint that could reshape how AI data centers are built, cutting energy waste and installation costs significantly while speeding up deployment. Trane Technologies and Eaton announced a first-of-its-kind integrated reference design that combines advanced cooling systems with intelligent power management, achieving up to 15% energy efficiency gains, reducing installation costs by up to 30%, and cutting copper use by as much as 80%.

Why Does AI Data Center Design Matter Right Now?

The timing of this collaboration is critical. Global data center capacity is expected to nearly triple by 2030, with artificial intelligence (AI) driving roughly 70% of that growth. Traditional data center designs treat power and cooling as separate systems, requiring slow, manual coordination between teams. This siloed approach wastes resources and delays deployment. The new reference design breaks down these barriers by creating a unified system where thermal management and electrical power systems communicate in real time, responding dynamically to changing demands.

The collaboration is built around NVIDIA's DSX (Data Center Scale) platform, a standardized blueprint for AI factories. By aligning with this industry standard, Trane and Eaton enable data center operators to deploy repeated, predictable solutions rather than custom-engineering each facility from scratch.

"AI and high-performance computing are transforming the demands placed on data centers, and customers want solutions that can keep pace with their needs. By combining our advanced thermal management solutions with Eaton's innovative power management solutions, we're delivering a coordinated design that helps customers accelerate deployment, improve efficiency and confidently plan to scale for the future," said Mauro J. Atalla, Senior Vice President, Chief Technology and Sustainability Officer, Trane Technologies.

Mauro J. Atalla, Senior Vice President, Chief Technology and Sustainability Officer, Trane Technologies

What Technical Changes Make This Design Different?

The reference design advances medium-voltage electrical architectures, which handle higher power densities than traditional low-voltage systems. This shift alone enables significant efficiency gains. The two companies have embedded their coordinated designs into the Trane Continuum Rubin DSX and Eaton Beam Rubin DSX platforms, creating pre-engineered solutions that data center teams can deploy repeatedly.

The approach also reduces material waste. By optimizing electrical distribution through medium-voltage systems, the design cuts copper consumption by up to 80% compared to conventional low-voltage designs. Copper is a critical material in electrical infrastructure, so this reduction has both environmental and cost implications.

Importantly, the design is built to evolve. As emerging technologies like liquid cooling and direct current (DC) architectures become mainstream, the reference design framework can accommodate them without requiring a complete redesign of the underlying system architecture.

How to Evaluate AI Data Center Infrastructure Improvements

• Energy Efficiency Metrics: Look for systems that achieve 15% or greater efficiency gains through integrated thermal and electrical management, compared to traditional siloed designs that treat cooling and power as separate functions.
• Installation Speed and Cost: Evaluate whether the design reduces installation costs by 30% or more and accelerates deployment timelines by using pre-coordinated reference designs rather than custom engineering for each facility.
• Material and Resource Optimization: Assess reductions in copper use and other materials, which indicate both environmental responsibility and long-term cost savings in procurement and installation labor.
• Scalability and Future-Readiness: Confirm that the design can accommodate emerging technologies like liquid cooling and direct current architectures without requiring complete infrastructure replacement.
• Alignment with Industry Standards: Verify that the solution integrates with widely adopted platforms like NVIDIA's DSX, ensuring compatibility with the broader AI infrastructure ecosystem.

"We're advancing the industry standard for speed of deployment by progressing reference designs into unified systems teams can deploy repeatedly. Aligned with the NVIDIA DSX platform, we're integrating our medium-voltage power systems and white space thermal management solutions with Trane's advanced thermal management system architecture to help accelerate AI-factory deployment at scale," said Michael Regelski, Senior Vice President and Chief Technology Officer, Electrical Sector, Eaton.

Michael Regelski, Senior Vice President and Chief Technology Officer, Electrical Sector, Eaton

What Does This Mean for the Broader AI Infrastructure Buildout?

The collaboration signals a shift in how the industry approaches AI data center construction. Rather than treating each facility as a unique engineering challenge, companies are moving toward standardized, repeatable designs that reduce complexity and risk. This approach mirrors how other industries, such as automotive manufacturing, have benefited from standardized platforms and modular architectures.

NVIDIA's Vladimir Troy, Vice President of AI Infrastructure, emphasized the importance of this coordination. "AI factories demand tightly coordinated power, cooling and compute infrastructure to operate efficiently at scale. By aligning with the NVIDIA Omniverse DSX Blueprint, Trane Technologies and Eaton are helping customers reduce complexity and accelerate deployment of next-generation AI data centers," he stated.

Vladimir Troy, Vice President of AI Infrastructure

The reference design also addresses a practical challenge: data center operators need to plan infrastructure that can scale reliably. By providing a unified, pre-coordinated system, Trane and Eaton give customers confidence that they can expand capacity without redesigning core systems or encountering unexpected compatibility issues.

As AI compute demands continue to surge, the efficiency gains from this collaboration could compound across hundreds of new facilities. A 15% improvement in energy efficiency across the projected tripling of global data center capacity by 2030 represents substantial reductions in power consumption, operational costs, and environmental impact.