Trane Technologies (NYSE:TT) and Eaton (NYSE:ETN) have launched a joint reference design that brings together thermal management and electrical infrastructure for the next generation of artificial intelligence data centres.
The collaboration combines the companies’ respective cooling and power-management technologies into a coordinated architecture intended to support increasingly power-dense AI facilities. The medium-voltage design has been developed to align with the Nvidia DSX AI Factory Reference Design platform.
Integrated system targets higher efficiency and lower costs
Trane Technologies and Eaton said their integrated approach could deliver energy-efficiency improvements of up to 15% compared with conventional low-voltage architectures.
The design could also reduce copper requirements by as much as 80% and lower installation costs by up to 30%, potentially addressing some of the infrastructure challenges associated with building increasingly large and power-intensive AI data centres.
The reference architecture has been incorporated into both the Trane Continuum Rubin DSX and Eaton Beam Rubin DSX platforms.
Power and cooling systems designed to work together
The architecture has also been designed for compatibility with the Nvidia Omniverse DSX Blueprint for AI data centres.
Under the collaboration, Eaton’s technology provides power distribution capabilities for the Trane platform. The companies’ electrical and thermal management systems are also designed to exchange operational data, allowing the infrastructure to respond dynamically as power and cooling requirements change.
“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 and Chief Technology and Sustainability Officer at Trane Technologies.
The coordinated approach is intended to simplify the design and deployment of facilities supporting increasingly demanding AI workloads, where cooling capacity and electrical infrastructure are becoming closely interconnected.
Nvidia highlights potential for faster AI infrastructure deployment
Vladimir Troy, Vice President of AI Infrastructure at Nvidia, said the integrated architecture could help customers reduce complexity while accelerating the deployment of AI data centres designed around the Nvidia Omniverse DSX Blueprint.
The collaboration comes as AI infrastructure developers face growing pressure to accommodate higher computing densities while controlling energy consumption, construction costs and deployment times.
By designing power distribution and thermal management as interconnected systems rather than separate infrastructure layers, Trane Technologies and Eaton are seeking to provide operators with a more streamlined approach to scaling AI capacity.
Architecture prepared for liquid cooling and DC technologies
The reference design has also been developed with future infrastructure changes in mind.
Trane Technologies and Eaton said the architecture can adapt to emerging liquid-cooling technologies as these become more widely used to manage the heat generated by increasingly powerful AI computing systems.
It is also designed to accommodate direct-current architectures, providing flexibility as data centre electrical systems evolve.
The companies’ joint approach therefore targets both current AI infrastructure requirements and the higher-density power and cooling technologies expected to become more prominent as the data centre industry expands.
Trane Technologies stock price
Eaton Corporation stock price
