Trane Technologies and Eaton unveil joint AI data centre reference design

What was announced

Trane Technologies and Eaton have introduced a joint reference design for AI data centres built around the NVIDIA DSX AI Factory Reference Design, combining thermal management and electrical system architecture in a single package. The design is included in the Trane Continuum Rubin DSX and Eaton Beam Rubin DSX platforms, with Eaton providing power distribution for the Trane platform and both companies aligning their systems with NVIDIA DSX.

Efficiency and cost claims

According to the companies, the integrated approach can improve energy efficiency by up to 15%, reduce installation costs by up to 30% and cut copper usage by up to 80% compared with conventional low-voltage designs. Eaton said these figures are estimates and may vary by installation.

Move toward medium-voltage systems

The collaboration is centred on advancing medium-voltage power systems for higher-power-density AI sites, replacing a more fragmented planning process. By linking electrical and cooling systems earlier in planning, the companies said operators can shorten development cycles and reduce deployment complexity for AI-focused facilities.

Alignment with NVIDIA tooling

The reference design works with the NVIDIA Omniverse DSX Blueprint for AI data centres, offering a coordinated way to plan and deliver electrical, thermal and digital control infrastructure. The architecture is designed to adapt as liquid cooling technologies and direct current systems become more widely used, and is intended to let power distribution and cooling systems exchange operational indicators and respond more dynamically to changing demands.

Stated objectives from company leaders

Trane framed the design as a way to give customers a coordinated approach from grid to chip, helping them accelerate deployment and scale. Eaton positioned the work as moving reference designs beyond specifications into repeatable systems that deployment teams can deploy repeatedly, integrating medium-voltage power systems and white space thermal management with Trane's thermal management architecture. NVIDIA described the collaboration as part of a broader effort to standardise infrastructure planning for AI computing environments.