Telcos Gain AI-Ready Data Centre Model from Siemens & nVent

Hyperscale infrastructure built for AI workloads demands a new kind of data centre.

Siemens and nVent now offer a blueprint that targets this shift directly, unveiling a reference architecture engineered to help operators roll out 100MW hyperscale campuses optimised for Nvidia AI infrastructure.

For telcos, this presents a modular, high-density model focused on speed, energy efficiency and system integration.

Telco operators facing pressure to scale quickly, manage rising rack densities and support AI-intensive services now get a model designed to handle exactly those conditions. 

The reference architecture supports Nvidia DGX SuperPOD systems using DGX GB200 hardware and offers a combined approach using Siemens' power systems and nVent’s liquid cooling portfolio.

Ciaran Flanagan, Global Head of Data Center Solutions at Siemens, explains the value of the partnership: “This reference architecture accelerates time-to-compute and maximises tokens-per-watt, which is the measure of AI output per unit of energy. It’s a blueprint for scale: modular, fault-tolerant and energy-efficient.”

Built for telcos targeting 100MW AI capacity

As telco operators ramp up investments in AI-based services, the Siemens–nVent model aims to deliver a standardised route to rapid deployment. 

Built for 100MW facilities, the architecture addresses power consumption, thermal control and operational risk through integration. 

It connects Siemens’ industrial-grade energy systems with nVent’s cooling technologies to support large-scale compute clusters aligned with Nvidia’s AI design standards.

Sara Zawoyski, President of nVent Systems Protection, says: “We have decades of expertise supporting customers’ next-generation computing infrastructure needs. 

“This collaboration with Siemens underscores that commitment. The joint reference architecture will help data centre managers deploy our cutting-edge cooling infrastructure to support the AI buildout.”

Telcos building out AI-enabled 5G infrastructure or edge-to-core services need systems that handle increased compute intensity. 

Liquid-cooled racks and dense hardware clusters built around the Nvidia DGX platform require high-capacity power, integrated control systems and precision thermal management. 

The reference model addresses these technical requirements with built-in scalability and consistency across multiple deployment environments.

Electrical resilience and liquid cooling for AI compute

Ciaran says the blueprint connects multiple parts of the supply chain: “Together with nVent and our broader ecosystem of partners, we’re connecting the dots across the value chain to drive innovation, interoperability and sustainability, helping operators build future-ready data centres that unlock AI’s full potential.”

The partnership’s design allows for modular scaling, which is essential as telcos explore AI workloads across centralised data centres and distributed edge locations. 

Siemens contributes a full stack of electrical systems: medium and low voltage distribution, intelligent automation, grid-aware control and AI-driven energy management tools. These tools are all aimed at ensuring uptime and operational consistency as rack densities rise.

Built-in IoT capabilities and cloud-based analytics help operators monitor and optimise system performance.

Voltage regulation and real-time fault detection provide electrical resilience essential to uptime as workloads shift towards GPU-heavy AI processing.

nVent’s contribution focuses on liquid cooling technologies tailored to high-density compute.

Shaped by partnerships with chip makers and OEMs, its systems deliver thermal control that adjusts to rising thermal design power (TDP) ratings without large-scale infrastructure changes.

The cooling architecture integrates direct liquid cooling, coolant distribution and modular thermal layouts. For telcos managing AI compute, the combination helps lower total cost of ownership while supporting the higher wattage loads associated with modern AI chips.

Reference design as standardised path for telcos

Reference architectures give telcos a clearer path forward in a fast-moving AI market. By offering a tested, deployment-ready model aligned with Nvidia’s system requirements, Siemens and nVent aim to eliminate fragmented engineering and shorten the overall build timeline.

Operators gain a framework for interoperability between vendors, easing the integration of new equipment and reducing the complexity that often slows hyperscale projects. 

For telcos looking to future-proof networks while managing energy performance, the design offers a route to AI readiness without starting from scratch.

The Siemens and nVent collaboration reflects an industry need for joined-up design.

With power availability, heat rejection and rapid scaling now core concerns for telco data centres, a unified approach combining power systems and cooling infrastructure becomes essential. 

Telcos aiming to deliver AI-based services with minimal latency and high reliability can now turn to a solution engineered for both performance and speed of deployment.