E D I T O R ' S Q U E S T I O N
TATE CANTRELL, CTO, VERNE
Sustainability in data centres has evolved from a peripheral consideration to a defining element of long-term infrastructure strategy. Yet, as AI drives exponential demand for high-performance computing, power availability is becoming a critical constraint – and many operators are turning to carbon-based fuels to meet immediate needs.
At Nvidia’ s annual GPU Technology Conference( GTC 2025), CEO Jensen Huang predicted that data centre buildouts will reach US $ 1 trillion by 2028 – underscoring the scale of the shift.
As the industry adapts to exponential growth in AI and high-performance computing, the focus is sharpening on three key areas: facility design, access to energy, and the growing influence of regulation and public accountability.
From a design perspective, the need to accommodate increasingly dense workloads is accelerating the adoption of new technologies. With AI server rack densities already exceeding 120kW – and future designs expected to reach 400 to 600kW – traditional cooling methods are no longer sufficient.
Operators are responding with innovations such as liquid cooling, improved airflow systems, energyefficient hardware and modular construction – solutions that not only reduce energy consumption but also increase scalability.
At the same time, the availability of renewable power and robust grid infrastructure is becoming a strategic differentiator. Locations that offer both are better positioned to support longterm sustainable growth. The Nordics stand out in this regard. Countries like Iceland, Sweden, Norway and Finland offer access to diverse renewable energy sources, including geothermal, hydro and wind energy, as well as stable grid capacity and naturally cool climates – all of which significantly reduce the carbon footprint of data centre operations. As a result, the region is emerging as a preferred destination for high-performance workloads with sustainability requirements.
But alongside innovation and smart site selection, the industry is facing a new kind of pressure: an AI arms race that’ s rapidly intensifying energy demand. As companies compete to train ever larger, more advanced models, AI workloads are pushing power consumption to new heights. AI infrastructure, in particular, brings a sharp increase in energy consumption – raising concerns that it is extending the life of coal or gasfired power plants, or necessitating new fossil-based generation to meet demand. This is prompting closer scrutiny from regulators and the public, with calls for greater transparency, carbon accountability and cleaner energy procurement.
The rapid acceleration of AI has, in many cases, put a pause on the industry’ s momentum toward holistic sustainability practices. Yet as energy constraints tighten, the strategic value of renewables is becoming clearer. Regions with stable, renewable-powered grids, such as the Nordics, are emerging as critical enablers of large-scale, sustainable AI infrastructure. In the long term, aligning performance with environmental stewardship won’ t just be an ethical imperative – it will be a competitive advantage. �
34 www. intelligentdatacentres. com