FM newsroom – energy management, data centres. As global data consumption surges, the race to power and future-proof data centres is accelerating, with AI, self-generation, and adaptive reuse leading the transformation.
Power is the lifeblood of digital infrastructure—and no sector feels this more acutely than data centers. According to JLL, data centres consume 2% of global electricity. Compounding the challenge, overall global energy demand is expected to double within the next five years, reaching approximately 100 gigawatts, tells Sean Farney, Vice President of Data Centre Strategy and Innovation at JLL, to FacilitiesNet
This explosion in energy demand is driven by the digital economy’s insatiable appetite for data. We are living in the Fourth Industrial Revolution, where data has overtaken currency as a core element of value. From cloud services to social media platforms, everything is underpinned by data centres. With every new generation becoming more data-dependent, operators are under increasing pressure to scale—and fast.
The Power Challenge
As the need for data services escalates, so too does the competition for electricity. However, traditional power grids are already strained, and the risk of outages due to extreme weather or climate-related disruptions is growing. For data centres requiring uninterrupted uptime, reliable power is no longer optional—it’s existential.
Without a dependable energy source, the digital backbone of society could face dangerous instability. As such, operators are exploring groundbreaking ways to decouple from traditional energy infrastructure.
Return to Self-Generation
One promising path forward takes a page from history: self-generation. A century ago, many businesses produced their own power before the modern grid was established. That model may be making a comeback.
Hyperscale data centre builders invest heavily in primary power generation using technologies like natural gas turbines, fuel cells, and linear generators. This approach allows them to build wherever they want, free from grid limitations. By making power onsite, data centres could eventually treat the grid as a backup rather than a primary source.
The Promise of Small Modular Reactors (SMRs)
Another forward-looking solution comes from the nuclear sector: small modular reactors (SMRs). According to the International Atomic Energy Agency, SMRs are advanced nuclear reactors that can generate up to 300 megawatts per unit—roughly a third of what traditional nuclear plants produce.
Because SMRs are modular, they can be built in factories and deployed wherever needed, dramatically improving flexibility and deployment speed. If commercialization accelerates, SMRs could offer a reliable, sustainable, and scalable source of primary power for data centers—especially in remote or power-constrained regions.
Adaptive Reuse: A New Model for Data Centres
Beyond high-tech power solutions, the industry is also exploring smarter real estate strategies. One emerging model is adaptive reuse—transforming existing properties like malls or factories into smaller, localized data centers. These facilities benefit from already-installed power connections and offer faster deployment and sustainability gains.
This trend is closely tied to the rise of AI inferencing—the process of deriving insights from previously trained data models. As AI moves from learning to real-time decision-making, enterprises need data centres closer to users. Smaller, distributed centres located in repurposed spaces meet this need while minimising new energy and infrastructure demands.
Capital Is Flowing Into Self-Deployed Energy
As the urgency to reduce reliance on strained power grids grows, SMRs and other self-generation technologies attract massive interest from institutional investors, infrastructure funds, and hyperscale tech giants. If the regulatory and technological barriers continue to fall, the market could witness a significant redirection of capital away from traditional grid expansion and toward on-premises, independently controlled power generation.
Why? Because these solutions solve multiple high-priority issues:
- Energy reliability and independence
- Faster deployment timelines
- Sustainability mandates and ESG goals
- Cost stability over time
Data centre operators, once entirely dependent on grid connectivity, may soon become power producers themselves. That shift not only redefines the business model but fundamentally reshapes the energy economy, turning data centres into hubs of both information and independent energy generation.
AI Is Reshaping the Data Centre Blueprint
The shift to AI has forced a fundamental rethinking of data centre operations. Traditional methods that served the industry for the past 25 years are being upended. AI workloads require entirely different architectures—from higher power density and liquid cooling to more localised deployment strategies.
Companies like Nvidia are leading the push for smaller, high-efficiency data centres optimised for AI processing. In response, major players like Microsoft, Google, Meta, AWS, and leading colocation providers such as CyrusOne are investing in research to innovate faster than ever before.
