The Soaring Demand for Data
The International Energy Agency (IEA) has sounded the alarm on the rapidly growing energy footprint of the digital economy. According to its latest analysis, global electricity consumption from data centres, artificial intelligence (AI), and cryptocurrencies
is on a path to potentially double by 2030. In 2022, data centres alone consumed an estimated 460 terawatt-hours (TWh) of electricity. The IEA's projections show this figure could surge to over 1,000 TWh in the coming years, an amount roughly equivalent to the entire annual electricity consumption of Japan. This sharp increase marks a pivotal moment, as for years, efficiency gains in computing hardware managed to keep energy demand relatively stable, even as our digital activities grew. That era appears to be over, with the sheer volume of data processing now outpacing technological efficiency improvements.
The Twin Engines: AI and Crypto
Two key technologies are overwhelmingly responsible for this projected surge: artificial intelligence and cryptocurrency mining. The rise of generative AI platforms like ChatGPT has created an unprecedented demand for computational power. Training a single large AI model can consume enormous amounts of electricity, and every query processed requires significantly more energy than a traditional web search. The IEA notes that AI is expected to account for a massive portion of the growth in data centre electricity demand. Simultaneously, the cryptocurrency sector remains a major energy consumer. Bitcoin mining alone consumed around 120 TWh in 2023, and the IEA forecasts the sector's total electricity use will continue to climb, despite efficiency gains in some areas. Together, these energy-hungry operations are fundamentally reshaping global electricity demand patterns.
Putting the Numbers in Perspective
The scale of this energy consumption can be difficult to grasp. A terawatt-hour (TWh) is a massive unit of energy, equivalent to consuming one trillion watts for one hour. The projected 1,000 TWh demand for data centres by the end of the decade is more than a third of the total electricity generated by all the world's nuclear power plants in 2023. This growth is not evenly distributed. The United States and China are the epicentres of this expansion, expected to account for nearly 80% of the global growth in data centre power consumption by 2030. In some regions, like Northern Virginia in the US, the concentration of data centres is so large that it is already straining local power grids. This concentration poses a significant challenge, as the demand is localised and intense, unlike the more distributed energy needs of electric vehicles.
The Efficiency Paradox
For years, the tech industry has made incredible strides in energy efficiency. Innovations in server design, software optimisation, and cooling technologies have allowed data centres to do more with less. Google, for example, reports its data centres deliver over three times the computing power for the same amount of energy compared to five years ago. This is measured by a metric called Power Usage Effectiveness (PUE), where a lower number is better. While the industry average PUE is around 1.55, leading hyperscale operators achieve PUEs below 1.2. However, the explosive growth in demand, particularly from AI, is now overwhelming these efficiency gains. The IEA estimates that while the energy needed for a single AI task is falling, the sheer number of users and applications is growing so fast that overall consumption continues to rise sharply.
The Challenge for a Greener Future
This escalating power demand creates a significant challenge for global climate goals. The IEA warns that this expansion is expected to increase the demand for electricity generated from fossil fuels, particularly in the US and China, leading to higher carbon emissions. In response, major tech companies have become the world's largest corporate buyers of renewable energy. However, the 24/7 reliability required by data centres is a challenge for intermittent renewable sources like solar and wind. This has led some tech giants to explore other options, including investments in advanced nuclear technologies like small modular reactors (SMRs) to provide a consistent, carbon-free power supply. Meeting this digital energy challenge will require a multi-faceted approach, combining aggressive efficiency measures, massive investments in renewable energy, and innovations in grid technology and energy storage to power our future sustainably.
















