“The evolution of AI is still in its infancy, and we are just scratching the surface of its potential,” says Youlian Tzanev, Co-founder & CSO, NexGen Cloud. With fears of energy use booming due to AI developments, green data centres could offer respite to help manage the load and diminish unsustainable operations in the future.
The G7 Summit in Apulia, Italy, in June 2024 covered some of the most pressing issues of our time. It is telling that both AI safety and energy security were on the agenda. These issues are far from mutually exclusive, with energy security and the green transition potentially coming under threat due to the power demand of AI.
Despite growing concerns, and these topics being debated almost simultaneously, their overlap was not discussed at length. Yet the question of how to sustainably power the AI revolution remains a hot topic within both the energy and data centre industries.
Artificial Intelligence (AI) is transforming industries and enhancing our daily lives with a wide range of diverse applications. From optimising supply chains, powering chatbots in customer service and managing traffic in smart cities, AI is already having a huge impact. But the evolution of AI is still in its infancy, and we are just scratching the surface of its potential. As the technology continues to advance, its capabilities will expand, leading to profound changes across all sectors of society.
The potential of AI to revolutionise society cannot be overstated. It promises to enhance efficiency, productivity and innovation in numerous fields such as healthcare, finance, manufacturing and beyond. For instance, AI algorithms can now accurately diagnose medical conditions, manage financial portfolios with unparalleled precision and optimise manufacturing processes to reduce waste. As AI systems become more sophisticated, they will continue to unlock new opportunities, driving economic growth and improving quality of life.
The energy challenge of AI and IT infrastructure
Despite its undoubted benefits, AI comes with significant challenges, particularly in terms of energy consumption. AI algorithms, especially those related to Deep Learning and neural networks, require substantial computational power. Training a single AI model can sometimes consume as much energy as several households use in a year.
The energy demands stem from the need for massive data processing capabilities and continuous operation of High-Performance Computing (HPC) systems, which require considerable electricity to function and maintain. This intensive energy use is driven by the complexity of AI models and the large datasets needed for training and operation. This significant power usage is exacerbated by the need for cooling systems to prevent overheating and to maintain optimal operating conditions for servers.
According to the International Energy Agency (IEA), data centres accounted for about 1% of global electricity demand in 2020. As the digital economy grows, so does the energy consumption of IT infrastructure.
As AI technologies proliferate, the energy demands of IT infrastructure are expected to skyrocket. The IEA predicts that electricity consumption from data centres could double by 2026, reaching levels equivalent to Japan’s total electricity consumption. In the UK, the CEO of the National Grid, John Pettigrew, has warned that AI could increase power consumption by 500% within the next decade.
This surge in energy requirements poses a significant challenge to sustainable development, as it risks outpacing the growth of renewable energy sources. Factors driving this increase include the expansion of AI applications, the growing number of data centres and the rising demand for cloud services and IoT devices.
Pioneering sustainable AI solutions
Addressing the energy demands of AI necessitates innovative solutions. One such approach is the development of green data centres. Europe is at the forefront of this initiative, with NexGen Cloud constructing Europe’s first AI Supercloud, a sustainable AI cloud designed to minimise environmental impact.
While traditional data centres have been located close to large urban areas to minimise latency for users, the same prerequisites do not apply to AI. We now have the option to build AI data centres in locations with abundant green energy, such as near hydroelectric plants, or to support them with dedicated microgrids that are entirely powered by renewable energy.
The new generation of AI systems that will power the AI revolution are also becoming increasingly energy efficient, using liquid-cooling of the GPUs that can be used to recover energy.
By leveraging renewable energy sources, optimising energy efficiency and implementing advanced cooling technologies, green data centres can significantly reduce the carbon footprint of AI operations. This can be a blueprint for how integrating sustainable practices can meet the high energy demands of modern AI applications.
As a NVIDIA Elite Partner, NexGen Cloud uses cutting-edge hardware that maximises performance while minimising energy consumption. Additionally, NexGen Cloud’s data centres are strategically located to take advantage of renewable energy sources, such as in Norway, where its data centre is powered by 100% hydroelectric power.
Across the industry, data centre owners are looking to harness renewable energy to power the energy appetite of AI. Recently, Google, Amazon and Microsoft made headlines for petitioning the Irish Government to allow them to construct their own power cables to connect data centres directly to renewable energy sites.
The future of sustainable data centres is bright, with advancements in technology and growing awareness of environmental issues driving progress. Innovations in energy-efficient hardware, renewable energy integration and cooling technologies will continue to enhance the sustainability of data centres.
In addition to technological developments, regulatory frameworks and industry standards are increasingly favouring green practices, incentivising companies to adopt sustainable solutions. Emerging technologies, such as AI-driven energy management systems and hydrogen fuel cells, are set to revolutionise the industry. NexGen Cloud seeks to be a part of this future, demonstrating that it is possible to meet the growing demands of AI without compromising on sustainability.
AI itself also stands ready to contribute to the energy transition through the acceleration in innovation that only it can provide. As an example, an electric vehicle battery typically takes 20 years from the inception of the research programme through to commercial launch. Using AI, it is forecast that the analysis of the permutations and combinations that are required to perfect a battery for commercial use could be reduced to two or three years through fully automated testing.
As AI continues to advance, its energy consumption is expected to grow exponentially. Green data centres offer a viable path forward, combining cutting-edge technology with renewable energy to meet the demands of the future.
By investing in sustainability today, we can ensure that AI continues to drive progress without compromising the planet’s resources. Collaboration between tech companies, policymakers and environmental organisations is also essential to ensure that the Digital Transformation driven by AI remains aligned with our sustainability goals.
