Sustainable Data Centers: Balancing AI Growth and Environmental Impact

Published on March 12, 2025 by Pradhumn Kumar

The demand for sustainable data centers has increased exponentially, driven by the rapid growth of AI data center infrastructure. Companies like Microsoft, Amazon and Google are responding to this demand, announcing multibillion-dollar plans to add or expand existing data centres across the US. Companies are also acknowledging the significant environmental challenges posed by rapidly expanding AI infrastructure. This presents a stark paradox: AI's potential to drive sustainability and innovation is undermined by its own infrastructure's growing carbon footprint.

With energy consumption projected to triple by 2030 and water usage straining local resources, the tech giants are pioneering sustainable solutions by investing in renewable energy, energy-efficient algorithms, and innovative cooling methods to mitigate climate change. By embracing these sustainable practices, companies can harness AI's potential to drive innovation while minimising its ecological footprint. Strong market demand, innovative technologies and strategic partnerships will drive future growth and profitability for companies operating and investing in such sustainable data centres.

The Rapid Growth of AI Data Centers

The growth of AI data centers is fueled by booming internet, cloud computing, and AI activities. Currently, there are over 8,000 data centres globally, with c.33% located in the US, c.16% in Europe and c.10% in China. Market research reports suggest that the data centre market size is set to grow at a CAGR of 11%-12% over the next 7-8 years, worth well over USD400bn-500 bn.

Microsoft and BlackRock formed a USD100bn partnership to develop data centres globally, while Google has committed to investing billions of dollars in data centres. Meanwhile, Amazon has pledged to spend USD148bn over 15 years to build data centres worldwide.

Powering AI Data Centers: Rising Energy Demands

Rapid AI data center growth has triggered a surge in power demand, highlighting the need for sustainable data centers that leverage renewable energy and energy-efficient technologies. Data centres rely on enormous amounts of electricity to operate and cool systems. The International Energy Agency (IEA) expects data centre electricity consumption to grow to 260TWh by 2026 from c.200TWh in 2022 (c.4% of total power consumed). McKinsey estimates it will reach 600TWh by 2030, accounting for 11-12% of total power consumed, from its estimate of c.290TWh in 2026.

Separately, US electric utility companies are benefiting from growing demand. Dominion Energy Inc., mentioned in its investor day in March 2024 that accelerating data centre expansion was driving significant demand growth. The company estimates that the rack power density of a typical data centre without AI demand was only 6-12kW, while it was 26-80kW for data centres supporting AI in training phase. Nvidia’s latest chip, the GB200, may require rack densities of up to 120kW. Other utilities, such as Oncor Electric Delivery Co. and Ohio Power Co., are also poised to benefit. McKinsey states that at least twice the data centre capacity built since 2000 would have to be built in less than a quarter of the time (over the next 6-7 years).

Water Usage in Sustainable Data Centers: Strategies for Conservation

The data centre industry is a substantial user of water resources, for server cooling and electricity generation. In the US, data centres are among the top ten industries driving water consumption. Google’s data centre in Dalles, Oregon consumed 355m gallons of water in 2021, nearly one-third of the town’s annual supply. Microsoft and Google reported significant water usage in their data centres, highlighting the need for sustainable water management practices.

Sustainable Solutions for AI Data Center Growth

Two primary strategies can help minimise the carbon footprint of AI data center growth: transitioning to sustainable data centers powered by renewable energy and implementing energy-efficient algorithms.

Energy-efficient algorithms can deliver equivalent performance with reduced power consumption.
Transitioning data centres to renewable energy sources like solar, wind or hydroelectric power can yield substantial emissions reductions.

Case studies of sustainable practices:

Energy Efficiency Optimization:

Google has achieved remarkable success in optimising energy efficiency, reducing non-computing energy usage to a mere 10%. Looking ahead, Google has set an ambitious target of powering all its global data centres entirely on 24/7 carbon-free energy by 2030, a vision that promises to reshape the industry's sustainability standards for data centres worldwide. Similarly, Microsoft has forged strategic partnerships with local energy providers to harness renewable energy from wind, solar and hydro power, with a bold commitment to powering all its New Zealand data centres with 100% renewable energy.

Innovative Energy Solutions:

Through the pioneering Granular Certificate Trading Alliance, Google and Microsoft are creating a platform for trading zero-carbon energy credits, paving the way for a cleaner energy future. The companies are also joining forces to invest in cutting-edge clean energy sources, including hydrogen, geothermal, battery storage and advanced nuclear technologies.

Small Modular Reactors:

Amazon Web Services plans to utilise Small Modular Reactors or compact nuclear reactors, to reduce carbon emissions and increase energy efficiency, a technology also being explored by Google and Microsoft through partnership with Kairos Power and Constellation Energy, respectively.

Water Footprint Reduction:

Google and Microsoft are reducing their water footprint by exploring alternative cooling methods, such as air cooling and liquid immersion cooling and utilising recycled or non-potable water in cooling systems. In 2023, about one-third of Google's data centres employed air cooling or non-potable water sources. Amazon also aims to become water positive by 2030 through water-efficient data centre cooling systems and restoration initiatives. These efforts prioritise minimising the impact on freshwater resources, promoting sustainable operations and conserving water in data centre operations, supply chains and communities.

Investment Case for Sustainable Data Centres and How Acuity Knowledge Partners Can Help

As the world prioritizes sustainability, investing in sustainable data centers offers a compelling opportunity for long-term growth and profitability. To capitalize on this trend, global investment banks and asset managers require timely and actionable insights. Acuity Knowledge Partners, committed to advancing sustainable business practices through our own ESG initiatives, can help. Our dedicated teams of analysts, content specialists, and MBA/CFA-qualified professionals provide tailored support across a range of activities such as idea generation, financial analysis, thematic research, database building, and regular sector coverage.

Furthermore, our AI-powered research platform, ResearchHub, facilitates the identification of ESG trends and themes, highlighting relevant key performance indicators (KPIs) for actionable intelligence. This approach ensures that you remain at the forefront by delivering comprehensive insights and data-driven recommendations essential for making informed investment decisions in the dynamic field of sustainable data centers. Whether you're looking to identify investment opportunities, assess market trends, or create insightful content, our team is here to support you. Contact us today to discover how Acuity Knowledge Partners can help you navigate the sustainable data center investment landscape and drive business growth.

Sources:

Tags:

AI Data Center Growth Sustainable Data Centers

What's your view?

Thank you for sharing your Comments

Share this on

About the Author

Pradhumn Kumar

Know More

Pradhumn Kumar is a seasoned equity research professional with over a decade of experience. As a part of Acuity Knowledge Partners’ Investment Research team, he covers the US Engineering & Construction sector supporting sell-side clients with research assignments including initiation reports, building and maintaining financial models, earnings support and tracking industry news. Previously, Pradhumn covered the European Food Retail and Luxury Goods sector supporting a bulge bracket Investment bank. He holds an MBA in Finance and has cleared Level II of the CFA Program.