Rising AI Demand in Data Centers Spurs $50 Billion US Infrastructure Investments

AI to fuel a 160% surge in data center power demand by 2030, driving significant environmental and infrastructure investment challenges.

By Max Weldon

5/15, 10:18 EDT
Verizon Communications Inc.

Key Takeaway

  • Goldman Sachs Research projects a 160% increase in data center power demand by 2030 due to AI, raising global consumption to 3-4%.
  • The surge in power demand necessitates $50 billion in US infrastructure investments and nearly €1.7 trillion in Europe for grid upgrades and renewable energy.
  • Verizon views AI as crucial for network demand growth, planning significant infrastructure investments to support the anticipated data traffic surge.

Rising Power Demand in Data Centers

The global shift towards artificial intelligence (AI) is set to significantly increase the power consumption of data centers, with Goldman Sachs Research projecting a 160% growth in data center power demand by 2030. Currently, data centers consume 1-2% of the world's overall power, a figure expected to climb to 3-4% by the decade's end. This surge is attributed to the expanding use of AI, which demands considerably more electricity than traditional computing tasks. For instance, a single ChatGPT query requires nearly ten times the electricity of a Google search. Despite advancements in computing efficiency, the broad adoption of AI technologies is anticipated to add approximately 200 terawatt-hours per year to the global power consumption of data centers by 2030, with AI expected to account for about 19% of data center power demand.

Environmental and Economic Implications

The increased power consumption by data centers is not without its consequences, particularly in terms of carbon dioxide emissions and social costs. Analysts predict that the carbon emissions from data centers could more than double between 2022 and 2030, posing a significant environmental challenge. The social cost of these emissions is estimated to be between $125-140 billion. Despite these challenges, technology companies remain optimistic about reducing energy intensity through investments in renewable energy sources and emerging nuclear generation capabilities. AI itself is seen as a potential tool for driving innovations that could lead to energy efficiencies and reductions in emissions across various sectors, including healthcare, agriculture, and education.

Investment in Infrastructure and Renewable Energy

To accommodate the rising demand for electricity, substantial investments in power infrastructure will be necessary, particularly in the United States and Europe. In the US, where power demand growth has been stagnant over the past decade, data centers are expected to consume 8% of the country's power by 2030, necessitating an estimated $50 billion investment in new generation capacity. Europe faces a similar challenge, with power demand potentially growing by 40-50% between 2023 and 2033 due to data center expansion and increased electrification. This will require nearly €1.7 trillion ($1.861 trillion) in investments for upgrading the power grid and expanding renewable energy sources. Additionally, the demand for critical materials like copper and uranium, essential for clean energy technologies, is expected to rise, presenting investment opportunities in these resources.

Verizon's AI-Driven Growth Strategy

Verizon Communications Inc. identifies AI as a key driver for future network demand growth, moving beyond the era dominated by video streaming. The company's consumer division chief, Sowmyanarayan Sampath, highlights AI's potential to double network demand over the next five years. Verizon is leveraging AI in its customer service operations, employing tools that analyze human sentiment and tone during calls, and automating routine inquiries. This strategic focus on AI is part of Verizon's broader infrastructure investment, including significant spectrum acquisitions, to prepare for the anticipated surge in data traffic.

Street Views

  • Carly Davenport, Alberto Gandolfi, and Brian Singer, Goldman Sachs Research (Neutral on data center power demand):

    "Conversations with technology companies indicate continued confidence in driving down energy intensity but less confidence in meeting absolute emissions forecasts on account of rising demand."