SUPERNEWS

Macro

AI Data Centers to Boost US Power Demand, Goldman Eyes 16 Stocks

Goldman Sachs forecasts a 15% CAGR in AI data center power demand, driving $50 billion in investments by 2030.

By Bill Bullington

ᐧ

4/29, 13:08 EDT
S&P 500
iShares 20+ Year Treasury Bond ETF
iShares 7-10 Year Treasury Bond ETF
First Solar, Inc.
NextEra Energy, Inc.
NVIDIA Corporation
Oracle Corporation
Vertiv Holdings, LLC
article-main-img

Key Takeaway

  • Goldman Sachs forecasts a 15% CAGR in data center power demand from 2023-2030, raising data centers' US power demand share to 8%.
  • This surge necessitates an additional 47 GW of power generation, driving $50 billion in capital investment across gas and renewable sources.
  • Goldman Sachs identifies 16 "buy" rated stocks across various sectors poised to benefit from the increased electricity demand for AI data centers.

AI Data Centers Fuel Power Demand Surge

Goldman Sachs predicts a significant increase in U.S. power demand due to the growth of AI data centers, forecasting a 15% compound annual growth rate in data center power demand from 2023 to 2030. This surge is expected to raise data centers' share of total U.S. power demand to 8% by 2030, up from approximately 3%. The investment bank highlights this as a growth not seen in a generation, with U.S. electricity demand historically growing at less than 0.5% annually over the last 20 years. To meet this demand, an estimated 47 GW of additional power generation capacity will be necessary, likely sourced from 60% gas and 40% renewable energies, driving about $50 billion in capital investment by 2030.

Investment Opportunities in Utilities and Renewables

Goldman Sachs has identified 16 stocks across various sectors, including Utility, Clean Technology, Midstream, Energy Services, Industrials, and Industrial Tech, that are positioned to benefit from the anticipated surge in power demand. Companies like Vertiv Holdings, NextEra Energy, and First Solar are among those expected to see significant growth due to their involvement in thermal cooling, power management, renewable energy generation, and utility-scale solar farms. This basket of stocks represents a broad spectrum of industries poised to capitalize on the increased electricity needs of AI data centers.

Nvidia's Role in the AI Revolution

Nvidia's critical role in powering AI technologies has led to a projected 20% increase in chip spending by tech giants, significantly impacting its stock performance with a nearly threefold increase in 2023. Oracle and Northern Data's pivot towards AI, with substantial investments in Nvidia chips for cloud services, underscores a broader market trend towards AI technology. The Spear Alpha ETF, heavily invested in Nvidia, has seen a 60% gain, reflecting the bullish market sentiment towards AI-focused companies.

Oracle and Northern Data's AI Strategies

Oracle's integration of generative AI into its cloud infrastructure aims to enhance productivity and reduce costs, supported by a 25% increase in cloud revenue. Northern Data's strategic shift from bitcoin mining to AI cloud services, with a significant investment in Nvidia AI chips, positions it as a major player in the AI cloud services sector by 2024. These moves by Oracle and Northern Data highlight the competitive landscape in AI and cloud services, with companies investing heavily to capitalize on the AI demand.

Street Views

  • Carly Davenport, Goldman Sachs (Bullish on utilities, renewable energy generation, and industrial sectors):

    "U.S. power demand (is) likely to experience growth not seen in a generation. Not since the start of the century has US electricity demand grown 2.4% over an eight-year period, with US annual power generation over the last 20 years averaging less than 0.5% growth." "Analysts estimate that about 47 GW of additional power generation capacity will be necessary to accommodate the growth in U.S. data center power demand by 2030." "There could be meaningful upside to our base case if appetite for purchase and utilization of servers is unconstrained." "There could be downside to our base case if power efficiency is higher than expected or if power/compute speed efficiencies lead to fewer servers purchased than expected."