As AI Companies Race for Power, Amazon and Google Have the Lead
Amazon.com Inc. Chief Executive Andy Jassy has identified power as the “single biggest constraint” in the company’s cloud and artificial-intelligence operations, according to a statement. This highlights the growing importance of energy availability for tech giants competing in the AI sector. Amazon, the world’s largest cloud provider, operates self-built data centers in the U.S. that consume up to 9 gigawatts of power, according to Aterio, a data provider. That level of consumption matches the generation capacity of North Dakota.
Microsoft Corp. and Alphabet Inc. each have self-built data centers using up to 5 gigawatts, while Meta Platforms Inc. operates facilities with about 4 gigawatts, Aterio data shows. Most hyperscalers rely on self-built infrastructure rather than rented space, according to the report. Amazon is projected to add the most data center and power capacity in the U.S. through 2030, but Google is expected to expand at the fastest rate. Including leased capacity, Google could narrow its gap with Amazon by that year, Aterio estimates.
Why the Power Constraint Matters
The focus on power reflects the escalating energy demands of AI workloads, which require vast computational resources. Data centers account for about 2% of global electricity use, according to the International Energy Agency. For companies like Amazon, securing reliable and cost-effective energy sources could determine their competitive edge in AI development. The scale of Amazon’s existing infrastructure gives it a head start, but rapid growth from rivals like Google may challenge its dominance.
What May Happen Next
Investors tracking the AI sector may closely monitor how companies secure additional power. Amazon’s extensive data center network could allow it to maintain a lead in total capacity, but Google’s faster growth rate might enable it to catch up in efficiency or innovation. The shift toward leased capacity could also reshape the data center market, with third-party operators playing a larger role in meeting demand.
Frequently Asked Questions
What did Amazon’s CEO identify as the biggest constraint in its cloud and AI business? The CEO stated that power is the “single biggest constraint.”
How much power do Amazon’s U.S. data centers consume? They use up to 9 gigawatts, according to Aterio.
How is Google expected to close its gap with Amazon by 2030? By adding capacity at the fastest rate and including leased data center space, according to Aterio.
How might the balance of power between tech giants evolve as AI demand grows?