A Deluge of A.I. Computing Power Is About to Come Online, Fueling Major Leaps
An unprecedented surge in AI computing power is driving the construction of hundreds of new data centers globally, with significant projects in the US. This expansion, while fueling rapid AI advancements, is sparking backlash over environmental impacts, electricity consumption, and water usage, becoming a contentious issue in US midterm elections. US companies are leading this massive infrastructure build-out with projected investments topping $750 billion this year.
The world is on the brink of an unprecedented surge in AI computing power, with the number of specialized AI chips expected to double every nine months, reaching 200 million by late 2028. This rapid expansion is driving the construction of hundreds of major data centers globally, including significant projects across the United States. Technologists like Rob Wachen, co-founder of microchip firm Etched, describe this as the "largest scale infrastructure build-out in the history of humanity," while Peter DeSantis of Amazon noted the company's computing capacity has doubled since 2022 and will double again by next year.
This "deluge of computing power" is expected to unlock major AI breakthroughs in drug discovery, robotics, and everyday applications, driven by the belief that more data and compute power lead to more advanced AI systems. Industry leaders such as Anthropic CEO Dario Amodei and Google DeepMind head Demis Hassabis foresee AI performing vast amounts of white-collar work and ushering in a technological revolution. However, this aggressive build-out has sparked significant backlash, leading to protests in communities concerned about environmental harm, rising electricity prices, and increased water consumption, making data centers a key issue in upcoming US elections.
US companies, including Amazon, Google, Microsoft, Meta, and Oracle, are leading this global AI infrastructure race, projected to spend $750 billion this year alone. Specific US examples include Amazon's complex in New Carlisle, Indiana; xAI's new multibillion-dollar data center in Memphis, Tennessee; and Microsoft's expansion of what could be the world's most powerful data center in Wisconsin. OpenAI also previously used a data center near Des Moines, Iowa. This US dominance is occurring despite concerns from economists like Oxford University's Carl Benedikt Frey about potential economic bubbles and geopolitical tensions with countries like China, which is heavily investing in its own AI infrastructure despite US export controls.
The demand for computing power is also driving a massive increase in semiconductor production, with Nvidia's H100 equivalent chips central to this expansion. The energy requirements are immense; data centers consumed 64 gigawatts globally last year, a figure expected to quadruple by 2030, raising significant environmental and economic concerns.