Google's Historic Deal to Power AI Data Centers with Nuclear Energy

In a groundbreaking move, Google has become the first company to strike a deal for powering its data centers with energy from small modular reactors (SMRs). The tech giant has placed an order with California-based Kairos Power for six or seven reactors, with the first expected to go live by 2030 and the others following by 2035.

As Google’s AI capabilities grow, so do its data centers' energy demands. This deal is not just about meeting those needs but also aligning with Google's long-standing sustainability goals. Nuclear power offers a stable, low-carbon solution, providing round-the-clock electricity essential for powering AI infrastructure. Michael Terrell, Google’s senior director for energy and climate, noted that AI is driving scientific breakthroughs and economic growth but also placing unprecedented pressure on the energy grid.

This move marks a major milestone in the tech sector's increasing reliance on nuclear energy for power-intensive operations. Microsoft recently announced a similar deal, sourcing energy from Pennsylvania's Three Mile Island, while Amazon acquired a nuclear-powered data center from Talen Energy earlier this year.

Although the financial terms of Google’s agreement are undisclosed, the company has committed to purchasing 500 megawatts of power from Kairos. The start-up, founded in 2016, is currently constructing a demonstration reactor in Tennessee, which is slated for completion by 2027. Kairos’ CEO, Mike Laufer, expressed confidence in delivering the reactors on time and on budget.

SMRs, which can generate up to 300 megawatts, offer a promising alternative to traditional, larger reactors. They are seen as faster and cheaper to build, though some critics argue that SMRs may face challenges related to costs and scalability compared to renewables like wind and solar.

Despite these challenges, SMRs are gaining traction as a key element in the global shift toward clean energy. Their smaller size and reduced cooling needs make them adaptable to a wider range of locations, and companies like Rolls-Royce are actively competing to develop SMR technology in the UK as part of efforts to revitalize the nuclear industry.

As countries and companies strive to reduce dependence on fossil fuels, SMRs are emerging as a flexible, game-changing solution. However, the debate continues, with proponents championing nuclear energy's role in the clean energy transition, while environmentalists argue for a greater focus on proven renewable sources like offshore wind.

Why Nuclear for AI? Google’s Shift to Small Modular Reactors

As artificial intelligence (AI) evolves, the demand for computational power is skyrocketing. AI models and cloud services hosted in data centers consume massive amounts of energy, pushing tech companies like Google to seek reliable, scalable power solutions. Nuclear energy, particularly Small Modular Reactors (SMRs), offers a key advantage—unlike wind and solar, it provides consistent, round-the-clock power.

SMRs are more cost-effective, quicker to build, and safer than traditional reactors, making them ideal for the energy demands of AI-driven infrastructure. Google's investment in SMRs underscores its commitment to sustainability, allowing it to scale AI operations while minimizing its carbon footprint.

By adopting SMR technology, Google—and potentially the wider tech industry—can balance the power demands of AI expansion with the need for environmentally responsible energy.