“From Chips to Power: Why the Next AI Race Is All About Electricity”

The turning point in the global AI race has officially shifted from chip supply timelines to power and grid infrastructure. As electricity demand from hyperscale data centers surges exponentially, power has not only become the core of hundreds of billions of dollars in capital expenditure for tech giants like Microsoft and Amazon, but is also emerging as a national security concern in the United States due to severe shortages of critical equipment such as transformers and rising geopolitical risks.

In the age of AI, computing power is equivalent to national power—and the prerequisite for computing power is a resilient and uninterrupted electricity grid. This AI-driven scramble for energy is rapidly reshaping the strategic importance of the global power supply chain.

Power Has Become the Top Capital Priority

For years, the biggest challenge in the tech industry was hardware and chip lead times. However, by 2026, the bottleneck has shifted from servers to substations.

With the rapid expansion of large-scale data centers, electricity has emerged as the primary growth constraint between 2023 and 2026. According to forecasts from the International Energy Agency (IEA) and Data Center Dynamics, global data center electricity consumption was approximately 460 terawatt-hours (TWh) in 2022, and is projected to exceed 1,050 TWh by 2026.

The growth of AI and high-performance workloads is far outpacing the ability of power systems to adapt. As a result, a new rule has taken hold in data center development:“No power, no deal.”

At the same time, a 2026 report by JLL highlights that AI training facilities require power densities up to ten times higher than traditional data centers. By 2030, AI workloads are expected to account for more than half of total global data center capacity—doubling compared to 2025.

Artificial intelligence has now become a core pillar of national strategy, prompting governments worldwide to invest heavily in sovereign infrastructure to strengthen domestic capabilities. This trend is expected to drive up to $8 billion in capital investment by 2030.

To feed these “power-hungry beasts,” U.S. hyperscale data center operators—including Microsoft, Alphabet, Amazon, Meta, and Oracle—plan to invest nearly $700 billion in AI-related capital expenditure between 2025 and 2026, with a substantial portion dedicated to ensuring stable infrastructure and reliable power supply.

Source of Information: Business Next