What Happened

The focus of artificial intelligence infrastructure investment is shifting from semiconductor manufacturing to electricity supply, as power generation becomes the primary constraint for data center expansion. Rand Corp. estimates that global AI data center power demand will reach 68 gigawatts by next year, a figure projected to surge to 327 gigawatts by 2030. This rapid growth is outpacing the ability to build generating capacity, creating a significant infrastructure gap. NextEra Energy, which operates Florida Power & Light, has announced a strategic deal with Alphabet's Google Cloud to build and power multiple new AI data centers. As part of this agreement, Google will assist NextEra in modernizing its digital systems to manage the increased load. NextEra Energy Resources, the company's wholesale power generation segment, plans to deliver 15 gigawatts of power to data centers by 2035. Of this target, 6 gigawatts will be supplied by new gas-fired power plants, reflecting a reliance on natural gas to bridge the immediate supply gap. The company's CEO, John Ketchum, stated he would be disappointed if the utility did not double its goals to deliver at least 30 gigawatts through this channel. Meanwhile, Credo Technology reported a 272% increase in second-quarter fiscal 2026 revenue, reaching $268 million, driven by demand for its Active Electrical Cables. These cables reduce signal degradation and power consumption in high-speed data connectivity. Credo issued guidance for its fiscal third quarter revenue between $335 million and $345 million. The article highlights that while America can construct data center buildings in months, securing the electricity to run them can take half a decade. This timeline mismatch is creating a new bottleneck for the AI industry. Analyst Patrick Sanders from The Motley Fool noted the critical nature of these infrastructure plays. The report identifies companies like NextEra Energy and Credo Technology as key beneficiaries of this power-centric shift in AI infrastructure.

Why It Matters

The transition from a chip-centric to a power-centric AI infrastructure race fundamentally alters investment logic and industrial planning. Investors can no longer assume that semiconductor availability is the sole constraint; the grid's capacity to deliver reliable, high-volume electricity is now the limiting factor. This shift means that utility companies and power generation firms are becoming as critical to AI growth as chipmakers like Nvidia and Broadcom. The reliance on gas-fired plants to meet immediate demand highlights the urgency of the supply gap and the environmental trade-offs involved in rapid AI expansion. For the broader economy, this indicates that energy policy and grid modernization are now directly tied to technological competitiveness. The half-decade timeline for power infrastructure versus months for building construction creates a structural lag that could slow AI deployment if not addressed. This bottleneck affects not just tech companies but also the energy sector, which must accelerate permitting and construction of new generation sources. The financial markets are already reacting, with semiconductor and energy stocks seeing significant movement as capital flows toward these new constraints. Understanding this power bottleneck is essential for predicting the pace of AI adoption and the profitability of infrastructure providers. It also suggests that future AI advancements may be capped by energy availability rather than computational power alone.

Local Vancouver / Burnaby Context

While this report focuses on global AI infrastructure and U.S. utility markets, the underlying demand for data center power has implications for regional energy grids and investment flows in British Columbia. Burnaby and the Greater Vancouver area are part of a larger North American energy market where utility companies like NextEra Energy operate. The shift toward gas-fired power generation to meet AI data center demands may influence regional energy pricing and policy discussions in BC, where the province is balancing its own clean energy targets with industrial growth. Local residents in Burnaby already contribute to regional infrastructure through property taxes and utility fees, which support the stability of the local grid. Any significant increase in regional power demand from data centers could impact local energy reliability and costs. Furthermore, the global trend toward power-constrained AI development may affect investment opportunities in BC's tech sector, particularly in companies involved in energy efficiency or data connectivity, such as Credo Technology. The Web Summit 2026 in Vancouver, which focuses on tech innovation, may also see increased discussion on sustainable AI infrastructure and energy solutions. Local policymakers and investors should monitor how global power bottlenecks influence local energy policy and tech investment strategies. The stability of the Greater Vancouver region's economy is linked to its ability to adapt to these global technological shifts, ensuring that energy infrastructure keeps pace with digital growth. This context highlights the interconnectedness of global AI trends and local energy realities in BC.