
Narvik, Norway, is poised to become a pivotal European artificial intelligence hub following a substantial $6.2 billion investment from Microsoft, Nscale, and Aker for sustainable AI infrastructure. This landmark deal, set to commence services in stages from 2026, exemplifies a burgeoning global trend towards establishing specialized "Compute Towns" where advanced AI infrastructure drives local micro-economies. The initiative highlights the escalating demand for efficient, scalable, and renewable compute power, attracting engineers and related industries to regions with abundant energy resources.
The rapid scaling of AI infrastructure is fundamentally reshaping physical and economic landscapes. As Trace Vertical Ai Cohen stated in a recent social media post, "We’re not just building data centers we’re building new civilizations of compute." This surge in demand is causing global data center electricity consumption to more than double by 2030, according to the International Energy Agency (IEA), with AI being the primary driver. Goldman Sachs projects a 160% surge in data center power demand by 2030, emphasizing the urgent need for reliable and carbon-free power sources.
Small Modular Reactors (SMRs) are emerging as a key component in powering these new AI-centric developments. Companies like Oklo Inc. and TerraPower are actively developing microreactors capable of providing decades of continuous, carbon-free power, a solution ideal for energy-intensive data centers. The U.S. Department of Energy is supporting advanced nuclear reactor programs, including SMRs, recognizing their potential to meet the urgent energy demands of AI. This shift points to a future where data centers are built around dedicated, resilient power generation.
The rise of "Compute Towns" also brings concerns about economic stratification, where access to advanced computing power could become a new form of inequality. "Inequality in the AI era won’t just be about wealth or education. It’ll be about proximity to compute," Cohen observed. Policymakers and researchers are advocating for "public compute utilities" or national AI clouds, drawing parallels to historical initiatives like rural electrification, to ensure broader and more equitable access to essential AI infrastructure for smaller businesses and academic institutions.
Furthermore, the concentration of AI compute capacity is leading to significant political reconfiguration, with AI becoming a critical national asset. Governments and tech giants are forming new alliances to secure land, energy, and infrastructure, transforming AI capacity into a matter of national sovereignty and strategic competitiveness. The Council on Foreign Relations and the Atlantic Council have both highlighted this escalating geopolitical competition, where control over AI compute resources is reshaping international relations and driving new industrial policies. The next "Silicon Valleys," as Cohen noted, "will be built where the power grid meets the cloud."