The tech industry would build AI data centers twice as fast if memory chip and energy supply constraints did not exist, according to Nokia CEO Justin Hotard.
Hotard said customers are willing to accelerate construction but cannot secure sufficient memory chips or power capacity. He made the comments on CNBC's The Tech Download podcast, emphasizing that even at the current pace, the AI infrastructure buildout remains in its infancy.
Nokia has positioned itself as a critical supplier in this market, providing the interconnect technology that links chip racks within data centers and connects facilities across regions. The company's stock has risen roughly 130 percent over the past year, reflecting investor confidence in its foundational role in the AI infrastructure wave.
Research presented at the Brookings Papers on Economic Activity estimates that AI infrastructure investments—data centers and chips combined—will total $10.3 trillion from 2025 to 2032, or an average of 3.63 percent of U.S. gross domestic product annually. Companies are deploying debt, joint ventures, and other financing structures to mobilize that capital.
Hotard pushed back against concerns that AI buildout may slow if frontier labs like Anthropic and OpenAI stop releasing breakthrough models. He argued that demand for AI infrastructure is not tethered to new model releases. Even without major advances for three years, he said, companies would deploy existing technology at scale, sustaining demand for data center capacity.
The current capabilities of these models and their early adoption stage ensure substantial workload growth, he added—meaning infrastructure constraints, not demand, are the limiting factor.
