Big Tech’s AI Splurge Faces Physical Reality Check
The tech giants’ ambitious plans to invest heavily in AI infrastructure are hitting a brick wall – physical reality. Four major companies, Amazon, Microsoft, Alphabet, and Meta Platforms, are projected to spend around $630 billion on data centers and AI chips in 2026 alone, but experts warn that this splurge may fall short due to physical constraints.
The rapid innovation in AI technology is creating a backlog, with Nvidia’s newest Blackwell chips generating more heat than previous versions, forcing data centers to shift from air cooling to more complex liquid systems. This shift requires new plumbing and water purification infrastructure, adding to the challenge.
Moreover, the broader industrial supply chain is struggling to keep up with overwhelming demand. The lead time for transformers supplied by groups such as Schneider Electric, Eaton, and Hitachi Energy is now up to 100 weeks in Europe, while generators in the United States can take around 50 weeks to arrive.
As a result, tech companies are exploring alternative solutions, such as renting capacity from agile “neocloud” operators like CoreWeave and Nebius, and building “island” data centers powered by on-site gas turbines. However, these workarounds have their own limitations, and the industry’s worst bottlenecks are not necessarily in semiconductors, but in physical infrastructure and local permits.
The physical constraints and rising construction costs are expected to compress profit margins and drag down tech giants’ returns on their investment. For instance, Alphabet’s return on invested capital after tax is expected to fall from 51% last year to about 36% by 2030.
Author: Korkor Anumu
