Google dumps another $40B into new AI data centers

Google will spend $40 billion to build three new data centers in Texas by 2027. This expands its footprint to support rising AI demand. The plan adds major capacity in Armstrong and Haskell counties. It also includes upgrades in Midlothian and Dallas. The buildout shows how fast cloud infrastructure is scaling to handle heavier model loads.

Is this investment actually worth it?

Google is making a large Texas investment that widens its U.S. network. It also deepens its push into cloud computing. The three new sites join its 42 global cloud regions. The company expects thousands of jobs. It also plans new training programs in electrical trades and digital skills.

Google is boosting investment at its Midlothian campus and in the Dallas cloud region. This move reinforces Texas as a key hub for high-density computing.

Analysts say hyperscale operators are racing to add capacity. They need more space to handle bigger models and rising demand for faster computing. This expansion supports Google’s need for more power, cooling, and advanced networking, if the company wants to stay competitive in the cloud market.

More capacity = larger and better LLMs

This construction wave shows how data centers have become core infrastructure for the next decade of computing. We see it as another milestone in the race to deliver fast, low-latency AI services at global scale. More capacity allows larger models. It also allows faster inference and more reliable training cycles.

AI is moving into everyday tools, from search to productivity to industrial automation. Because of this shift, the physical backbone behind these systems matters more. Demand for power, land, and cooling technology continues to shape regional economies and utility planning.

For us, the size of Google’s plan highlights a clear trend. AI workloads are becoming an essential service, not a niche tool. This shift may change how people use software and how companies run core operations. It also raises the bar for every cloud provider competing on performance and availability.

How to read this cloud-building wave

We see clear winners and new risks in U.S. public markets. Alphabet (GOOGL) is the most direct beneficiary. It is extending its long-term moat by building the infrastructure needed for dependable AI throughput. More capacity improves product stickiness. It also strengthens Google Cloud against Amazon’s AWS (AMZN) and Microsoft Azure (MSFT).

Utility-linked firms that supply high-voltage equipment, grid upgrades, and power management may gain if demand keeps rising. Semiconductor and accelerator suppliers, especially Nvidia (NVDA) will benefit from steady orders for the systems used to train AI models.

Data-center REITs like Digital Realty (DLR) and Equinix (EQIX) face mixed effects. Demand remains strong. Yet hyperscalers are building more of their own campuses, which could pressure long-term pricing.

We also see risks. If enterprise adoption slows, the sector could feel pressure. This is more likely if companies increase spending faster than they grow revenue. Even so, our base case still calls for steady growth as AI becomes a standard part of enterprise software. As this buildout continues, we expect AI spending to stay strong and support cloud providers and component suppliers for several years.

Google (GOOG) has a Disruption Score of 5 and is part of the Disruption Aristocrats.