Breaking

WSJ: Power Crunch Puts Amazon and Google Ahead in AI Race

June 24, 2026 at 06:32 EDT

Infra & Chips
Sales & Business

As the biggest bottleneck in building and running AI shifts to power supply, Amazon (AWS) leads through an incumbent advantage in massive data-center infrastructure, while Google stands out with differentiation in its own hardware—so argues a market analysis by the Wall Street Journal.

The AI Power Race · Amazon vs Google

Power, Not Chips, Is the New AI Battleground

As AI loads strain the US grid, Amazon leads on existing scale and utility ties, while Google bets on a distinctive move — buying its way into owning power generation itself.

~9 GW

Amazon US data center power use — comparable to all of North Dakota's generating capacity

$100B+

Amazon data center investment planned over the next decade

$4.75B

Google's planned Intersect Power buy — could make it the only Big Tech to own a generator

The grid squeeze ahead

Data centers' share of US electricity by 2030

Low
forecast

17%

High
forecast

Gigawatt-scale AI loads are landing on a grid where demand was long flat — turning power procurement into the central factor in hyperscaler investment.

5–15% premium — what hyperscalers reportedly pay for flexible power purchase agreements.

Amazon — Scale & Ties

20+ years building its own data centers
Long-standing relationships with utilities
~$40B pledged in Spain through 2035
Edge: existing scale and stability

Google — Vertical Integration

$4.75B move into wind & solar generation
In-house TPU chips and proprietary data
Breadth of research, funded by non-AI revenue
Edge: owning the power itself

The bottom line

"Power is the single biggest constraint" in the cloud and AI business — making both companies' early energy bets practical preparation, not just spending.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.