Which companies are securing power for AI right now?

Microsoft is helping restart a nuclear reactor at Three Mile Island. Google has signed agreements tied to next-generation nuclear power. Amazon, Meta, Oracle, OpenAI, and others are pursuing long-term power arrangements and dedicated generation assets.

How much power capacity will AI data centers need?

JLL projects developers may require roughly 100 gigawatts of new data-center capacity over the next several years through 2030.

Why is Bitzero positioned differently from other companies?

Bitzero secured low-cost electricity, land, permits, and grid access years ago—before AI factories became a major investment theme—and now controls a development pipeline exceeding one gigawatt of potential capacity. Most other companies are still waiting in interconnection queues.

Back to articles

AI's Real Bottleneck: Power, Not Chips

Yahoo Finance AI8h ago4 min read

Key takeaway

As artificial intelligence companies race to build massive data centers, electricity has become the industry's scarcest and most limiting resource—more critical than chips or software. Major tech firms including Microsoft, Google, Amazon, and Meta are now competing aggressively for long-term power supplies and dedicated energy assets, while utilities face multi-year backlogs for infrastructure upgrades. This fundamental shift means that energy security, not computing power alone, will determine which companies can scale their AI operations.

Summaries like this, in your inbox every morning.

3 Key Points

What happened
As AI companies race to build massive data centers, electricity has emerged as the scarcest resource. McKinsey estimates AI-related infrastructure spending could exceed $5 trillion（約800兆円） by 2030, while utilities across North America and Europe are receiving requests for hundreds of megawatts from individual AI campuses. Microsoft, Google, Amazon, Meta, Oracle, and OpenAI are now pursuing long-term power arrangements and dedicated generation assets to secure electricity. Bitzero (NASDAQ: AIBZ), which controls a development pipeline exceeding one gigawatt of potential capacity, positioned itself ahead of this scramble by securing low-cost electricity and grid access years ago.
Why it matters
For the past three years, investor focus has centered on chips, models, and software. But electrical demand now poses a hard physical constraint on AI expansion. Many utilities face multi-year waits for interconnection studies and transmission upgrades before data centers can even operate. For businesses dependent on AI infrastructure—or energy providers—this shift from chip scarcity to power scarcity represents a fundamental change in how AI expansion will be funded and constrained.
What to watch
JLL projects developers may require roughly 100 gigawatts of new data-center capacity over the next several years through 2030. Companies that secured power agreements early, like Bitzero, now hold a strategic advantage over those still waiting in interconnection queues.

FAQ

Which companies are securing power for AI right now?: Microsoft is helping restart a nuclear reactor at Three Mile Island. Google has signed agreements tied to next-generation nuclear power. Amazon, Meta, Oracle, OpenAI, and others are pursuing long-term power arrangements and dedicated generation assets.
How much power capacity will AI data centers need?: JLL projects developers may require roughly 100 gigawatts of new data-center capacity over the next several years through 2030.
Why is Bitzero positioned differently from other companies?: Bitzero secured low-cost electricity, land, permits, and grid access years ago—before AI factories became a major investment theme—and now controls a development pipeline exceeding one gigawatt of potential capacity. Most other companies are still waiting in interconnection queues.

Discussion

No discussion yet for this article

Stay ahead with AI news

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.

Get Started Free

Free · takes 30 seconds · unsubscribe anytime

1 minute a day. The AI essentials.

200+ sources · Email / LINE / Slack

Get it free →