ChipsThe story, in brief

Inside the race to power AI data centers with fusion energy — and the surprise detours along the way

Not a pilot. Tech giants are betting fusion energy will power the next generation of AI data centers — and startups like Helion and Zap have a 2028 deadline to deliver.

Paper-cut illustration of an amber microchip with circuit paths extending into a row of data-center cabinets.
The infrastructure powering AI.AI illustration by KeyNews
The KeyNews take

Why it matters

AI data center power consumption is becoming a critical infrastructure bottleneck. The race to secure reliable, scalable energy sources — from fusion to traditional nuclear — is now a direct competitive advantage for companies building at hyperscale.

The key facts

5 to know
  1. Helion and Zap Energy (Seattle-area fusion startups) actively pursuing commercial fusion for AI infrastructure

  2. 2028 deadline for fusion energy commercialization cited as critical milestone

  3. Tech giants exploring fusion as solution to AI data center power demands

  4. Strategy includes mix of fusion and traditional nuclear power expansion

  5. Down-scaling of fusion approaches mentioned as key pivot in commercialization path

Go to the source

GeekWiregeekwire.com

Publisher excerpt: While some tech giants are looking to the stars for nearly limitless clean electricity, the actual path to commercial fusion energy involves an unexpected down-scaling, an expansion to traditional nuclear power, and a tight 2028 deadline. Go inside Seattle-area fusion startups Helion and Zap as…
Read original report
Back to today's editionMore chips news

Keep reading

Related stories

More from Chips