Musk vows to put data centers in space and run them on solar power but experts have their doubts

Elon Musk vowed this week to upend another industry just as he did with cars and rockets — and once again he’s taking on long odds.

The world’s richest man said he wants to put as many as a million satellites into orbit to form vast, solar-powered data centers in space — a move to allow expanded use of artificial intelligence and chatbots without triggering blackouts and sending utility bills soaring.

To finance this ambitious effort, Musk combined SpaceX with his AI business on Monday and plans a significant initial public offering of the merged company, potentially creating one of the largest technology IPOs in recent years.

“Space-based AI is obviously the only way to scale,” Musk wrote on SpaceX’s website Monday, adding about his solar ambitions, “It’s always sunny in space!”

The move comes amid growing concerns about AI’s massive energy consumption. Data centers already account for approximately 1% of global electricity use, with that figure expected to rise significantly as AI applications expand. Terrestrial facilities require enormous resources for cooling and operation, creating environmental and infrastructure challenges.

However, scientists and industry experts say even Musk — who outsmarted Detroit to turn Tesla into the world’s most valuable automaker — faces formidable technical, financial and environmental obstacles.

One major challenge is heat management. While space seems ideal with its cold temperatures, it’s also a vacuum that traps heat inside objects, similar to how a thermos keeps coffee hot.

“An uncooled computer chip in space would overheat and melt much faster than one on Earth,” explained Josep Jornet, a computer and electrical engineering professor at Northeastern University.

Engineers would need to build giant radiator panels that emit infrared light to push heat “out into the dark void,” Jornet noted. While this technology has worked on a small scale on the International Space Station, Musk’s data centers would require “massive, fragile structures that have never been built before.”

Musk remains confident despite these challenges. “You can mark my words,” he said in a preview of a Cheeky Pint podcast episode. “In 36 months, but probably closer to 30 months, the most economically compelling place to put AI will be space. And then it will get ridiculously better to be in space.”

Space debris presents another significant hurdle. With plans to launch approximately 100 times more satellites than currently in his Starlink network, the risk of collisions increases dramatically. A single malfunctioning satellite could trigger a cascade of crashes, potentially disrupting emergency communications, weather forecasting and other critical services.

“We could reach a tipping point where the chance of collision is going to be too great,” warned John Crassidis, a University at Buffalo professor and former NASA engineer. “And these objects are going fast — 17,500 miles per hour. There could be very violent collisions.”

Maintenance poses yet another challenge. On Earth, technicians can easily replace damaged components in data centers, but no such repair crews exist in orbit. The specialized graphics processing units (GPUs) crucial for AI applications are particularly vulnerable to damage from high-energy solar particles.

Baiju Bhatt, CEO of space-based solar energy company Aetherflux, suggests one workaround is to equip satellites with redundant chips to replace ones that fail. However, this approach substantially increases costs, with each chip potentially costing tens of thousands of dollars. Current Starlink satellites have a relatively short lifespan of about five years, adding to the economic challenges.

Musk isn’t alone in this pursuit. Starcloud, a Redmond, Washington-based company, launched a satellite in November carrying a single Nvidia-made AI chip as a test. Google is exploring orbital data centers through its Project Suncatcher, while Jeff Bezos’ Blue Origin announced plans for more than 5,000 satellites starting next year, though its focus has been more on communications than AI.

What gives Musk a significant competitive advantage, however, is his ownership of rockets. Starcloud used one of Musk’s Falcon rockets for its test launch, and Aetherflux plans to send its “Galactic Brain” chips to space on a SpaceX rocket later this year. Even Google may need to rely on Musk to launch its first two prototype satellites by early next year.

Pierre Lionnet, a research director at trade association Eurospace, points out that Musk typically charges competitors far more for launches than his internal cost — as much as $20,000 per kilogram versus $2,000 for SpaceX projects.

“When he says we are going to put these data centers in space, it’s a way of telling the others we will keep these low launch costs for myself,” said Lionnet. “It’s a kind of powerplay.”

With the AI market expected to reach $1.3 trillion by 2032 according to Bloomberg Intelligence, the race to establish space-based computing could reshape the technology landscape for decades to come.