SpaceX CEO Elon Musk is back at it with his bold prediction. This time its about AI data centres. On a podcast, Musk says that space will become the most cost-effective location for AI data centres within
two to three years. This comes just as India sets up its aggressive terrestrial strategy. The government hs even rolled out sweeping tax incentives to attract global cloud giants to build facilities on Indian soil.
Speaking on the 'Cheeky Pint' podcast with Stripe co-founder John Collison and AI podcaster Dwarkesh Patel, Musk issued a stark warning, "Mark my words" – space-based computing will dominate by 2027.
'Magical Electricity Fairies' Won't Save Ground Data Centres
Musk painted a dire picture of the energy crisis facing AI expansion on Earth. "The availability of energy is the issue," he stated bluntly. "The output of chips is growing pretty much exponentially, but the output of electricity is flat. So how are you going to turn the chips on? Magical electricity fairies?"
The mathematics are sobering. The US currently consumes about half a terawatt of electricity on average. Deploying an additional terawatt for AI infrastructure would require doubling the country's entire electricity generation capacity. "Those who have lived in software land don't realize they're about to have a hard lesson in hardware," Musk warned. "It's actually very difficult to build power plants."
The bottleneck extends beyond just building plants. "The limiting factor is the vanes and blades," Musk explained, referring to turbine components. "There are only three casting companies in the world that make these, and they're massively backlogged."
His prediction, "Towards the end of this year, chip production will probably outpace the ability to turn chips on."
Five Times More Power, Zero Batteries Required
Musk's case for orbital data centres rests on fundamental physics advantages. "You're also going to get about five times the effectiveness of solar panels in space versus the ground, and you don't need batteries," he explained. "The atmosphere alone results in about a 30 percent loss of energy."
The vision extends beyond incremental improvements. "The only place you can really scale is space," Musk insisted. "Once you start thinking in terms of what percentage of the Sun's power you are harnessing, you realize you have to go to space."
His headline prediction was unequivocal: "My prediction is that it will be by far the cheapest place to put AI. It will be space in 36 months or less. Maybe 30 months… mark my words."
TeraFab: The 100 Gigawatt Goal
Musk revealed plans for what he calls "TeraFab" – Tera being the new Giga. "You need 100 gigawatts worth of chips," he said, outlining an ambitious 2030 target. "We want 100 gigawatts of power and chips that can take 100 gigawatts by 2030… We'll take as many chips as our suppliers will give us."
The scale is staggering. "You've got to match mass to orbit, power generation, and the chips," Musk explained. Meeting this goal would require matching orbital infrastructure deployment with unprecedented manufacturing capability.
One Starship Launch Every Hour
The logistics of Musk's vision are equally audacious. "100 gigawatts… is on the order of 10,000 Starship launches," he calculated. "You want to do that in one year. That's like one Starship launch every hour."
Far from seeing this as impossible, Musk views it as achievable with relatively modest fleet sizes. "If you can use a ship every, say, 30 hours, you could do it with 30 ships," he said. "SpaceX is gearing up to do 10,000 launches a year, and maybe even 20 or 30,000."
The ambition doesn't stop there. "SpaceX is going to become a hyperscaler. Hyper-hyper," Musk declared. "SpaceX will launch more AI than the cumulative amount on Earth of everything else combined."
Looking five years ahead, Musk went further, "Five years from now, my prediction is we will launch and be operating every year more AI in space than the cumulative total on Earth."
India's Competing Vision
Musk's orbital predictions arrive as India pursues an opposite strategy, doubling down on terrestrial data centre infrastructure. Finance Minister Nirmala Sitharaman announced in the Union Budget 2026 a 21-year tax holiday extending through 2047 for foreign companies that provide cloud services globally using Indian data centre infrastructure.
The incentive package addresses longstanding concerns among multinational cloud providers including Google, Microsoft, and Amazon Web Services about potential tax exposure. IT Minister Ashwini Vaishnaw projects the tax holiday could attract investments worth $200 billion.
India's push addresses a critical vulnerability: the country generates approximately 20 percent of global data but currently stores and processes 95 percent of it abroad. The country's data centre capacity reached 1.5-1.7 gigawatts by the end of 2025 and is projected to expand five-fold to 8 gigawatts by 2030.
Google's October announcement of a $15 billion investment in a 1-gigawatt AI data centre in Visakhapatnam marked a watershed moment for India's digital infrastructure ambitions.
