Musk's orbital data centers: a repeat of microsoft's underwater dream?
Elon Musk, never one to shy away from ambitious projects, is aiming for the stars—literally. His latest venture involves constructing data centers in orbit, a move that echoes Microsoft's ill-fated Project Natick, and raises serious questions about the viability of space-based infrastructure in the age of relentless AI growth.

The ai boom fuels a data center frenzy
The explosive growth of artificial intelligence is driving an unprecedented surge in demand for data centers worldwide. These colossal facilities, essential for training and deploying AI models, are consuming vast amounts of energy and water, sparking fierce opposition from local communities. Just last month, protests erupted in Zaragoza, Spain, as Microsoft and Amazon sought to build new data centers in the region – a sign of the growing tensions between technological advancement and environmental concerns.
Musk's vision, however, proposes a radical solution: taking these power-hungry behemoths off Earth. But is it a stroke of genius, or a costly, impractical fantasy? The parallels with Microsoft's Project Natick, launched in 2015, are striking. That initiative envisioned submerged data centers cooled by seawater, touting sustainability and efficiency. A prototype, roughly the size of a shipping container, was deployed off the coast of Scotland.
The initial promise of Project Natick—a self-sufficient, eco-friendly data center—ultimately floundered. The costs associated with maintaining and repairing the underwater facility, even with renewable energy sources like wind power, proved prohibitive. Roy Chua, founder of AvidThink, warns that the challenges facing Musk's orbital data centers could be even more acute. “It’s probable that these problems are more serious in space than underwater,” he told Reuters.
The core concept behind Musk’s plan—modular units deployed at a high cost with minimal operational expenses—mirrors Project Natick. But a growing chorus of experts are raising concerns about the 'hidden' costs of space-based data centers. The inability to easily update or repair these systems poses a significant hurdle to attracting clients willing to make multi-million dollar investments with no guarantees.
While Tesla’s sales recently saw a boost, its stock dipped, failing to meet expectations – a stark reminder of the risks involved in pursuing futuristic technologies. The fundamental flaw? Data centers in space risk becoming obsolete the instant they’re launched. As Roy Chua bluntly puts it, these satellites will be “locked down for life,” a prospect that’s hard to sell to businesses needing to constantly adapt to the rapid evolution of AI chip technology. Terrestrial facilities offer the agility of faster deployment and hardware upgrades—a critical factor in today's fast-paced landscape.
Musk’s orbital data center, like Microsoft’s underwater experiment before it, may ultimately prove to be an expensive lesson in the realities of scaling technology beyond Earth. The International Space Station, a marvel of engineering, is dwarfed by the sheer scale of ambition Musk is pursuing. The question isn't whether Musk is aiming high, but whether he's overlooking the harsh economic realities that sunk a similar project years ago.
