SpaceX is planning 10 gigawatts of data center capacity by 2027, according to SemiAnalysis, with Microsoft positioned as the largest offtaker in what could become a $500 billion annual revenue line for Elon Musk's space company.
The Scale of 10 Gigawatts
Ten gigawatts is roughly the power consumption of Portugal. SpaceX intends to build this capacity using the rapid-construction methodology proven at xAI's Colossus facilities, where prefabricated compute shells can be assembled in months rather than years. SemiAnalysis reports that Colossus 1 and 2 were built remarkably fast through retrofits, and SpaceX's first true greenfield site suggests the company can move even faster on subsequent builds.
The Microsoft connection is critical. As the largest offtaker, Microsoft would gain access to compute capacity outside its existing Azure data center footprint: a potentially cheaper alternative to building its own facilities from scratch. For SpaceX, Microsoft's offtake commitment provides the revenue guarantee needed to finance the build-out.
Why SpaceX, Not Just Big Tech
The AI data center buildout has been framed as a competition between Google, Microsoft, Amazon, and Meta. SpaceX changes the calculus. Musk's company brings three structural advantages: vertically integrated power procurement from Starlink's satellite and ground station infrastructure, rapid physical construction at scale from a company that builds and launches rockets weekly, and a willingness to operate in locations that Big Tech has overlooked.
SemiAnalysis notes that SpaceX's construction speed could deliver compute capacity 12 to 18 months faster than traditional hyperscaler timelines. In an industry where training runs cost hundreds of millions of dollars and time-to-compute determines competitive position, that speed premium is itself a moat.
The $500 Billion Revenue Question
The $500 billion ARR estimate assumes SpaceX fills the 10GW capacity with enterprise AI workloads at current market rates. This figure would make SpaceX's data center business alone worth more than all but a handful of companies in the S&P 500. For context, Microsoft Azure generated approximately $110 billion in annualized revenue in 2026.
The assumption deserves scrutiny. Current AI compute demand is growing rapidly, but the $500 billion figure requires sustained demand growth at rates that imply AI infrastructure spending will consume a meaningful share of global GDP. If demand plateaus or competitors build out faster, SpaceX's capacity could sit partially idle.
Energy as the New Bottleneck
The SpaceX buildout underscores a structural reality: AI scaling is no longer limited by algorithms or chips. It is limited by power. The companies that control access to gigawatt-scale electricity, cooling infrastructure, and construction speed will determine how fast AI scales over the next decade.
For Bitcoin miners, the implications are mixed. More data center capacity means more competition for cheap power in the very locations where mining operations have clustered. But it also validates the thesis that energy infrastructure is the macro trade of this decade.
Follow @UTXOMacro for daily Bitcoin, macro, and AI breakdowns.