From Rockets to Rack Units: SpaceX's Computing Power Gambit
While the world watches Starship launches, SpaceX is quietly orchestrating a move into a different frontier: artificial intelligence infrastructure. A new industry analysis suggests the company is targeting a staggering deployment of over 10 gigawatts (GW) of dedicated computing capacity by the end of 2027.
The Roadmap to 10GW
The report assesses this goal as achievable. Elon Musk has indicated that SpaceX's baseline target is to deliver 6 to 8 GW of new computing power by 2027, with a clear path to exceed 10 GW.
The Capital Behind the Compute
Building at this scale requires monumental investment. Using an industry benchmark of approximately $50 billion in capital expenditure per GW deployed, the report estimates that SpaceX's related spending in 2027 could range from $300 billion to a staggering $500 billion.
The financial rationale hinges on the booming demand for AI inference. The analysis models that when leading AI firms like OpenAI or Anthropic provide API services on high-end clusters (such as GB300-class systems), each GW of capacity could generate over $100 billion in annual revenue.
The Economics of AI Infrastructure
On the cost side, operating 1 GW of capacity, priced at an estimated $3 per GPU-hour, would incur roughly $12 billion in annual expenses. This gap highlights the substantial margin potential in premium compute services.
This isn't just theoretical. The report points to Microsoft's $250 billion infrastructure deal with OpenAI in October 2025, which analysts estimate covers about 7 GW of capacity. By similar logic, a future contract between Microsoft and SpaceX for around 3 GW, potentially worth $150 billion, is seen as a plausible scenario.
The $300 Billion Horizon
Synthesizing these figures, the core projection emerges: by the end of 2027, SpaceX's annual recurring revenue from compute services could reach $300 billion. This would not only transform SpaceX's financial profile but could position it as a dominant force in global computing supply.
Behind these numbers lies the insatiable demand for compute driven by generative AI. If SpaceX can successfully translate its expertise in engineering, energy, and scale from aerospace to terrestrial data centers, it will secure a commanding position in the next phase of the technology race. This cross-industry venture is quietly redrawing the map of technological power.