SpaceX Targets 10GW of AI Compute by 2027
SpaceX aims to expand its AI compute capacity from about 1.4GW in Q2 2026 to roughly 10GW by the end of 2027, with a near-term target of more than 2GW by the end of 2026. The SpaceX AI compute expansion is set to rely primarily on Nvidia’s Vera Rubin chips, including NVL72 and Kyber systems.
The company is also developing Starmind, an orbital AI platform, with initial satellite launches planned for 2027 and a larger rollout expected in 2028. SpaceX plans to reserve about 60% of its compute for internal operations, including xAI and Tesla, and offer the remaining capacity to outside customers such as Google and Anthropic.
Elon Musk has forecast annual recurring revenue of $300 billion to $500 billion at the proposed capacity and pricing. However, the plan requires substantial investment: Q2 2026 capital expenditure was estimated at $15.8 billion to $18.4 billion. Reliance on a single chip supplier and supply constraints affecting components and cooling systems could put the timeline at risk.
Neutral
The announcement has no clear direct effect on cryptocurrency prices: it concerns AI infrastructure investment, not crypto assets, blockchain adoption or token demand. It may contribute to broader risk appetite around technology and AI, but any spillover into crypto is likely to be indirect and sentiment-driven.
In the short term, traders may focus on Nvidia’s chip demand, the scale of SpaceX’s planned spending and whether the company meets its 2026 capacity milestone. Similar announcements of large AI infrastructure commitments have often supported technology-sector sentiment, but they do not reliably produce sustained moves in Bitcoin or other crypto assets. In the longer term, the project could influence energy demand, computing costs and competition for data-centre resources—factors that may matter to crypto mining and infrastructure businesses. The scale of capital expenditure, dependence on Nvidia and supply-chain risks also leave room for delays or a reassessment of projected returns. Overall, the article does not provide a strong enough catalyst to classify the crypto-market impact as bullish or bearish.