Bernstein Says Starship Reusability Central To SpaceX AI Valuation
Bernstein identifies full Starship reusability as critical to SpaceX's AI-driven valuation, forecasting $600 billion in 2031 revenue based on 3,500 launches. This falls short of Elon Musk's $1 trillion by 2030 target. The firm also notes upside from the $60 billion Anysphere acquisition and potential orbital AI deployments starting in 2028.

*this image is generated using AI for illustrative purposes only.
Bernstein states that full reusability of the Starship rocket is absolutely central to Space Exploration Technologies Corp. (NASDAQ: SPCX) long-term artificial intelligence valuation and current price levels. Analyst Douglas Harned told CNBC on Monday that solving this problem is necessary for orbital AI infrastructure at scale, noting that the path toward full reusability anchors the kinds of valuations the firm expects.
Revenue Forecasts And Launch Assumptions
Bernstein forecasts roughly $600 billion in SpaceX revenue by 2031. This projection assumes about 3,500 Starship launches in that year. The figure sits below Elon Musk’s earlier projection that SpaceX could generate $1 trillion in revenue by 2030, a target Reuters confirmed in June.
| Metric | Bernstein Forecast (2031) | Musk Projection (2030) |
|---|---|---|
| Revenue | $600 billion | $1 trillion |
| Launch Volume | 3,500 Starship launches | Not specified |
SpaceX filings warn that AI compute satellites at scale need full Starship reusability to become economically compelling. Slower turnaround times could raise costs and delay deployments. Reuters reported that SpaceX has spent more than $15 billion developing Starship, which remains central to orbital AI data centers. Musk has previously described achieving fully reusable rockets as a "crazy hard problem."
Orbital AI And Software Upside
SpaceX aims to begin orbital AI compute demonstrations by late 2027, ahead of possible deployments in 2028. Vice President Stephanie Bednarek recently said Starship could make this a reality by carrying the mass needed for large orbital platforms.
Harned noted that SpaceX’s software business could add upside not currently modeled by Bernstein. He pointed to Grok and the recent acquisition of Cursor parent Anysphere as potential wildcards. The company completed its $60 billion all-stock acquisition of Anysphere this month after announcing the deal in June. Reuters indicated the move could give xAI a stronger foothold in AI coding, while Cursor’s access to developer requests could improve models like Grok. Cursor serves more than 50,000 enterprises, including 64% of the Fortune 500.
What The Numbers Show
The divergence between Bernstein’s $600 billion revenue forecast for 2031 and Musk’s $1 trillion target for 2030 highlights the sensitivity of valuations to launch frequency and cost structures. With over $15 billion already spent on Starship development, the economic viability of orbital AI hinges on achieving the rapid turnaround rates required to support the assumed 3,500 launches annually. Failure to demonstrate full reusability would directly impact the cost basis of these deployments, potentially widening the gap between analyst models and management aspirations.
How might the technical challenges of achieving full Starship reusability impact the timeline for SpaceX's 2027 orbital AI compute demonstrations?
What specific synergies could arise between xAI's Grok models and Anysphere's Cursor platform to accelerate enterprise AI adoption?
If launch costs remain higher than projected due to slower turnaround times, how would that affect the competitive landscape for orbital data centers against terrestrial alternatives?

































