Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
At the heart of Elon Musk’s ambitions to send humans to Mars — and propel SpaceX to the largest initial public offering in history — is a gargantuan 407-foot-tall rocket known as Starship.
The AI-to-satellite conglomerate led by the world’s richest person has ploughed more than $15bn into developing the largest rocket ever built. In regulatory filings for its IPO this month, SpaceX described the vehicle as central to its plans for future growth.
SpaceX hopes Starship will become the world’s first fully reusable launch system. Recycling both stages of the rocket, while dramatically increasing payload capacity, would allow the company to launch more equipment into orbit, carry larger satellites and eventually undertake missions to the Moon and Mars at far lower cost.
The promise is central to some of the company’s most ambitious projections. In September 2024, Musk posted on X: “Being multiplanetary will vastly increase the probable lifespan of consciousness, as we will no longer have all our eggs, literally and metabolically, on one planet.”
Yet Starship remains far from operational. US regulators grounded the rocket after the company lost control of its booster during a test flight in May. Turning it into a reliable transport system will require technical breakthroughs, including the ability to refuel the spacecraft while in orbit.
Powered by 33 of the company’s raptor booster engines, Starship dwarfs even the Saturn V rocket that Nasa used in the Apollo Moon missions.
“Starship is fundamental to every future endeavour that SpaceX wants to pull off,” said Caleb Henry, an analyst at Quilty Space.
SpaceX argues its existing Falcon rockets cannot support many of its future plans. The company has said neither its Falcon 9 nor Falcon Heavy can deploy its next generation of Starlink satellites at the scale required, nor support proposals for space-based data centres.
The company has had some progress. In October 2024, it successfully caught a returning booster using giant “mechazilla” mechanical arms at its Starbase facility in Texas, a milestone in its effort to create a rapidly reusable launch system.
Over the past decade, SpaceX has turned reusable boosters from a controversial idea into the foundation of the commercial launch market. Falcon 9 has flown hundreds of missions, helped drive down launch costs and established the company as the dominant provider of commercial access to space.
“SpaceX has had no reason to lower its prices,” said Greg Autry, associate provost at the University of Central Florida. “Market demand currently exceeds capacity.”
That position has helped SpaceX build Starlink, the satellite broadband network that has become one of the company’s most important businesses. The service now reaches more than 10mn customers worldwide and generated $4.4bn in operating income last year.
SpaceX says Starship is needed to continue expanding the network, allowing it to launch larger and more capable satellites than its current rockets can carry.
Before that can happen, however, the company must overcome engineering hurdles.
SpaceX has launched Starship 12 times, using the flights to test major components and systems. Several missions have ended in failure, including incidents in which the company lost control of the vehicle and debris fell over parts of the Caribbean. One rocket exploded on the launch pad during a test firing.
The setbacks have highlighted the challenge of building a fully reusable heavy-lift rocket. SpaceX is not alone. Rivals including Jeff Bezos’s Blue Origin and United Launch Alliance have also encountered difficulties developing next-generation launch systems.
Even if Starship succeeds technically, the scale of SpaceX’s ambitions remains extraordinary.
The company says Starship will eventually be able to carry up to 100 metric tonnes of cargo into orbit, with later versions capable of carrying more. That capacity would enable the company to deploy larger satellites, transport substantial amounts of equipment beyond Earth and support projects that are impractical because of launch costs.
Musk has spoken for years about building self-sustaining cities on Mars. More recently, SpaceX has told investors that Starship could also support vast orbital data centres designed to run AI systems.
The vision reflects one of the most striking assumptions in the company’s long-term growth plans: that AI computing infrastructure could eventually move into space, where access to solar power is unlimited.
“My prediction is that by far the cheapest place to put AI will be space in 36 months,” Musk said in February, driven by the rising costs of energy and associated data centre infrastructure on Earth.
Analysts at SemiAnalysis have argued that orbital data centres remain significantly more expensive than terrestrial alternatives and would require major advances in engineering and materials science before becoming commercially viable.
The challenge is not simply building Starship. It is operating the system at a frequency no launch provider has achieved. SpaceX has transported thousands of tonnes of payload into orbit across its history. Realising its most ambitious plans would require launching vastly larger amounts of material, potentially through several thousand Starship flights each year.
The company’s willingness to embrace risk and rapid testing has helped it move faster than competitors. But the future presented to investors — of humanity expanding its horizons across the solar system — depends on proving that Starship can become a reliable and routine means of reaching space.
“Is it feasible to launch 5,000 flights a year? Nobody knows yet,” Henry at Quilty Space added. “SpaceX has taken industry precedent and turned its head multiple times before.”
Additional reporting by Tim Bradshaw in London
Additional graphics by Cleve Jones
