SpaceX Plans First Orbital AI Computing Tests by Late 2027, Ahead of IPO Schedule

SpaceX has told investors it plans to launch initial demonstrations of space-based artificial intelligence computing infrastructure by late 2027, ahead of the "as early as 2028" timeline disclosed in the company's IPO filing, according to two people who attended investor presentations held prior to the offering.

The orbital compute effort sits at the centre of SpaceX's long-term growth pitch. In its IPO documents, the company claims to be "the only company with a commercially viable path to building orbital AI compute at scale." SpaceX has also requested regulatory permission to launch up to one million space-based data centre satellites.

During investor presentations featuring President Gwynne Shotwell and Chief Financial Officer Bret Johnsen, executives described the 2027 deployments as demonstrator systems, intended to validate the technology before any broader commercial rollout. The distinction is meaningful: the IPO filing did not separate demonstration missions from commercial deployments, giving management room to manage expectations on Starship's development timeline.

SpaceX stock is scheduled to begin trading on the Nasdaq on Friday under the ticker symbol SPCX, with the IPO price targeted at $135 per share and a total fundraise target of $75 billion.

The orbital AI compute strategy has direct relevance for enterprise technology buyers in the Gulf region. Saudi Arabia and the UAE have both made significant commitments to space and AI, and any viable commercial orbital compute infrastructure would have implications for data sovereignty, latency-sensitive AI workloads, and connectivity in underserved markets.

Starship, the fully reusable rocket that underpins the orbital computing plan, remains years behind Elon Musk's original targets and has yet to demonstrate the rapid reusability required to make large-scale deployment economically viable. In a video released on Monday, Musk said that building orbital AI data centres is not a difficult engineering challenge, noting that much of the required technology already exists within the Starlink satellite network. The first version of the AI satellite is expected to use Nvidia chips, with computing power equivalent to a single Nvidia GB300 rack.

Separately, OpenAI is in talks to lease a 10-gigawatt data centre facility in Ohio, also involving Nvidia backing, underscoring how concentrated the infrastructure layer of AI is becoming around a small number of key hardware and financing players.