SpaceX is days away from one of the most anticipated initial public offerings in stock market history. The company is expected to price on June 11 and begin trading on June 12 at a valuation of approximately $1.75 trillion, which would make it the most valuable IPO ever completed.
Before the roadshow closed, Elon Musk chose to make a specific argument about what kind of company investors are actually buying.
It was not primarily about rockets or satellites. Musk used a 30-minute technical presentation to argue that SpaceX is building the infrastructure to solve the problem that every major AI company on Earth is currently struggling with: where to get the power to run the next generation of AI systems. The answer he gave was orbit.
What Musk said about SpaceX’s orbital AI data center plans
Speaking in a video discussion released by SpaceX on June 8, Musk said building orbital AI data centers is not a difficult engineering challenge.
“Part of what we want to convey here is that there is not some magic that is necessary, that doesn’t exist,” he said. “A lot of this is technology we’ve already made for the Starlink V3 satellites. We don’t think this is a super hard problem compared to the things we already do,” he added, according to Investing.com.
More Elon Musk:
- Elon Musk’s Terafab bet: what it means for Tesla investors
- Bank of America revamps Tesla stock price
- UBS has a message for Tesla stock investors
Alongside the comments, SpaceX unveiled its first-generation orbital compute satellite, designated AI1. The satellite generates 150 kilowatts of peak power and approximately 120 kilowatts of sustained compute, roughly equivalent to a single Nvidia GB300 AI server rack.
Musk described the AI1 design as simpler than a Starlink satellite, consisting largely of solar cells without the complex phased-array antennas that make Starlink hardware expensive to produce at scale, according to Reuters.
Why SpaceX is making this argument right before its IPO
The timing is deliberate. SpaceX’s IPO filing describes the total addressable AI market as potentially reaching $26.5 trillion, but identifies terrestrial power constraints as the primary factor that could prevent that market from materializing. A power grid that does not need to exist because the data centers are in orbit is a different business proposition entirely.
Orbital data centers address the two constraints most directly threatening AI infrastructure buildout on Earth: power availability and permitting timelines. Solar power in low Earth orbit is available continuously without grid connections, land permits, or utility negotiations.
For investors reading the SpaceX S-1, the orbital data center narrative positions the company not merely as a launch provider but as a potential AI infrastructure company competing directly with Amazon Web Services, Microsoft Azure, and Google Cloud for the $26.5 trillion market it identifies in the filing.
T. Fallon/Getty Images
The technical and financial specifics behind SpaceX’s AI satellite plan
SpaceX’s AI satellite factory in Bastrop, Texas, spans more than 1,000 acres and 11 million square feet of manufacturing floor space. Musk said SpaceX expects to reach meaningful production volumes at the facility by the end of 2027, according to Reuters.
The production targets are ambitious even by SpaceX standards. The company is targeting an annual deployment pace of 1 gigawatt of space-based AI computing by end of 2027, growing by orders of magnitude each year and eventually reaching 1 terawatt of computing power, according to NPR.
The financial scale is visible in the IPO filing. xAI, which merged with SpaceX last year, more than doubled capital expenditure to $12.7 billion in 2025. SpaceX separately spent $3 billion developing Starship.
In Q1 2026, two of SpaceX’s three main business units reported losses, underscoring that this is a company investing heavily in a future that has not yet arrived commercially, according to NPR.
Key figures from SpaceX’s IPO filing and Musk’s June 8 technical presentation:
- SpaceX raised approximately $1.5 billion in a pre-IPO funding round in March 2026 at an implied valuation of $450 billion for just the commercial launch and Starlink businesses alone; the $1.75 trillion IPO valuation therefore implies investors are pricing in roughly $1.3 trillion of value specifically for the orbital data center and xAI businesses that do not yet have meaningful commercial revenue, Investing.com indicated.
- JPMorgan CEO Jamie Dimon met with Musk in the days before the IPO filing; Goldman Sachs is the lead underwriter; both relationships signal that the most powerful institutional capital allocators in the world are actively supporting the $1.75 trillion valuation, according to Investing.com.
- More than 95% of SpaceX’s revenue comes from the private sector; the company explicitly rebutted characterizations of dependence on NASA or government contracts, a disclosure aimed at retail investors who may associate SpaceX primarily with government missions, NPR reported.
- Starlink currently serves more than 5 million subscribers globally and generated approximately $8 billion in revenue in 2025; SpaceX’s IPO filing indicates it expects Starlink revenue to double by 2027 as it expands into maritime, aviation, and enterprise markets, separate from the orbital data center business entirely, according to NPR.
- SpaceX’s Starship rocket, which is central to making orbital data centers economically viable, completed its eighth integrated flight test in March 2026 and achieved full catch of the booster for the first time; Musk has said Starship will reach operational status with full reusability before the end of 2026, which is the prerequisite for the low launch costs the AI satellite business model requires, Investing.com noted.
What the SpaceX IPO and Musk’s orbital ambitions mean for investors
The orbital data center narrative answers the central investor question about why a rocket and satellite company deserves a $1.75 trillion valuation when Amazon, which operates the world’s largest cloud infrastructure, is valued at roughly $2.5 trillion. The answer Musk is giving is that SpaceX is not competing with terrestrial cloud providers on their terms.
Whether that argument sustains the valuation after trading begins depends on execution. The target of 1 GW by end of 2027 is the first milestone investors can anchor to. If SpaceX’s Bastrop facility begins producing AI1 satellites at meaningful volume and early orbital deployments demonstrate viable performance, the $26.5 trillion AI market framing becomes harder to dismiss.
For investors deciding whether to participate, Musk’s June 8 technical presentation is the most specific public statement the company has made about how its AI infrastructure ambitions actually work.
Rather than being a bet on Starlink or Starship, the IPO is a bet on whether a company that has never built a data center can become one of the defining AI infrastructure providers of the next decade.
Related: Morningstar drops bombshell message on SpaceX IPO
