Author: iFanr

Just now, SpaceX officially acquired xAI.

The combined valuation of the two companies is estimated at $1.25 trillion. SpaceX would be valued at $1 trillion and xAI at $250 billion, in an all-stock transaction with no cash involved.

This is not the first time Musk has done this kind of "left hand taking over right hand" thing, assembling the companies he controls like "Lego bricks".

In November 2016, Tesla acquired SolarCity, another new energy company effectively controlled by Musk, for a staggering $2.6 billion. This deal was met with a lawsuit, with investors accusing Musk of using Tesla's then-abundant cash reserves to "rescue" the underperforming SolarCity.

In March 2025, xAI, which had been established not long before, acquired X (formerly Twitter), which was controlled by Elon Musk, with the aim of forming a closed loop of social platforms, data, models, computing power, and talent. After that, X became the testing ground for xAI's Grok model.

And then today, SpaceX acquired xAI.

In the new architecture, Grok and X represent model and social content platform scenarios, while Starlink and Starship represent orbital networks and launch capabilities. Taken together, SpaceX's acquisition of xAI is far more than a simple capital operation; it's more like a reorganization that packages together the "upstream and downstream of the AI ​​industry chain."

This news was announced by Musk himself on the SpaceX website, accompanied by a lengthy open letter. The core idea in the letter is clear: move the AI ​​data center to space.

The reason is simple and direct:

The growth of terrestrial computing power faces energy and heat dissipation constraints; further expansion is both expensive and slow, and could also trigger community and environmental controversies. For an AI company burning through billions of dollars every month, this pressure is unlikely to disappear in the short term.

He did the math: to use even one millionth of solar energy, we would need a million times the current total energy consumption of human civilization. Earth simply doesn't have the resources for that. Therefore, the only solution is to move AI into space.

What are the advantages of space? Unlimited solar energy, and perpetual sunshine.

Musk's plan is to launch one million orbiting data center satellites. This would amount to 1 million tons of AI satellites launched into orbit annually, generating 100 kilowatts of computing power per ton, equivalent to an additional 100 gigawatts of AI computing power each year.

The ultimate goal is to launch 1 terawatt of AI satellites from Earth annually. Furthermore, he envisions establishing a manufacturing base on the moon, using lunar resources to build satellites and launching them into deep space via electromagnetic mass actuators. This would allow for the deployment of 500 to 1000 terawatts of AI satellites annually.

Musk did not disclose details of the acquisition in his public letter. Reports indicate that the deal was communicated to employees via internal memos, and the full terms and conditions were not publicly disclosed. To outsiders, it appears more like a merger that has already occurred, rather than a transaction with ample room for negotiation. This pace aligns with Musk's past operating style: first set the direction, then fill in the details.

According to Bloomberg, citing sources familiar with the matter, the merged company is valued at approximately $526.59 per share. xAI just completed a $230 billion funding round in January, while SpaceX planned to sell its shares at an $800 billion valuation in December.

The merger was completed entirely in stock, with no cash payment involved.

The Structure Behind Mergers and Acquisitions

Looking back at mergers and acquisitions over the past few years, a clear integration trajectory emerges.

In 2022, Musk, like a "barbarian at the gate," bought Twitter and renamed it X. In 2025, he had xAI acquire X in a stock swap, tying the content platform and the model together.

Now that SpaceX has entered the picture, it's like adding "orbital infrastructure" to this closed loop.

Musk has always been adept at reorganizing his companies into a complete, interdependent system; this move is simply putting a piece of the puzzle in a more complete position.

The pressure from the capital side is also very real. On the one hand, xAI needs to catch up with competitors like OpenAI and Anthropic, and on the other hand, it is burning through cash at a rapid pace.

While SpaceX is more mature, its revenue structure also suffers from a concentration problem, with Reuters reporting that a significant portion of its revenue comes from Starlink's own satellite launches.

Merging the two companies means that xAI has secured stable launch capabilities, while SpaceX has obtained long-term satellite launch contracts, creating a mutually supportive financial structure.

With the addition of funding, the puzzle is more complete.

According to Bloomberg, xAI's monthly cash burn could reach $1 billion, while Tesla and SpaceX have invested approximately $2 billion in xAI, making the merger seem more like a "systemic capital injection."

Meanwhile, SpaceX is preparing for its IPO, with market rumors suggesting it may launch in 2026, but it remains unclear whether the acquisition will affect the timeline.

For investors, there is a logic to creating a "layered asset" deal before an IPO: SpaceX's aerospace attributes combined with its AI narrative make it easier to obtain a higher valuation anchor.

Musk's open letter was very direct:

"SpaceX has acquired xAI, and the combined company will become the most ambitious and vertically integrated innovation engine on Earth (and beyond), combining artificial intelligence, rocket technology, space-based internet, direct-to-mobile device communication technology, and X, the world's most important real-time information platform. This is not only the next chapter in the mission of SpaceX and xAI, but also the next book: building a sentient 'sun,' understanding the universe, and extending the light of consciousness to the stars."

Sending data centers into space

The most compelling part of the announcement was an extreme judgment made by Musk:

He wrote, "In the long run, space AI is the only way to scale," and gave a timeline—within 2 to 3 years, space will become the cheapest source of AI computing power.

Space has a nearly constant solar energy source; in his own words, "It's always sunny in space."

To support this approach, SpaceX plans to launch satellites with computing capabilities using Starship, directly transferring computing power into orbit.

Moreover, the scale is not "a few hundred": public reports mention that SpaceX has submitted an application to the FCC, hoping to obtain approval to launch up to 1 million data center satellites.

These satellites will generate long-term and huge launch demand, which in turn will provide stable revenue for SpaceX's launch system —this sounds reasonable, but objectively speaking, it also crosses the line of "related-party transactions." Of course, as mentioned earlier, this is not the first time Musk has engaged in such transactions, and today's US regulators have more pressing matters to attend to and don't have time to deal with him.

Starlink currently has over 9,000 satellites in orbit, making it the world's largest satellite network. This scale itself is a competitive advantage, and it also explains why only a handful of companies can attempt this route.

In a grander narrative, Musk also described this route as the beginning of the "sentient sun," and even mentioned extending the "light of consciousness" to the stars.

It sounds a bit abstract, but you can understand it like this: he imagines the scope of human civilization's influence and being influenced as an expanding beam of light—this is also the core emotional driver in this merger narrative.

Of course, not everyone believes in this route.

Microsoft President Brad Smith publicly questioned the feasibility of space data centers; Google also proposed a research project like Project Suncatcher, but it remains in the experimental stage.

From a practical perspective, xAI itself has not given up on expanding its ground computing power. It is advancing a data center project worth approximately $20 billion in Mississippi, indicating that "going to the sky" is more of a long-term solution than a short-term replacement.