Will SpaceX Cease to Be a 'Rocket Company'? The Triple-Threat Strategy for Space Infrastructure Advancing Behind the Stock Price Plunge
SpaceX's stock price has temporarily fallen by nearly 50% from its high immediately after going public.
While the June 2026 IPO garnered high expectations as a growth company spanning rockets, satellite communications, and AI, its market capitalization shrank by over $1 trillion by the end of July. With concerns over the profitability of AI investments and delays in Starship development, the market has rapidly shifted toward a cautious stance.
However, it is premature to judge that it was 'all hype' just by looking at the stock price.
What SpaceX is advancing is vertical integration that connects 'transportation' via Starship, 'communications' via Starlink, and 'computation' via future orbital AI data centers. If this vision is realized, the company could transform from a mere space firm into a massive infrastructure company that competes with telecommunications carriers and cloud providers.
It is necessary to view short-term stock prices and long-term business structures separately.
Why 'AI x Space Infrastructure' Now?
The first piece supporting SpaceX's vision is the large reusable rocket, 'Starship'.
In the space business, the cost of transporting satellites and equipment into orbit has long been one of the greatest constraints. If Starship achieves full reusability and significantly lowers the cost of orbital insertion, the economic rationale for transporting large quantities of satellites and computing equipment into space will be completely transformed.
This is directly linked not only to the expansion of Starlink but also to the future orbital AI data center concept.
Admittedly, Starship is not yet a completed technology. With recent test flights failing to fully recover the booster, uncertainty remains in the development schedule. A significant reduction in transportation costs is merely a prerequisite for making the entire vision viable.
Avoiding the Constraints of Terrestrial Data Centers in Space
Due to the rapid expansion of generative AI, power supply and cooling capacity have become major bottlenecks for terrestrial data centers.
This has led to the idea of 'orbital data centers' that utilize solar power and heat dissipation into space.
Several companies are already in motion. In November 2025, the U.S. company Starcloud launched a satellite equipped with NVIDIA's 'H100' GPU to demonstrate AI processing in orbit. Google is also advancing its 'Project Suncatcher' satellite concept, which will carry AI 'TPU' chips, with plans to launch a prototype in early 2027. Blue Origin is also reported to have applied to U.S. authorities for an orbital data center concept.
SpaceX is also advancing the 'StarMind' orbital AI constellation concept, premised on collaboration with xAI.
According to reports, the first-generation 'AI1' unit has an average power consumption of about 120kW and a peak of about 150kW. It is equipped with large solar arrays, and a vision for deploying an extremely large satellite constellation in the future has been presented.
However, this is not a business where commercialization is imminent at this point.
There are challenges completely different from those on the ground, such as radiation-hardened semiconductors, maintenance in case of failure, orbital maintenance, and launch costs. It is generally viewed within the industry that full-scale practical application will not occur until the 2030s, after Starship operations have matured.
'While orbital data centers are attractive in terms of power and cooling, they incur other costs such as radiation resistance and maintainability. It will likely take several generations of demonstration to judge their economic viability.' (Daisuke Okazaki, Information and Communications Expert)
Why Starlink
Will Not End with 'Communications'
The second piece is Starlink.
By combining 'Direct to Cell,' which communicates directly from satellites to standard smartphones, with optical inter-satellite links (ISL) that connect satellites via laser, a wide-area network can be built that reduces reliance on terrestrial base stations and fiber optics.
Direct to Cell launched service in July 2025 through a partnership with T-Mobile and has expanded its functionality from messaging to broader data communications.
On the other hand, in April 2026, the U.S. FCC withheld part of the frequency expansion application, and the situation where regulatory decisions dictate the speed of expansion continues.
What is important is that SpaceX is not simply trying to provide a 'satellite communication service'.
Transporting computing resources to space with Starship, communicating with Starlink, and executing AI processing in orbit. The essence of this strategy lies in designing transportation, communication, and computation as a single network.
How Will the Role of Telecommunications Carriers Change?
As this structure advances, the relationship with existing telecommunications carriers will also change.
NTT Docomo, KDDI, Verizon, and AT&T have invested huge sums in base stations and fiber optic networks over many years. Meanwhile, direct satellite communication can cover areas where terrestrial networks are difficult to make profitable, such as mountainous regions, remote islands, and the ocean.
Therefore, satellite operators and telecommunications carriers will not necessarily be in total confrontation.
Just as KDDI is deploying direct satellite communication using Starlink, a form where existing carriers incorporate satellite operators as 'last-mile complements' has already begun.
In other words, in the future communications market, a hybrid network that combines both may become mainstream rather than a binary choice of 'terrestrial network or satellite network'.
The Next Competitor Might Be AWS Rather Than Telecom Companies
An even greater change could occur in the cloud market.
Hyperscalers like AWS, Microsoft Azure, and Google Cloud continue to invest heavily in terrestrial data centers.
In contrast, SpaceX is a rare company that owns the physical infrastructure itself, including rockets, satellites, and communication networks.
If orbital data centers become viable, they will be able to provide communication lines and computing resources as a set. This could become a competitive axis that cloud providers cannot ignore.
SpaceX's prospectus reportedly indicates that providing computing resources to AI companies like Anthropic and Google is a future revenue source, suggesting that the company is conscious of a position like a 'space-based cloud'.
Evaluating the current SpaceX as a rocket company means overlooking these structural changes.
The Risk of One Company Controlling National Infrastructure
However, vertical integration has other problems.
Communication and AI computing infrastructure are infrastructures directly linked not only to the economy but also to national security.
If one company broadly controls transportation, communication, and computation, that company's management decisions or system failures could have national-level impacts.
Therefore, competition policy and security policy are unavoidable in SpaceX's business expansion.
The fact that the U.S. FCC is carefully reviewing frequency usage should be viewed not just as a technical procedure, but in the context of vigilance against infrastructure concentration.
In Europe and Japan as well, as reliance on satellite communications increases, the debate over 'how much should we rely on a specific private company for core infrastructure' will become unavoidable.
This is a risk that should be viewed calmly, separate from SpaceX's growth potential.
Can Huge Investments Really Be Recovered?
Another risk is, of course, funding.
Expanding Starship, a massive satellite network, AI satellites, and communication infrastructure all at once requires extremely large upfront investment.
If the full operationalization of Starship is delayed, launch costs will not decrease as expected. If orbital AI cannot achieve the expected efficiency, the recovery period for AI investment will also be extended.
It is highly likely that these uncertainties are reflected in the current stock price decline.
The market immediately after the IPO valued SpaceX highly based on the growth story of 'Rocket Company + Starlink + AI'. Since then, investors have begun to return to the realistic question of 'can those three businesses really be monetized at the same time?'
A stock price adjustment is not synonymous with the failure of the business concept itself. Conversely, having a grand vision is not synonymous with the current stock price being undervalued.
These two must be considered separately.
Three Indicators to Watch More Than Stock Price
When considering the future of SpaceX, the essential variables are relatively clear.
The first is how close Starship can get to full reusability and actually lower the launch unit cost.
The second is whether orbital AI data centers can move beyond the demonstration stage and show economic rationality compared to terrestrial facilities.
The third is how much regulatory authorities in the U.S. and other countries will tolerate SpaceX, which vertically integrates communication, computation, and transportation.
If these three move forward, the company could exceed the current category of 'space company'.
Conversely, if even one of them stumbles significantly, a revision of the long-term growth expectations currently priced in will be unavoidable.
Summary
SpaceX's stock price plunge can be seen as a result of the market re-evaluating concerns over AI investment and the uncertainty of Starship development.
On the other hand, behind the scenes, a very long-term infrastructure strategy is underway that integrates 'Starship = transportation', 'Starlink = communication', and 'orbital AI = computation'.
What should be noted is not the daily stock price itself, but whether this triple-threat model can really reach the point of having economic viability.
SpaceX's next competitor is not just rocket companies. It is becoming a global infrastructure company itself, including telecommunications carriers and cloud providers.
(Reference Article = BUSINESS JOURNAL)
