SpaceX Just Solved Starmind and the Space Industry’s Biggest Bottleneck

On August 25, 2026, SpaceX unveiled plans to construct a new US$100 billion spaceport in Louisiana (United States), dubbed Starbase, LA, with construction slated for the beginning of 2027. Starbase LA is planned to be a self-sustaining spaceport, with 5 complexes with 2 Starship towers each, for 10 launch pads total, each with its own propellant farm, with fuel produced on-site. Power generation, deep water shipping capabilities, vehicle processing facilities, and factories will also be located on-site. The site will span 130,000 acres and bring Starship launch pads available from 1 active + 4 under construction today, to 15 by the time Starbase, LA becomes operational—5X the operational pads available for Falcon 9 at any given time. SpaceX has indicated this would amount to 2 flights per day, per pad, or up to 30 launches per day.

This announcement comes after speculation recently hit the industry around SpaceX ending commercial Falcon 9 launches around 2028 to shift resources to Starship. SpaceX public prospectus filings had always confirmed an eventual wind down to the Falcon 9 program in favor of Starship, but “when” is still an open question. What these developments highlight, however, is that SpaceX and other new space companies have already begun allocating capital for new space infrastructure like Orbital Data Centers (ODCs) and Space-Based Solar Power (SBSP), which are contingent on a commercially ready launch platform (Starship) still under development. The business case for these new systems rides on the success of Starship and overcoming the launch bottleneck has never had higher stakes.

This announcement is a significant development for the space industry and a boon for emergent space infrastructure like ODCs. Launch is one of the single largest cost inputs for any space system, and the launch demand required to support the scale and number of space systems filed far exceeds the supply. This means launch is not only expensive (~US$4,000 /Kilogram (kg) on a reusable booster) but increasingly difficult to book, as commercial launch companies begin to vertically integrate and use their own launch supply for their systems. Starbase, LA changes the launch supply outlook and launch cost input significantly. 

From SpaceX’s own Initial Public Offering (IPO) prospectus, Starship V3 is designed to carry up to 100 metric tons to Low Earth Orbit (LEO) and deploy 60 V3 Starlink satellites or 50 V2 Mobile satellites per launch, over 5X the launch capacity of a single Falcon 9. Based on current specifications for Starmind ODC satellites (AI1)—SpaceX’s ODC systems—this equates to roughly 30 to 50 satellites and between 4 Megawatts (MW) and 6 MW of sustained orbital compute each launch. The expansion of usable launch sites by 3X and moving more manufacturing and supply to the spaceport will also dramatically improve launch costs, as the consistent reuse of Starship compounds cost savings, exchanging many inputs that would normally be for constructing an entirely new rocket for refurbishment instead. From ABI Research’s analysis, the point where launch costs turn favorable for ODC systems would require close to 100 completed reuse flights by a single Starship (bringing costs to sub-US$500/kg), and the realistic timelines until space companies see the economic advantages of Starship over Falcon 9 are still a story for the 2030s.

This is the signal the industry has been banking on. The argument against space systems as a primary solution is often attributed to much higher costs, but with launchers like Starship and spaceports the scale of Starbase, LA, this gap will close more rapidly than skeptics think. Assuming SpaceX does end support for Falcon 9 by 2028, it would almost certainly be because Starship has reached parity or surpassed Falcon 9 launch volume (roughly 165 Falcon 9 launches, or 2,400 tons to orbit, equivalent to ~20 Starship flights per year)—not before. A lot of ODC enthusiasm is capital chasing an infinite space Total Addressable Market (TAM), capitalized today on what gets built tomorrow. This development signals that capital is flowing exactly where it needs to ensure that the promise of tomorrow does in fact get built.

ABI Research has done some modeling on the economics of ODCs. Given that the major inputs fall, principally launch and manufacturing scale, which these developments have positively signaled they will, levelized cost per watt approaches terrestrial data center levels by the mid-late 2030s. Our assumptions banked on a few things: the space supply chain embracing higher-volume and lower-cost solutions, while Starship becomes commercially operational. This development reinforces our outlook. But this doesn’t matter because orbital wins on substitution. It does matter because once the pricing penalty shrinks, space-native workloads move into a wider addressable market, likely some terrestrial but more space-native.

The development of Starbase, LA is the space industry’s proof point that launch capacity is on a trajectory to meet launch demand, and at a speed that brings some SpaceX timescales, particularly for Starlink upgrades and Starmind deployment, into reality. The focus for stakeholders, therefore, should not be if these systems launch, but when. The first successful mission with a full Earth orbital launch, orbital insertion burn, and landing by Starship will be the key signal that the launch cycle is effectively ready. All of this points to an interesting future, and a new direction that vertically-integrated space companies may go.

In this regard, Starbase, LA won’t be the last spaceport built by SpaceX and other space companies, and the scale ambition of the space industry dictates that there will be more. Space companies need to increasingly grow their ground footprint to service increasing scale in space, the terrestrial launch facilities, and ground stations, and massive manufacturing sites are a requisite. The terrestrial supply chain, logistics, and connectivity for these operations is substantive and not inexpensive. Not everybody will be able to do it, but the new space companies with deep pockets, alongside AI houses and data center players, are poised to be significant drivers for infrastructure development on the ground.