SpaceX flew Starship on its 15th test flight Friday evening from Starbase, Texas, and put the upper stage into a nominal low-Earth orbit. It’s only the second time the rocket has reached orbit. According to Ars Technica, the main goal of the mission went off cleanly: all 26 Starlink V3 satellites deployed. The first-stage booster didn’t make it home, though. SpaceX diverted it into the Gulf of Mexico instead of trying to catch it at the launch tower.
The Quick Version
- What happened: Starship’s 15th flight reached orbit and deployed 26 Starlink V3 satellites about 40 minutes after liftoff.
- What went wrong: Because of an unspecified issue, SpaceX ordered the booster to divert into the Gulf, where it exploded. Observers more than 10 miles away felt the blast.
- Why it matters: Starship can now deliver real payloads to orbit. Reusability is the part SpaceX still hasn’t solved.
- What’s next: Ars Technica’s headline says Saturday could bring a possible catch attempt.
A Rocket That Actually Delivers
The big news here is the payload, not the launch. Starship carried 26 Starlink V3 satellites and released them through a narrow door at the base of the spacecraft. These satellites are huge. Their solar panels span 60 meters end to end once they’re fully unfolded.
All 26 reached orbit safely. Each one will now use its own onboard propulsion to climb to its operating altitude. As Ars Technica puts it, that makes Starship “an operational, orbital rocket.” It’s also the largest and most powerful machine humans have ever launched into space.
That’s a real milestone. For more than three years, Starship flights were tests of the vehicle. This one put working hardware into orbit for paying service.
The Booster Problem
The ascent went perfectly. The first stage released the upper stage over the Gulf of Mexico right on schedule. Then SpaceX hit a snag during the return.
The plan was to bring the booster back to Starbase, catch it with the launch tower, refurbish it, and fly it again. Because of an issue SpaceX hasn’t explained yet, the booster was sent into the Gulf instead.
Some context helps here:
- SpaceX has landed and reflown Starship boosters before.
- It hasn’t done that yet with Version 3, the newest version of the vehicle.
- V3 is the version SpaceX needs working at scale, so this loss matters more than the fireball suggests.
Why the AI Industry Should Care
This looks like a space story. It’s also an infrastructure story, and AI runs on infrastructure.
Bandwidth. Starlink V3 satellites are built to add a lot of capacity to SpaceX’s network. More orbital bandwidth means more places where cloud AI services, edge inference, and remote data pipelines can run reliably. Think ships, mines, farms, disaster zones, and rural areas with weak fiber.
Cost per kilogram. Starship’s whole business case depends on cheap, frequent launches. If SpaceX gets full reusability working, the cost of putting heavy hardware in orbit drops sharply. That’s the economic assumption behind ideas like orbital data centers and space-based compute. Those ideas have moved from science fiction into serious pitch decks over the past year.
Flight rate. Ars Technica describes the next challenge plainly: make Starship reusable and raise its flight rate dramatically. Until boosters come back reliably, every flight burns very expensive hardware. That limits how fast SpaceX can deploy capacity.
My Take
What stands out is how mixed this result is. Getting a payload to orbit was the hard part, and Starship did it without drama. Losing the booster is a setback, but a familiar one. Every reusable rocket program has gone through this phase, and SpaceX went through it with Falcon 9 years ago.
If you’re a practitioner, the timeline is the thing to watch. A working Starship with reliable booster catches turns SpaceX into the dominant hauler of orbital infrastructure, including the networks AI products will rely on outside traditional data centers. One clean catch on V3 would shift that timeline in a big way.
What to Watch
- Saturday: Ars Technica’s headline points to a possible catch attempt.
- Root cause: Whatever SpaceX says about the booster issue will show whether this was a quick fix or a design problem.
- Starlink V3 ramp: Watch how fast these satellites reach their operating orbits and come online.
Saturday could show whether V3 is close to reusable or still a few flights away. You can find the full launch-day report at Ars Technica.