The Rocket Problem Just Became A Starship Factory Problem
Two Failures, One Flight
Starship has spent five years failing in two separate, unrelated ways. Sometimes the ship survives reentry and burns up something else; sometimes it carries a payload that isn't really a payload, a stack of steel mass simulators standing in for satellites nobody had built yet. Flight 13, which launched from Starbase, Texas on July 24, is the first time both problems were solved in the same mission. All 33 Raptor 3 engines on the upgraded Super Heavy booster lit as planned, according to SpaceNews. The upper stage splashed down intact in the Indian Ocean about 65.5 minutes after liftoff, the softest and first fully survived reentry the vehicle has managed, per Fortune's account of the flight. And riding inside it were 20 functioning Starlink V3 satellites, not simulators: SpaceX's Michael Nicolls confirmed they were talking to the ground over both radio and laser links, with telemetry down within about 20 minutes of separation. Those are three different engineering questions, structural survival, thermal protection, and payload integration, that had each been answered on their own before. Flight 13 is the first time they were answered together, in the same mission.
Where The Limit Moves
The reason this matters is what a Starlink V3 satellite is actually built to do. Engadget and Norwegian ground-station operator KSAT put its design ceiling at 1 terabit per second downlink and 160 gigabits per second uplink, roughly ten and twenty-two times what the current V2 satellites manage. Multiply that by a full constellation and the throughput jump is not incremental, it changes what SpaceX can sell to a rural ISP, an airline, or a defense customer. Until Friday, that upgrade lived behind an unresolved question: would Starship actually fly V3 satellites, or just keep flying test mass while the ship itself got fixed? That question is now closed. The open one is production, how many V3 satellites SpaceX can build and how often Starship can fly, not whether the rocket can carry them. Super Heavy's landing burn again lit fewer engines than planned and came in hot, meaning booster recovery, the other half of full reusability, is still the unsolved twin of this story. A rocket that can deliver the constellation is a different machine from one proven to relight enough engines to land intact and fly again next week.
SpaceX now has a rocket that can carry the next Starlink generation and a satellite that proves it works within twenty minutes of release. What remains is cadence: Elon Musk has targeted flying Starship as often as every three to four weeks, with dozens of V3 satellites per flight. That target, not this weekend's splashdown, decides whether the capacity race with rival constellations was actually won here or merely opened.