Gigabit Speeds at Anchor and What Actually happened with Starlink V3 Launched

The pitch was simple: SpaceX launches V3, capacity goes up tenfold, superyacht owners with the right dish get gigabit speeds by fall. The reality, now that it’s actually happened, is messier. That messiness changes what captains should be telling owners this season.

What Actually Happened on September 28

The first batch of fully operational Starlink V3 satellites didn’t reach orbit on September 22, the originally targeted date. It happened six days later than planned, on September 28, aboard Starship’s fourteenth test flight, and even that flight nearly didn’t finish the job. One of the vehicle’s engines was lost shortly after liftoff from Starbase, Texas. Mission control pressed on anyway, and Starship reached orbit for the first time in the program’s history roughly 26 minutes later, releasing all 26 of its V3 satellites as planned. But the engine loss forced an abbreviated flight. Instead of the six orbits and roughly 10-hour mission originally planned, Starship splashed down in the Northern Pacific after about 3 hours and 9 minutes. The 26 satellites are, as of this writing, still gradually raising themselves to their operational altitude. They are not yet fully in position.

It Took Three Tries to Get Here

The path to that September 28 flight is worth knowing, because it’s the real story of why this slipped and why some caution is warranted before anyone promises owners a fast timeline. SpaceX’s first attempt to fly V3 hardware, in May, ended when the Super Heavy booster failed during separation from the upper stage. A second attempt, Starship Flight 13 on July 16, aborted on the pad entirely. Four of the booster’s 33 engines failed to ignite before liftoff. SpaceX fell back on a familiar workhorse in the interim, launching a routine Falcon 9 mission four days later to keep adding older V2 Mini satellites to the constellation while Starship was grounded for repairs. The next Starship attempt, on July 24, got 20 prototype V3 satellites to deployment altitude and confirmed SpaceX could communicate with all of them in space, a real milestone, but the booster failed again during its landing attempt and exploded in the Gulf of Mexico. Because that flight was sub-orbital by design, the satellites themselves burned up on reentry about 20 minutes after release. September 28 was, in other words, the third attempt at getting V3 hardware to orbit, and the first time the full commercial batch, not prototypes, actually got there and stayed.

What V3 Actually Delivers, Per SpaceX’s Own Numbers

It’s worth being precise about what those 26 satellites are capable of, since SpaceX’s own technical page on V3 is more specific than most of the trade coverage around the launch. Each V3 satellite is built for 1 Tbps of downlink and 160 Gbps of uplink. That’s roughly 10 times the downlink and about 22 times the uplink of a V2 Mini satellite, a bigger jump on the upload side than most summaries of this launch mentioned. That capacity is delivered through upgraded phased-array antennas supporting 2,048 beams in each direction, versus 192 downlink and 144 uplink beams on V2. That means each V3 satellite can paint far more individual, adjustable connections across the ground it passes over, which is the piece of the puzzle that actually matters for a busy anchorage full of competing terminals, not just the headline bandwidth number.

Two other specs are relevant to how this reaches a boat at all. Each V3 satellite carries six 400-gigabit space lasers forming part of a petabit-scale mesh network between satellites. That’s what lets traffic move between distant points on Earth without routing through a ground station directly below, which is useful context for why coverage in genuinely remote cruising grounds can work at all. Getting that capacity back down to Earth relies on quad-band ground-station backhaul antennas (spanning the Ka, E, V, and W frequency bands) delivering up to 1.2 Tbps per satellite, more than eight times V2’s backhaul capacity. That’s the connection between the satellite and SpaceX’s ground infrastructure, not the one to a yacht’s dish, but it’s the bottleneck that would otherwise cap what any of this is worth, regardless of how fast an individual terminal can go. All of this draws roughly twice the power of a V2 satellite, which SpaceX says is why the new solar arrays use a different manufacturing process entirely: a continuously produced blanket cut into 19-meter segments and stitched into each finished array. SpaceX also says the V3 satellite platform is what Starship’s cargo capacity is built to exploit at scale, adding roughly 20 times the constellation capacity per Starship launch compared with a Falcon 9 launch of V2 satellites. That’s the entire economic logic behind putting up with Starship’s rockier flight record in the first place. One successful Starship flight is worth many Falcon 9 flights once it’s routinely reaching orbit.

None of this changes the caveat below. These numbers describe satellite and network capacity. They don’t describe what any single terminal can actually pull from it.

The Bigger Complication: Even a Working Dish Has a Ceiling

Reporting from just before the launch made a point worth taking seriously: a V3 satellite pushing a terabit per second toward a yacht doesn’t mean a terabit per second actually reaches it. The terminal’s own radio hardware, not the satellite’s capacity, is what caps the speed a dish can receive, regardless of how much bandwidth SpaceX puts in orbit. As of that reporting, SpaceX had not announced pricing, availability, or a concrete timeline for any terminal upgrade, and the guidance for existing hardware was that “current service performance will not change for your immediate needs.”

There’s one more data point worth including with appropriate caution. An unverified social-media post from an account that tracks SpaceX closely claimed SpaceX has told enterprise customers it expects to roughly double throughput on the Performance-tier maritime dish, from around 475 Mbps today to somewhere near 950 Mbps, by December. That’s a meaningful, welcome jump if it holds, but it’s not gigabit, it’s not yet confirmed by SpaceX directly, it’s aimed at enterprise customers rather than Maritime-plan yacht owners specifically, and AAS has not independently verified it. Treat it as a rumor worth watching, not a promise to relay to an owner.

What to Actually Tell Owners This Season

The honest assessment is a conservative one. The satellites that make gigabit possible have only just reached orbit and are still moving into position. The rocket program that’s supposed to deliver dozens more of them at a time has failed or aborted on three of its last four attempts. And the dish hardware question, whether today’s Performance terminal can even receive V3’s extra capacity without a swap, doesn’t have a confirmed answer yet. Nothing changes at the dock this week. The realistic timeline for anything resembling gigabit speeds at a yacht’s actual terminal has, if anything, moved further out and gotten less certain than it looked two weeks ago. That’s worth knowing before setting any owner’s expectations for this charter season.

captain has written for ALL AT SEA since 2020.