What exactly did Musk claim?
On November 6, Musk posted on X: "~100 lasers per satellite, each operating above 10 Tb/s, with over 100k satellites should be able to outcompete undersea cables. Speed of light in space is ~40% faster than cable and the path is direct, instead of following the outline of land masses." In a separate remark shared the same week, he said "Starlink version 4 will have 10x the capability of V3 (which itself is 10x better than V2)."
These are claims in social media posts, not a product announcement: no V4 specifications, deployment schedule or service pricing have been published. We covered a related question - why a 10 Tb/s laser link is not your home download speed - earlier this week. This article is about the newer claim: taking on the backbone cables themselves.
How dominant are undersea cables today?
Very. TeleGeography, which tracks the industry, counts more than 600 active and planned submarine cables and over 1.5 million kilometres of cable in service as of early 2026. A single modern cable can carry enormous capacity: TeleGeography cites the MAREA cable at 224 terabits per second of potential capacity. Google, Meta, Microsoft and Amazon are all major investors in new cables, and private network operators have been deploying capacity faster than traditional backbone carriers.
Satellites, by contrast, are a rounding error in international connectivity. Statistics from the US Federal Communications Commission, cited by TeleGeography, indicate satellites account for just 0.37% of all US international capacity. Cables also win on resilience by design: operators spread traffic across multiple cables so a single break - usually a dragged anchor or fishing vessel - reroutes without users noticing.
What would Starlink have to prove?
First, the hardware has to exist. V4 comes after V3, and Starship only first carried V3 satellites to orbit on September 28, according to TIKR's analysis. SpaceX president Gwynne Shotwell expects a few hundred V3 satellites in orbit by mid-2027 - a fraction of the 100,000-plus fleet in Musk's scenario. Musk's own Starship cadence guidance has slipped from a flight a day within a year (August) to weekly flights in the first half of 2027 (October).
Second, the economics have to work. TIKR notes the cables are already in the water: to take their traffic, Starlink would have to beat them on price as well as speed, and that price must cover the satellites, the lasers and the launches. Cables are typically amortised over decades, and their owners upgrade endpoint equipment gradually as demand grows - a very cheap cost base to compete against. Musk's physics point about light moving faster through space than through glass fibre is real and matters for specialised latency-sensitive routes, but most internet traffic is bought on price per bit, not milliseconds.
Would it change anything for regular internet users?
Not directly, and not soon. Backbone capacity is a wholesale market: if a Starlink backbone business ever materialised, the plausible effects for consumers would be indirect - more competition on some intercontinental routes, marginally lower latency on certain long-distance paths, and another buyer-side option for the carriers and content companies that actually purchase capacity. Your home fibre, mobile or Starlink plan would not change because of it.
The nearer-term context is terrestrial: TIKR reports Starlink's consumer revenue rose 57% year over year in the third quarter to $3.12 billion, but consumer average revenue per user fell to $69 from $83 as subscribers grew faster. A backbone business would be a revenue line that does not depend on signing up more households - which explains why SpaceX is talking about it, not evidence it is close.
What if you are choosing an internet connection now?
Decide on the service available at your address today - satellite backbone ambitions should not factor into a purchase decision in 2026. If you are signing up for Starlink, check the price and terms shown at your own checkout, and our referral link is below. Check Starlink's eligibility and the reward shown for your order; compatibility with other offers is not assumed.
Frequently asked questions
Did SpaceX announce a V4 satellite or a cable-replacement service?
No. The claims come from Elon Musk's posts on X in November 2026. No V4 specifications, launch schedule or pricing have been published.
How much internet traffic do undersea cables carry?
Nearly all of it. FCC figures cited by TeleGeography put satellites at about 0.37% of US international capacity, with more than 600 submarine cables in service or planned.
Is satellite latency really lower than cable?
Light does travel faster through space than through fibre, as Musk notes, but real-world latency depends on routing, ground stations and network design - not just physics.
When could V4 satellites launch?
No date has been given. SpaceX expects a few hundred of the preceding V3 satellites in orbit by mid-2027, and Starship flight cadence guidance has repeatedly slipped.
Could Starlink make cables obsolete?
Cables are cheap, proven and constantly upgraded, so most analysts treat full replacement as unlikely. A more realistic outcome, if V4 succeeds, is Starlink winning some specialised traffic.