On November 6, 2026, Elon Musk suggested in a post on X that satellites equipped with laser links should be able to beat subsea cables in competition once their number exceeds 100,000. He cited the vast capacity of inter-satellite lasers, and the advantages that light travels faster in space than inside optical fiber and that communication paths can be shorter. However, he gave no specific figures for the capacity of the link between satellites and the ground.
Meanwhile, the next-generation V3 satellites that SpaceX first put into orbit with Starship numbered only 26 as of September 28. The plan to deploy a constellation on the scale of 100,000 satellites is still at the application stage with the U.S. Federal Communications Commission (FCC). Given this situation, the near-term role of satellite communications is likely to be not a full replacement for subsea cables but a backup line for four Gulf states that have no subsea communication route bypassing the Strait of Hormuz. This view rests on the gap between inter-satellite and satellite-to-ground capacity in V3's published specifications, the progress of the deployment plan, and the Gulf countries' structural dependence on subsea cables.
V4 laser count: about 100 in the X post, 10 in the interview
Musk's post was published at 19:32 UTC on November 6, quoting a post by Sawyer Merritt introducing the next-generation V4 satellite. Musk wrote: "With about 100 lasers per satellite, each exceeding 10 Tbps, more than 100,000 satellites should be able to beat subsea cables in competition." He also explained that in space light travels about 40% faster than inside a cable, and that communication paths are not constrained by the shape of land and can be set more directly.
However, a conversation between investor Ron Baron and Musk, released by Baron Capital at 18:01 UTC the same day, gives a different number. According to a transcript published by Merritt, Musk explained that V4's capability would be 10 times that of V3, and said of the next-generation inter-satellite lasers, "10 Tbps each, and we'll put 10 of them on." This "10" came up in the context of long-distance inter-satellite communication, and it cannot be confirmed whether it refers to the same kind of laser as the "about 100" in the X post.
If both refer to the same lasers and we simply calculate at 10 Tbps each, the total capacity per satellite would exceed 1,000 Tbps based on the X post and be 100 Tbps based on the interview figure, a gap of more than tenfold. It is also unclear whether the "10 times V3's capability" that Musk mentioned refers to inter-satellite capacity, satellite-to-ground capacity, or some other performance metric. For now, the only material for judging V4's performance is what Musk said in the interview and the X post.
Behind these remarks is rising tension over subsea communication cables in the Persian Gulf. On October 23, Reuters reported, citing two Iranian security officials, that detailed plans exist to attack Persian Gulf subsea cables and obstruct repair work if the United States and Israel strike again. Maritime data analytics firm Windward also analyzed that between October 5 and 28, three U.S.-sanctioned tankers and one Iranian-flagged fishing support vessel behaved suspiciously along subsea cable routes in the Gulf of Oman. These vessels repeatedly slowed and circled over the cable routes and had switched off their AIS (Automatic Identification System) transmissions.
Specifically, a vessel called CELESTIAL SEA circled for 4.3 hours at a point 0.65 nautical miles from the 2Africa subsea cable route, then stopped transmitting AIS for 86 hours. Windward said this behavior is hard to explain as ordinary cargo shipping and is consistent with cable surveying, or with placing explosives while position broadcasting is off. However, no direct evidence of the vessels' intent has been presented, and there are no reports that any cable was actually damaged. It therefore cannot be concluded that these actions were preparation for an attack on subsea cables.
Inter-satellite laser capacity differs from capacity down to the ground
What Musk emphasized in his November 6 post were figures for the scale of inter-satellite communication: about 100 lasers per satellite, more than 10 Tbps each, and more than 100,000 satellites. He also mentioned the speed of light in space and the reduced latency from more direct paths, but gave no specific figures for satellite-to-ground capacity. He did not address ground station equipment or communication costs, nor did he define what would count as an advantage in competing with subsea cables.
In satellite internet, data sent from a user's terminal goes up to a satellite and, as needed, is relayed to another satellite by inter-satellite lasers. It is then connected to terrestrial networks and data centers via ground stations. To replace subsea cables in carrying high-capacity intercontinental traffic, a system must not only relay data quickly in space but also send and receive large volumes of data to and from the ground at both ends of the path. The number of lasers and the speeds Musk cited are mainly figures for the relay portion in space, and they alone cannot be used to judge the effective capacity of the whole system.
In fact, V3's published specifications set different values for inter-satellite and satellite-to-ground communication. According to specifications published by Starlink and reported by industry outlet Converge Digest, a V3 satellite carries six inter-satellite lasers of 400 Gbps each, for a total capacity of 2.4 Tbps. Meanwhile, downlink capacity for users is 1 Tbps and total backhaul capacity between the satellite and ground stations is 1.2 Tbps, all of them design values.
In other words, already with V3, capacity toward the ground is less than half the total inter-satellite laser capacity. Even if V4 raises inter-satellite laser capacity 10-fold or 100-fold, capacity to the ground will not necessarily grow at the same rate. Without ground-side facilities that can process large volumes of data, increasing relay capability in space does not mean international communication capacity comparable to subsea cables can be secured.
On the other hand, the claim that satellite communications can have an advantage in long-distance latency has a technical basis. In a 2018 paper, Mark Handley of University College London (UCL) showed that light in a vacuum travels about 47% faster than inside optical fiber, and argued that a major advantage of low-Earth-orbit satellite networks is reduced latency on long-distance links. Musk's "about 40%" is close to this difference. However, while that advantage grows for long distances such as intercontinental links, it is limited over short distances. In the same paper, Handley also noted that the bandwidth of a single satellite path alone would probably be insufficient to have a major impact on the long-distance fiber network business.
The scale of international communication used as a comparison is also enormous. According to telecom market research firm TeleGeography, global international bandwidth usage exceeded 6.4 Pbps (6.4 petabits per second; one petabit is 1,000 terabits) in 2024, with a compound annual growth rate of 32% from 2020 to 2024. Content and cloud providers such as Google and Meta account for about three-quarters of it. These providers' traffic mainly starts and ends at data centers, so for satellite networks to carry a large share of international traffic, the ground-side facilities connecting to data centers must be able to handle large volumes of traffic.
V3 starts with 26 satellites; the 100,000-satellite Gen3 plan is at the FCC application stage
To assess Musk's vision, Starlink's deployment status and the situation around subsea cables need to be considered separately. The main developments since July 2026 are as follows.
| Date | Event | Source |
|---|---|---|
| July 2026 | SpaceX applies to the FCC for up to 100,000 Gen3 satellites | DCD |
| September 28 | Starship reaches orbit for the first time and releases 26 V3 satellites | CircleID, Spaceflight Now |
| October 23 | Iranian security officials tell Reuters about plans to attack subsea cables | Reuters |
| November 6 | Musk posts on X that more than 100,000 satellites should be able to beat subsea cables | X |
| November 8 | Starlink satellites in orbit reach 11,207, of which 11,193 are operational | KeepTrack |
When Musk mentioned competition with subsea cables on November 6, Starship's delivery of V3 satellites into orbit had only just begun with the 26 on September 28. SpaceX expects those 26 satellites to add up to 26 Tbps of capacity, but this too is a design figure, and actual operating performance has not been confirmed. Moreover, the 100,000-satellite Gen3 deployment plan is at the FCC application stage, and when the planned constellation would be operational has not been determined.
Meanwhile, the security risk to subsea cables is already a concrete concern. Alan Mauldin of TeleGeography told Reuters that international bandwidth passing through the Strait of Hormuz is less than 1% of the world total, so the impact on a global scale could be limited. However, he said the actual impact depends on which routes operators can reroute traffic to and how much spare capacity the alternatives have. The figure of under 1% of the world total is only the share of traffic passing through the Strait of Hormuz and does not represent an upper limit on the scale of damage if cables are cut.
According to an analysis Windward published on October 26, there are six active subsea cables crossing the Strait of Hormuz in total, including 2Africa, AAE-1 and FALCON. Among them, Iraq, Kuwait, Bahrain and Qatar all have every subsea cable they use landing inside the Persian Gulf, and have no subsea cable route bypassing the Strait of Hormuz. Terrestrial routes exist, and Iraq is also building an overland route toward the Turkish border. But CircleID points out that if multiple subsea cables went down at the same time, terrestrial alternatives alone might not be able to handle the rerouted traffic.
The fact that the Iranian side has also mentioned obstructing repair work means a risk of prolonged communication outages for these four countries. With V3 deployment only just started and V4's satellite-to-ground capacity not yet disclosed, it is hard to imagine Starlink immediately replacing subsea cables. Rather, for Iraq, Kuwait, Bahrain and Qatar, which have no subsea route bypassing the Strait of Hormuz, satellite communications are expected to serve as a backup line complementing the existing international network.
To actually function as a backup line, however, several conditions must be met: satellite terminals and ground stations must be authorized to operate in each country, traffic must be relayable to ground stations outside the region, and there must be enough capacity to take over part of the traffic when subsea cables go down. Whether these conditions can actually be met cannot be judged from the published specifications of V3 or Gen3 alone.
Dependence on subsea cables for international communication is not an unrelated issue for Japan either. According to the Ministry of Internal Affairs and Communications' 2023 White Paper on Information and Communications, about 99% of Japan's international communication passes through subsea cables, and landing points are concentrated on the Boso and Shima peninsulas. Diversifying routes through satellite communications is an important issue to consider in Japan too, from the viewpoint of increasing the redundancy of communication infrastructure.
A counterargument: 100,000 satellites would mean 100 Pbps. Three conditions will decide practicality
The view that satellite communications will remain a backup line for now can also be countered from the scale of the constellation SpaceX plans. According to DCD, the Gen3 plan SpaceX submitted to the FCC in July 2026 calls for deploying up to 100,000 satellites, each with 1 Tbps of downlink capacity. Simply multiplying these gives 1 Tbps × 100,000 = 100 Pbps, far above the 6.4 Pbps of global international bandwidth usage in 2024 compiled by TeleGeography.
In his 2018 paper, Handley also examined the possibility that in low-Earth-orbit satellite networks, many communication paths can be built between the same two cities and used in parallel to increase capacity. Starlink's satellite deployment capability has also already reached considerable scale. According to KeepTrack's tally, cumulative Starlink satellites launched from the first launch in May 2019 through November 8, 2026 reached 13,096.
Still, the reason to regard satellite communications as a backup line rather than a replacement for subsea cables at this point is that 100 Pbps and 6.4 Pbps are different metrics. 100 Pbps is the total of the design downlink capacity of satellites placed around the world, whereas 6.4 Pbps is a tally of international communication bandwidth actually used. The former is capacity spread across a wide area of the globe, while the latter represents bandwidth used for international communication. A simple comparison of size therefore cannot support the conclusion that satellite networks can replace the traffic of subsea cables.
Particularly important is how much of a satellite's downlink capacity can be concentrated at a specific location. Subsea cables connect intercontinental data centers and major communication hubs with high-capacity lines. Satellite capacity, by contrast, is distributed over wide areas, and whether the same level of capacity can be concentrated at places equivalent to subsea cable landing points cannot be judged from published figures. On this point, the assessment could change if SpaceX discloses ground-station capacity and specific operating plans.
To judge whether satellite communications can become an alternative to subsea cables, or whether their role will center on backup lines, the following three developments will be important.
- Disclosure of V4 satellite and ground-station capacity, and wholesale contracts for international transmission capacity. If SpaceX discloses the specific capacity of V4 satellites and ground stations, and announces contracts to wholesale capacity to carriers and cloud providers on international lines linking Europe, the Middle East, Asia and elsewhere, the debate will move from comparing design figures to evaluating an actual commercial service. This would be an important basis for revising the current view.
- FCC approval of the Gen3 plan and announcement of V4 launch timing. If the plan to deploy up to 100,000 satellites is approved and V4 deployment timing becomes concrete, it will be easier to estimate how long until a high-capacity satellite network is available. That could shorten the "for now" period in which satellite communications are positioned as a backup line.
- Formal adoption of satellite links as a backup for international communication by the four Gulf states. If the governments or carriers of Iraq, Kuwait, Bahrain and Qatar contract for Starlink as a backup line for international communication, it would be a concrete case of satellite communications being adopted to compensate for the vulnerability of subsea cables.
The 100,000-satellite network Musk envisions could, judging by design capacity alone, rival the existing international communication network. But no matter how much data can be relayed between satellites, it cannot replace subsea cables unless it can be connected to terrestrial networks where it is needed. Whether the actual capacity of V4 satellites and ground stations becomes clear and commercial contracts for international transmission capacity are concluded will be an important indicator of whether Starlink can develop from a backup line into a network that seriously carries intercontinental communication.