Most of the “wireless” world is not wireless
Many people picture international internet traffic as bouncing off satellites. In practice, the great majority of data between continents travels through fiber-optic cables lying on the seabed. Exact shares are hard to pin down, but commonly cited estimates put it well above 90 percent, and it is safe to say that cables carry the bulk of intercontinental traffic. Satellite internet matters for remote regions, ships and aircraft, but satellites carry a small fraction of the total volume. The reasons are mostly capacity and cost: a single fiber cable can carry far more data, and with lower delay, than a satellite link.
How data crosses the ocean
A submarine cable carries data as pulses of light through glass fibers. A modern cable typically contains a number of fiber pairs, each of which can carry many separate channels of light at slightly different wavelengths, a technique that multiplies capacity. At each landing point, the cable comes ashore at a station that connects it to the land network and on to the wider internet.
Light fades as it travels, so over thousands of kilometers the signal would be lost without help. Along the route, at intervals often tens of kilometers apart, devices called repeaters amplify the light. These repeaters need electric power, which is sent down the cable itself through a copper or similar conductor from the stations on shore, so a cable is also a very long power line.
The result is a trade-off. Adding repeaters and fibers costs money, and the cable’s total capacity depends on the equipment connected at each end, which can be upgraded over the years without replacing the cable.
What a cable looks like
Cable design depends on where it lies. In the deep ocean, where hazards are few, a cable can be quite slender, in the region of the thickness of a garden hose, with the fibers protected inside layers of metal, plastic and insulation. Near shore, where fishing gear, anchors and currents are a danger, cables are given extra armoring with steel wires and are often buried in the seabed to protect them. Cables are laid by specialized ships that carry long lengths of cable and pay it out along a planned route, sometimes with machines that plow a trench for shallow sections.
Planning the route
Routes avoid steep slopes, earthquake-prone zones, busy shipping lanes and fishing grounds where possible. Surveys map the seabed before laying. Governments and regulators issue permits, and landing a cable involves local agreements. Politics can shape routes as well, since some countries are reluctant to rely on paths through particular regions.
Delay and distance
Light in glass travels at about two thirds of its speed in a vacuum, and cables rarely follow a perfectly straight line. A trans-oceanic round trip therefore takes tens of milliseconds at the very least, and often more once equipment and the rest of the route are included. For most uses such as browsing and streaming that is not noticeable, but for some financial trading and real-time applications, even small differences in route length can matter.
Who owns them
Historically, cables were often built by consortia of telecommunications carriers sharing the cost. In recent years, large technology and content companies have invested in or built cables to carry their own traffic, and some now own a significant share of new capacity. The exact balance changes as new projects are announced. Ownership matters because it affects who controls capacity and how access is priced, and it raises policy questions about security and competition.
What goes wrong
Contrary to dramatic stories, most faults are accidental. Common causes include:
- Fishing activity, such as trawling gear dragged across the seabed in shallow water.
- Ship anchors that are dropped or dragged in the wrong place.
- Natural events, including earthquakes and underwater landslides, which can break several cables in one region at once.
- Equipment failure or aging components.
Cable breaks happen several times a year somewhere in the world, according to commonly cited industry figures, and they usually do not cause widespread blackouts because traffic is rerouted over other cables. Regions served by few cables are more exposed; when a small number of routes serve a country or island, a single break can badly slow or interrupt service for days. The geography of cables is therefore a real factor in national resilience.
Intentional damage and spying are also discussed by security analysts and governments. Public claims in individual incidents are sometimes hard to verify, so it is wise to read reports carefully. Cables are hard to defend because they stretch across vast stretches of remote ocean.
Repairing a break
Specialist repair ships are stationed in ports around the world. When a fault is detected, test signals help locate it with reasonable precision. The ship goes to the site, grapples the cable from the seabed, lifts the damaged part on board, splices in a new section and lowers it back. Weather, permits and the depth of the break all affect how long this takes, and repairs can take days to weeks. The work is slow, specialized and relatively unglamorous, yet essential.
Why this matters for everyday internet use
When you load a page hosted on another continent, your request commonly passes through a landing station and under an ocean. Companies reduce reliance on distant connections by caching popular content closer to users, which makes the system feel faster and lessens pressure on long routes. Still, global services like messaging, video calls and cloud computing services rely on the same few thousand kilometers of glass, and the broader story of how it joins up with the rest of the network is part of the picture of internet infrastructure. Increasing demand from video and AI is pushing operators to build new routes and upgrade older ones, but new cables take years to plan, fund and lay.