Global Internet Cable Map

How Does the Internet Travel Between Countries?

When you load a page hosted on another continent, the request almost never travels through the air. It crosses the sea inside a fibre-optic cable lying on the ocean floor. Your traffic leaves your home over copper, fibre or radio, reaches your internet provider, and is handed between networks until it reaches one that operates or leases capacity on a submarine cable. On the other side, the same process runs in reverse.

Which cable your traffic actually uses is not fixed. Networks announce routes to one another with the Border Gateway Protocol, and the path is chosen by commercial agreements and current conditions, not by geography alone. Two requests a second apart can take different routes. That is why this map shows where the cables are, and never claims to show where your packets go.

What Are Submarine Internet Cables?

A submarine communications cable is a bundle of optical fibres, each thinner than a human hair, wrapped in layers of protection: a copper conductor that carries power to the repeaters, a steel wire sheath, polyethylene insulation, and, near shore where anchors and trawlers are a real hazard, heavy armouring. In deep water a cable can be no thicker than a garden hose.

Light travels down the fibres carrying data. Because the signal weakens over distance, optical repeaters are spliced in every 50 to 100 kilometres and powered by a constant current sent along the copper conductor from the shore. A single modern cable system can carry hundreds of terabits per second.

How Are Undersea Fiber Cables Installed?

A cable route is surveyed first, because the seabed is not flat: the survey looks for canyons, volcanic activity, fishing grounds and existing cables and pipelines. Permits are then required from every country whose waters the cable crosses.

A cable ship loads the cable into large circular tanks and pays it out as it steams along the route. In shallow water, roughly the first 1,500 metres of depth where damage from anchors and fishing gear is most likely, a sea plough buries the cable a metre or more beneath the seabed. In deeper water it is simply laid on the bottom. Ashore, the cable is pulled through a duct into a landing station, where the fibres are connected to terrestrial networks.

What Is a Cable Landing Station?

A cable landing station is the building where a submarine cable comes ashore and meets the terrestrial network. It houses the power feed equipment that energises the repeaters out at sea, the terminal equipment that turns light into data, and the cross-connects to backhaul networks running inland.

Landing stations are deliberately unremarkable buildings, and several cables often share one. This dataset contains 209 landing points, of which 13 are buildings explicitly tagged as cable landing stations in OpenStreetMap; the rest are the points where a mapped cable route reaches the coast. Browse the landing station directory.

How Much of the Internet Uses Submarine Cables?

The figure most often quoted is that submarine cables carry roughly 99% of international data traffic. It comes from the research firm TeleGeography, which publishes it in its submarine cable FAQ, and it is repeated by the International Telecommunication Union and the International Cable Protection Committee.

Two things are worth being precise about. The figure describes international traffic, not all internet traffic: most data never leaves the country it started in. And this map measures nothing about traffic at all: it maps physical infrastructure. No traffic volumes, capacities or utilisation figures are shown anywhere on this site.

What Happens When a Submarine Cable Breaks?

Cable faults are routine. Most are caused by fishing gear and ships' anchors in shallow water, with earthquakes, undersea landslides and abrasion accounting for most of the rest. The industry handles well over a hundred faults a year.

Usually nobody notices, because networks are built with redundancy: traffic reroutes onto other cables within seconds, and a busy corridor such as the North Atlantic has many parallel systems. Where a country is served by only one or two cables, a break is far more serious and can slow or cut international connectivity for weeks.

Repair means sending a cable ship, grappling the cable off the seabed, lifting the damaged section to the surface, splicing in new cable and lowering it back. Depending on weather, depth and how far the nearest repair ship is, that takes days to months.

Who Owns Submarine Internet Cables?

Traditionally, submarine cables were built by consortia of telecommunications carriers who shared the cost and the capacity. That model still builds many systems, especially those serving several countries along one route.

Since the 2010s, large content providers have become major owners in their own right, financing private cables to link their data centres. Ownership is often layered: one group owns the cable, another operates it, and capacity is sold or swapped between many carriers.

Where ownership is recorded in Wikidata this site shows it, for 14 of the 168 systems in the dataset. For the rest the owner field reads Data unavailable, because guessing would be worse than saying nothing.

How Long Are Submarine Cables?

They range from a few kilometres to well over 20,000. A cable crossing a strait between two islands may be shorter than a city street network, while a trans-Pacific system runs further than the distance from the North Pole to the South Pole.

Two different lengths appear on this site and they are not interchangeable. Where Wikidata states an official system length, that number is used. Otherwise the length shown is measured from the route actually mapped in OpenStreetMap and marked with an asterisk. For a partly mapped system that can be a small fraction of the real length. The distinction is labelled on every cable page.