How Underwater Internet Cables Work

When you load a webpage hosted on a server in another country, your request likely travels through more ocean than you would think. The internet is not wireless. Despite what the word suggests, approximately 99% of all data crossing oceans moves through physical cables resting on the seafloor, TeleGeography’s submarine cable map documents. These cables stretch over 1.4 million kilometers, enough to circle Earth 35 times. They carry every email, video call, financial transaction, and Netflix stream that crosses an ocean. Without them, the global internet would simply stop working.

The short answer

Underwater internet cables work by transmitting data as pulses of light through thin glass fibers encased in protective layers. These fibers are laid on the ocean floor by specialized ships, following routes carefully chosen to avoid geological hazards. The cables land at beach manholes on coastlines, where they connect to terrestrial networks. A single cable can carry terabits of data per second, making them vastly more efficient than satellites for long-distance traffic.

The full picture

The physical construction

A submarine cable is essentially a sophisticated telephone wire, but built for conditions no telephone wire ever faced. At the core are hair-thin optical fibers, made of ultra-pure glass, that carry data as pulses of light. These fibers are so transparent that if you held a 100-kilometer stretch up to your eye, you could see through it almost like air.

Around this fragile core, engineers layer protection like an onion. First comes a plastic coating. Then comes copper or aluminum for power conduction. Above that, layers of petroleum jelly buffer against water pressure. Finally, a steel armoringWrap protects against fishing nets, anchors, and the occasional shark bite. The final cable is about as thick as a garden hose, Submarine Cable Networks explains.

Laying cable across oceans

The process begins with route planning. Survey ships map the seafloor using sonar, avoiding underwater mountains, earthquake zones, and areas where tectonic plates might shift. The ideal route is flat and stable, which explains why cables often take surprisingly indirect paths.

The actual laying happens from cable ships carrying thousands of kilometers of cable on massive spools. As the ship moves at walking pace, the cable unspools and sinks to the ocean floor. The cable does not get buried in most places; it simply rests on the bottom, held in place by its weight and the pressure of the water above. In shallow waters near coastlines, machines bury the cable into the seabed to protect it from anchors and fishing gear.

Landing the cable requires construction work on the beach. Workers dig a trench from the shoreline out to where the water deepens enough for ships to approach. This connection point, called a beach manhole, is where the submarine cable links to the terrestrial fiber network that distributes data inland.

How the light carries data

Optical fibers work on a simple principle: total internal reflection. Light pulses bounce off the walls of the glass fiber as they travel, losing almost no energy along the way. This allows signals to travel over 100 kilometers before needing amplification.

Modern cables use wavelength-division multiplexing, a technique that sends multiple streams of data simultaneously on different colors of light through the same fiber. Think of it as having many lanes on a highway instead of a single lane. A single fiber pair in a modern cable can carry over 300 wavelengths, each carrying 400 gigabits per second. That adds up to over 100 terabits per second per fiber pair.

Every 60 to 100 kilometers, the signal needs boosting. Underwater repeaters, cylindrical devices about the size of a small car, sit along the cable path. These are the only points requiring power along the entire submarine route, and they are powered from landing stations on each end.

The scale of the network

Over 400 submarine cables currently span the world’s oceans. This infrastructure represents hundreds of billions of dollars in investment, maintained by a consortium of telecom companies, technology giants, and governments. Google alone owns or leases capacity on over 20 cables, The Economist reports.

The Pacific route between the US and Asia carries the most traffic, followed by the transatlantic route between North America and Europe. These routes are deliberately redundant, with multiple cables following different paths. If an earthquake severs one cable, traffic automatically reroutes through others within milliseconds.

Why it matters

When a cable cuts, the consequences are immediate and expensive. In 2023, a ship anchor damaged three cables in the Red Sea, disrupting internet service across multiple countries for days. Repair ships had to navigate active conflict zones to reach the breaks. The incident highlighted how a single physical failure can affect millions of people and billions of dollars in transactions.

This matters because the internet is now critical infrastructure. Every financial trade crossing borders, every business video call between continents, every cloud application running on overseas servers depends on these cables. The underwater network is arguably the most critical, and least visible, piece of the global internet infrastructure.

For businesses, understanding this dependency matters for disaster planning. For policymakers, it shapes decisions about cybersecurity and infrastructure investment. For anyone using the internet, knowing where their data actually travels provides useful context for understanding global connectivity.

Common misconceptions

“The internet is in the cloud.” The cloud is just someone else’s computer, and that computer is connected by physical cables. The cloud does not float; it sits in data centers, connected by fiber running through buildings, underground tunnels, and across ocean floors. Every “cloud” operation involves physical infrastructure.

“Satellites will replace submarine cables.” Starlink and other satellite constellations serve remote areas well, but they cannot match the bandwidth or cost efficiency of cables for high-volume traffic. Satellites have higher latency, limited capacity, and weather vulnerabilities. For foreseeable needs, satellites complement cables rather than replace them.

“Cables are safe from surveillance.” Intelligence agencies have a long history of tapping submarine cables. The NSA’s MUSCULAR program, The Washington Post revealed, accessed data directly from cables through partnerships with telecommunications companies. Cables are physically secure from casual access, but they remain a target for sophisticated state actors.

Key terms

Submarine cable: A fiber optic cable laid on the ocean floor to connect continents, carrying the majority of intercontinental internet traffic.

Repeater: An underwater device placed every 60 to 100 kilometers along a cable that amplifies the optical signal to prevent it from degrading.

Wavelength-division multiplexing: A technique that sends multiple data signals simultaneously on different light wavelengths through a single fiber, dramatically increasing capacity.

Landing station: The facility on each coast where submarine cables connect to terrestrial fiber networks and where power is fed into the system.

Beach manhole: The underground chamber where the submarine cable transitions from water to land, connecting to the terrestrial network infrastructure.