The Hidden Network: How Subsea Cables Power the Global Internet

You’ve probably heard that the internet is “in the cloud.” It’s not. The cloud is underwater, buried in mud, and crawling with sharks.

Over 95% of the world’s intercontinental data travels through a hidden network of fiber-optic cables laid on the ocean floor. These cables are the literal backbone of the global internet. They’re why you can video call someone in Tokyo from New York, stream a movie hosted in Dublin, or send a message that bounces through London in under a second.

But how do you build a data highway across the abyss? It’s one of the most impressive engineering feats of the modern age, and it’s almost invisible.

The Scale of the Subsea Network

There are over 400 active submarine cable systems, stretching roughly 1.3 million kilometers (800,000 miles) in total. That’s enough to wrap around the Earth 30 times. These cables carry everything: emails, financial trades, video calls, streaming content, and even the data that powers your smart home.

The cables aren’t just a few strands of glass. They’re complex, multi-layered structures designed to survive crushing pressure, corrosive saltwater, and the occasional anchor drag.

What’s Inside a Submarine Cable?

A modern submarine cable is about the thickness of a garden hose, but its construction is a marvel of materials science.

• Core: The center contains one or more pairs of optical fibers. Each fiber is a strand of ultra-pure glass, thinner than a human hair, that carries data as pulses of light.
• Buffer tubes: These protect the fibers from moisture and physical stress.
• Strength members: Layers of steel or Kevlar wires provide tensile strength. A cable must withstand the tension of being laid from a ship and the pressure of deep ocean water.
• Copper or aluminum sheath: This conducts electricity to power the repeaters (amplifiers) along the cable.
• Outer insulation: A thick layer of polyethylene protects against saltwater, abrasion, and marine life.

In shallow water (near coastlines), cables are armored with steel wire and sometimes buried under the seabed to protect against fishing trawlers and ship anchors. In the deep ocean, they’re thinner and lighter, relying on the depth to keep them safe.

The Engineering Challenge: Laying Cable Across the Abyss

Laying a cable across the Atlantic or Pacific isn’t like running a wire through your attic. It requires specialized ships, precise navigation, and months of work.

Route Planning

Engineers spend years planning a route. They avoid: - Underwater mountains and trenches – The ocean floor isn’t flat. They map the seabed to find a path that minimizes stress on the cable. - Shipping lanes and fishing grounds – Anchors and trawlers are the biggest threat to cables. Routes are chosen to avoid high-traffic areas, or the cable is buried deep in the seabed. - Earthquake zones – Subsea landslides can snap a cable. Routes are designed to cross fault lines at safe angles.

The Laying Process

A cable-laying ship, like the CS Responder or Ile de Sein, carries thousands of kilometers of cable coiled in massive tanks. The process is slow and precise:

1. Near shore: The cable is buried 1–3 meters deep using a plow towed by the ship. This protects it from anchors and fishing gear.
2. Continental shelf: The cable is laid on the seabed, often with a protective layer of rock or concrete mattresses.
3. Deep ocean: The cable is simply laid on the seafloor. At depths of 8,000 meters, no ship anchor can reach it.

The ship moves at about 6–8 knots, paying out cable at a controlled tension. If the cable is laid too fast, it can kink; too slow, it can snap. The entire operation is monitored by GPS and acoustic sensors.

The Repeaters: Keeping the Signal Alive

Light traveling through glass fiber loses strength over distance. After about 100 kilometers, the signal would be too weak to read. That’s where repeaters come in.

These are electronic amplifiers placed every 50–100 kilometers along the cable. They’re powered by the copper sheath, which carries high-voltage DC electricity from shore stations. A single cable can carry up to 10,000 volts.

Repeaters are the most failure-prone part of the system. They’re sealed in titanium or stainless steel housings to withstand pressure, but they contain lasers, electronics, and power supplies. If one fails, the entire cable can go dark.

How Data Travels: Wavelength Division Multiplexing

Modern cables use a technique called Wavelength Division Multiplexing (WDM) . Instead of sending one signal per fiber, they split the light into dozens of different colors (wavelengths), each carrying its own data stream. A single fiber pair can carry hundreds of terabits per second.

For example, the MAREA cable (connecting Virginia, USA to Bilbao, Spain) uses 8 fiber pairs, each capable of 200 terabits per second. That’s enough bandwidth to stream 80 million HD movies simultaneously.

The Hidden Vulnerabilities

Despite their strength, subsea cables are surprisingly fragile.

Shark Bites

It sounds like a joke, but it’s real. Sharks have been known to bite cables, possibly attracted by the electromagnetic field. Modern cables are wrapped in steel tape to deter them, but it still happens.

Ship Anchors and Fishing Trawlers

This is the #1 cause of cable damage. A trawler dragging a net across the seabed can snag a cable and snap it. In shallow waters, cables are buried, but in deeper areas, they’re exposed. Repairing a broken cable costs millions and takes weeks.

Natural Disasters

Earthquakes and underwater landslides can sever multiple cables at once. In 2006, a 7.0 earthquake near Taiwan snapped seven cables, disrupting internet service across East Asia for weeks.

How Cables Are Repaired

When a cable breaks, the first sign is a sudden drop in data traffic. The cable’s operators use optical time-domain reflectometry (OTDR) to pinpoint the break’s location within a few meters. A repair ship is dispatched, which can take days to reach the site.

The repair process is like deep-sea surgery:

1. The ship drags a grapnel along the seabed to hook the cable.
2. The cable is cut and brought to the surface.
3. A new section is spliced in, using a fusion splicer to join the glass fibers with microscopic precision.
4. The repaired section is lowered back down, often with extra slack to prevent future stress.

A single repair can cost $1–2 million and take weeks. That’s why redundancy is built into the system—multiple cables connect major hubs so that a single break doesn’t take down the internet for a continent.

Who Owns the Cables?

You might think governments own them. In reality, most submarine cables are owned by consortia of tech companies and telecom carriers. Google, Facebook, Amazon, and Microsoft are major investors. They need the bandwidth for their own services—search, cloud computing, video streaming—so they build their own cables.

For example, Google’s Dunant cable (connecting the US to France) is entirely owned by Google. It’s a private highway for their traffic. This gives them control over latency, capacity, and security.

The Security Angle: Tapping the Unthinkable

Can you tap a submarine cable? Technically, yes. But it’s incredibly hard.

To tap a fiber-optic cable, you need to physically access it, strip the layers, and bend the fiber to extract a tiny fraction of the light signal. This requires a deep-sea submersible, precise engineering, and the risk of detection. The cable’s operators monitor signal strength constantly; any drop in power or unusual reflection is a red flag.

That said, intelligence agencies have done it. The US Navy’s USS Jimmy Carter is a modified submarine designed to tap undersea cables. It’s a cat-and-mouse game: cable owners encrypt their data, and spies try to intercept it.

The Power Problem

Cables need electricity to power the repeaters. This is supplied from shore stations, which send high-voltage DC (typically 3,000–10,000 volts) down the copper sheath. The current travels thousands of kilometers, powering repeaters along the way.

If a cable is cut, the power is lost, and the repeaters go dark. That’s why cables are designed with redundant power paths. If one shore station fails, the other can take over.

The Future: More Cables, More Capacity

The demand for bandwidth is growing exponentially. Streaming, cloud computing, AI, and the Internet of Things are all hungry for data. To keep up, new cables are being laid constantly.

Key Trends

• Higher fiber counts: Early cables had 2–4 fiber pairs. Modern cables like 2Africa have 16 pairs, each capable of 200+ Tbps.
• Space-division multiplexing: Instead of one fiber per pair, cables now use multiple fibers in a single sheath, dramatically increasing capacity.
• Direct routes: Cables are being laid to connect emerging markets directly, bypassing traditional hubs. For example, the 2Africa cable will connect 33 countries in Africa, Europe, and Asia.
• Arctic cables: Melting ice is opening new routes. The Far North Fiber project plans to connect Europe to Asia via the Arctic, cutting latency by 30% compared to the southern route.

The Environmental Impact

Laying cables does disturb the seabed, but the impact is surprisingly small. Cables are thin, and the seabed recovers quickly. In fact, cables often become artificial reefs, attracting marine life.

The bigger environmental concern is the power consumption of the shore stations and repeaters. A single cable system can draw several megawatts. But compared to satellite internet (which requires massive ground stations and rocket launches), subsea cables are remarkably energy-efficient.

The Geopolitics of Cables

Cables are strategic assets. Countries want to control where they land, who owns them, and who can access the data.

• Landing rights: Every cable must get permission from the country where it makes landfall. This can be a political bargaining chip.
• Surveillance: Intelligence agencies have been known to tap cables. The NSA’s Tempora program reportedly intercepted data from undersea cables in the UK.
• Sabotage risk: In 2022, the Nord Stream gas pipeline explosions raised fears about critical undersea infrastructure. Cables are vulnerable to similar attacks, though they’re harder to target because they’re smaller and more numerous.

The Human Side: Who Maintains This?

A small, global community of engineers and technicians keeps the network running. They work for companies like SubCom, Alcatel Submarine Networks, and NEC. When a cable breaks, they’re on a plane within hours.

The job is grueling. Repair ships can be at sea for weeks, working in all weather. The work is precise: splicing a fiber-optic cable requires a clean room environment, even on a rolling ship in a storm.

The Future: What’s Next?

The subsea cable industry is booming. By 2025, over $10 billion will have been spent on new cables in the previous five years. Here’s what’s coming:

• Space-division multiplexing: Using multiple cores in a single fiber, or multiple fibers in a single cable, to multiply capacity.
• Repeaterless cables: For short distances (under 400 km), new amplifiers can boost signals without repeaters, reducing cost and failure points.
• Quantum-safe encryption: As quantum computing threatens current encryption, cables are being designed with quantum key distribution (QKD) to secure data.
• Deep-sea mining cables: Some companies are exploring cables that also carry power for underwater mining equipment, though this is controversial.

The Human Cost

The cable industry is a small, tight-knit world. Engineers, surveyors, and ship crews spend months at sea. The work is dangerous: handling high-voltage equipment, working in rough weather, and diving to repair cables in zero visibility.

But there’s also a sense of mission. These cables connect continents. They enable global commerce, education, and communication. Without them, the internet as we know it would collapse.

The Bottom Line

The internet isn’t wireless. It’s a physical network of glass and steel, laid with painstaking precision across the planet’s most hostile environment. Every time you load a webpage, you’re relying on a cable that might be lying in a trench 5,000 meters deep, patrolled by sharks, and maintained by a handful of engineers who spend months at sea.

Next time you complain about buffering, remember: your data just crossed an ocean.