More than 95% of all intercontinental internet traffic does not travel through space at all. It travels through roughly 600 fiber-optic cables lying directly on the ocean floor, some spanning entire oceans, carrying everything from a video call between Cairo and London to a multi-billion-dollar currency trade executed in milliseconds.
The idea that most of the internet's physical backbone sits at the bottom of the sea, laid and repaired by a small fleet of specialized ships, surprises most people who assume satellites do the heavy lifting. Understanding how these cables actually carry data, why they occasionally break, and who is responsible for fixing them explains a surprisingly fragile layer beneath the world's digital economy.
Why Almost All International Data Travels by Cable, Not Satellite
Fiber-optic cables carry vastly more data at vastly lower cost per bit than any satellite system currently in operation, because a single cable can carry many terabits per second across thousands of kilometers with minimal signal loss when engineered correctly.
Satellites remain useful for reaching remote regions cables cannot economically serve, but for the enormous volume of data moving between major population centers and financial hubs, cables are simply the only technology that can keep pace with global demand for bandwidth.
What a Modern Submarine Cable Actually Contains
A finished submarine cable is remarkably slim, often no thicker than a garden hose in open ocean, built around a core of hair-thin glass fibers, sometimes over a hundred per cable, each capable of carrying multiple wavelengths of light simultaneously.
Surrounding that fragile core are layers of steel wire for tensile strength, copper conductors to carry electrical power to equipment along the route, and a tough polyethylene jacket, with additional armor layers added in shallow coastal waters where fishing trawlers and ship anchors pose a real threat.
How Light Signals Actually Carry Data Across Oceans
Data travels as pulses of laser light shone into the glass fiber, which are then bounced along its length through total internal reflection, a physical property of glass at the right angle that keeps the light contained inside the fiber core for enormous distances.
Modern systems layer dozens of different light wavelengths onto a single fiber simultaneously, a technique called wavelength-division multiplexing, which multiplies the effective capacity of each physical strand many times over without adding a single additional cable.
Why Repeaters Are Needed Every Few Dozen Kilometers
Light signals weaken as they travel, a process called attenuation, so cables include repeater units roughly every 60 to 80 kilometers that amplify the optical signal using erbium-doped fiber amplifiers rather than converting it back into electricity.
These repeaters sit sealed inside pressure-resistant titanium housings on the seafloor for their entire operational life, drawing power sent down the cable's copper conductors from massive power feeding equipment installed at the landing stations on either end.
How Cable Routes Are Actually Planned and Surveyed
Before any cable is laid, specialist survey ships spend months mapping the proposed seafloor route in detail, identifying underwater canyons, fault lines, shipwrecks, and existing cables that must be avoided or crossed at a safe angle.
Route planners also weigh geopolitical considerations, since landing points require agreements with coastal governments, and the shortest geometric path is frequently rejected in favor of a longer route that avoids unstable seabed, protected marine areas, or politically sensitive waters.
How Cables Are Physically Laid on the Seafloor
Purpose-built cable ships feed the cable off enormous rotating drums at a controlled speed matched to the ship's course, using dynamic positioning systems to hold an exact track even in rough seas over routes that can take months to complete.
In shallow water near shore, a plow towed behind the ship buries the cable up to two meters beneath the seabed for protection, while in deep ocean beyond about 1,500 meters the cable is typically left resting directly on the seafloor since the risk from anchors and trawlers there is minimal.
Why Cables Break More Often Than Most People Assume
Global cable networks experience roughly 100 to 200 faults every year, and the leading cause is not the popular assumption of shark bites but ordinary human activity, above all fishing trawlers dragging nets and ships dropping anchor in the wrong place.
Underwater landslides, earthquakes, and volcanic activity account for a smaller but more dramatic share of breaks, sometimes severing several cables simultaneously in a single event, as happened when a 2006 earthquake off Taiwan disrupted regional internet connectivity for weeks.
How a Cable Fault Is Actually Located and Repaired
Operators first pinpoint a fault's approximate location using optical time-domain reflectometry, sending a light pulse down the fiber and precisely timing how long it takes for reflections to return from the break, narrowing the search to within a few kilometers.
A specialized repair ship then sails to that location, uses a grapnel to hook and haul the damaged section to the surface, splices in a new length of cable using rigorous fiber-alignment techniques, and tests transmission quality before lowering the repaired section back down.
Who Actually Owns and Operates These Cables
Ownership has shifted dramatically over the past fifteen years from telecom carrier consortia toward large technology companies, with major cloud and platform operators now owning or co-owning a substantial share of new transoceanic cable capacity to serve their own data centers directly.
Most cables are still built and financed through consortium agreements among several companies who each fund a share of construction in exchange for dedicated fiber pairs, spreading the enormous capital cost of a project that can run into hundreds of millions of dollars.
How Landing Stations Connect Cables to National Networks
Where a submarine cable reaches shore, it terminates at a landing station, a secured facility that houses the power-feeding equipment, optical terminal hardware, and the point where the international cable's traffic is handed off to domestic fiber networks.
Landing stations are chosen partly for coastal geography and partly for political and regulatory stability, which is why a relatively small number of coastal cities worldwide host a disproportionate share of the planet's international cable landings.
Why a Handful of Geographic Chokepoints Matter So Much
Certain narrow waterways, including the Strait of Malacca, the Red Sea approaches to the Suez Canal, and the waters off Singapore and Marseille, host a remarkable concentration of the world's submarine cables because they are the shortest practical routes between major landmasses.
This concentration creates genuine vulnerability, since damage to just a few cables passing through one of these narrow corridors can meaningfully degrade internet connectivity across entire regions, a risk that has drawn increasing attention from security planners in recent years.
How Satellite Internet Compares in Capacity and Cost
Even the newest low-earth-orbit satellite constellations carry only a small fraction of the data capacity of a single modern submarine cable, and the cost per gigabit transmitted by satellite remains far higher than the equivalent cost by fiber.
Satellites genuinely excel at reaching ships, aircraft, and remote regions where laying cable is not economically justified, but they are not currently, and are not expected to become, a realistic substitute for cables on the world's major international data routes.
What Actually Happens When a Major Cable Is Cut
Modern international networks are typically engineered with enough redundancy that a single cable cut reroutes traffic automatically onto other cables within seconds, usually with only a temporary increase in latency rather than a visible outage for most users.
The picture changes considerably for countries connected to the wider internet by only one or two cables, where a single break can cause severe national slowdowns or outages lasting days or weeks until a repair ship becomes available and reaches the site.
How Redundancy Is Deliberately Engineered Into the Network
Network planners deliberately route new cables along different physical paths than existing ones, avoiding the same narrow chokepoints where possible, specifically so that a single natural disaster or human error cannot simultaneously sever every path between two regions.
Major cloud providers and large enterprises further protect themselves by purchasing capacity across multiple independent cable systems and providers, ensuring their own traffic can shift automatically if any single cable, landing station, or provider experiences an outage.
What the Next Generation of Submarine Cables Looks Like
Newer cable systems are being engineered with far higher fiber counts and improved amplifier technology, substantially increasing the total data capacity carried by each individual cable compared to systems built even a decade earlier.
Investment in new transoceanic capacity has accelerated sharply as demand for cloud computing and artificial intelligence workloads grows, with several of the largest projects now funded primarily by the same technology companies that consume the bulk of the resulting bandwidth.
How Cable Projects Are Actually Insured and Financed
Building a single transoceanic cable system can cost several hundred million dollars once survey work, manufacturing, ship time, and permitting are included, so financing is typically structured around long-term capacity purchase agreements that let investors recover costs over fifteen to twenty-five years of operation.
Specialized marine insurers cover the laying and repair operations against weather delays, equipment loss, and third-party damage claims, while the completed cable itself is generally treated as a depreciating infrastructure asset on the owning consortium's balance sheet rather than insured as a single unit.
Because so much capital is tied up before a single byte of revenue-generating traffic flows, project sponsors often presell a portion of the cable's future capacity to anchor customers years before construction even begins, reducing the financial risk of the entire undertaking.
Why Governments Increasingly Treat Cables as Strategic Infrastructure
National governments have grown considerably more attentive to submarine cables in recent years, recognizing that a country's economic activity, financial system, and military communications increasingly depend on infrastructure it does not directly own or control.
Several countries now require security reviews before foreign entities can invest in cable landing stations on their territory, and naval forces in a handful of regions have begun openly discussing cable protection as part of broader maritime security planning.
This strategic attention has also encouraged some governments to directly subsidize new cable construction to specific allied or friendly landing points, treating diversified international connectivity as comparable in importance to energy or transport infrastructure.
How the Small Global Fleet of Cable Ships Actually Operates
Only a few dozen specialized cable-laying and repair ships exist worldwide at any given time, operated by a small number of companies, meaning the global capacity to lay new cable or respond to faults is genuinely limited relative to the size of the network being maintained.
These vessels carry enormous rotating cable tanks below deck, precision winches, remotely operated underwater vehicles for seabed inspection, and accommodation for specialist engineering crews who can remain at sea for extended repair missions far from any port.
Because demand for repair capacity is unpredictable and geographically scattered, ship operators maintain standby agreements covering broad ocean regions, and a single ship's schedule disruption in one part of the world can genuinely delay repairs to an unrelated cable thousands of kilometers away.
How Environmental Reviews Shape Where Cables Can Be Laid
Before a cable route receives final approval, environmental assessments examine potential disturbance to seabed habitats, protected coral formations, and migratory patterns of marine species, since the physical footprint of cable laying, though narrow, can still affect sensitive coastal ecosystems.
Regulators in many coastal jurisdictions now require operators to avoid marine protected areas entirely or to use directional drilling techniques that bring the cable ashore beneath sensitive habitats rather than across them, adding cost and time to projects near ecologically significant coastlines.
Once installed, the completed cables themselves are generally considered to have minimal ongoing environmental impact, and studies examining seabed recovery around older cable routes have found limited long-term disturbance once the initial laying activity is complete.
How Cable Capacity Is Upgraded Without Laying a New Cable
Increasing an existing cable's capacity does not necessarily require laying an entirely new line beside it; operators can often substantially raise effective capacity simply by upgrading the electronic terminal equipment at the landing stations on either end, without touching the physical cable buried under the ocean itself.
Newer signal-processing techniques and more efficient optical modulation formats allow more bits to be extracted from every light wavelength passing through the same aging glass fiber, an upgrade that costs a small fraction of an entirely new cable project.
This incremental approach explains why some of the oldest cables still operating in the global network today retain genuine economic value despite their age, since their effective capacity has kept growing through successive equipment upgrades even as the original glass fiber itself remains unchanged.
Why Some Countries Remain Far More Vulnerable Than Others
A country's genuine internet resilience depends heavily on how many independent submarine cables reach its shores and how many separate landing stations those cables use, since a nation served by a dozen diverse cables can absorb a single fault without most users noticing anything at all.
Island nations and countries at the end of long single-cable routes face a fundamentally different risk profile, where one fault can genuinely sever the bulk of international connectivity for the entire population until a repair ship arrives, sometimes after a delay of days or weeks.
Development banks and regional infrastructure funds have increasingly prioritized financing additional cable diversity for exactly these vulnerable nations, recognizing that a single point of failure in international connectivity carries direct economic consequences for banking, trade, and essential government services.
How Traffic Is Prioritized During a Partial Outage
When a fault temporarily reduces available capacity on a route, network operators generally prioritize latency-sensitive traffic such as financial trading and voice calls over less time-critical bulk transfers like large file backups, which can simply wait a few extra minutes without any user noticing.
This prioritization happens largely automatically through pre-configured routing policies rather than manual intervention, since the volume and speed of internet traffic makes real-time human decision-making about individual data flows completely impractical during an active fault.
Sources
- Wikipedia β overview of submarine cable construction, history, and technology
- International Telecommunication Union β global data on telecommunications infrastructure and cable systems
- TeleGeography Submarine Cable Map β independent tracking of global cable routes and ownership
- International Cable Protection Committee β industry body on cable fault statistics and protection practices
- National Oceanic and Atmospheric Administration β oceanographic data relevant to seafloor cable routing
FAQ
Do sharks really bite undersea cables?
Shark bites have been documented but are a minor cause of faults overall; fishing gear and ship anchors are responsible for the large majority of cable damage worldwide.
How long does it take to repair a broken cable?
Repairs typically take one to several weeks depending on ship availability, weather, water depth, and how far the repair ship must travel to reach the fault location.
Could satellites eventually replace undersea cables?
Not for major international routes in the foreseeable future; cables carry far more data at far lower cost per bit, and satellite capacity is not close to matching that scale.
Who pays to build a new submarine cable?
Costs are usually shared by a consortium of telecom carriers and, increasingly, large technology companies, each funding a portion of construction in exchange for dedicated fiber capacity.
Why do so many cables pass through the same narrow waterways?
Certain straits and coastal corridors offer the shortest practical route between major landmasses, so cable operators concentrate there even though it creates a shared vulnerability.
About the Author
We reference Wikipedia, International Telecommunication Union, TeleGeography Submarine Cable Map, International Cable Protection Committee, and National Oceanic and Atmospheric Administration to explain the background and current understanding of this topic.
Loved This Article?
Share it on WhatsApp β Share it on WhatsApp
Get more guides in your inbox β Subscribe to our newsletter for weekly surprising stories from Egypt, Saudi Arabia, Dubai, and beyond.