There are two critical things that every data centre in the world needs to operate successfully: power and connectivity—often referred to as power and fibre.
Connectivity provides the interconnection of information between end users and the compute/storage hosted in the facility. Let’s call that the data.
Power allows the compute/storage hosted in the facility to remain energized and cooled. Let’s call that the centre.
When we design a data centre, we ensure that there is an “N+1 redundancy,” which is resilience achieved by adding additional equipment to the architecture to support a failure or allow equipment to be serviced while the data centre is still operational. In the event of a power grid failure, standby generators are on-site to keep the data centre running without interruption.
The one thing that a data centre operator can’t control is the connectivity once it leaves the campus. This is why Hyperscalers insist on a minimum of three telecom service providers, using a minimum of three physical routes/points of entry to a data centre campus.
International connectivity can come in the form of Terrestrial Cable, Submarine Cable, and Satellite:
- Satellite has bandwidth limitations and suffers from high latency. It’s often used as a backup for critical data or deployed in areas where there is no communications infrastructure. However, this is the most expensive form of international connectivity.
- Terrestrial cables are typically laid alongside gas/oil pipeline corridors, on overhead power lines, or next to railway lines/highways. Border crossings can be more complicated, and cable deployment can be expensive as land negotiations have to take place. However, international terrestrial cables are very reliable and rarely experience outages.
- Submarine cables carry 98% of the world’s internet traffic, spanning 1.5 million kilometres across 560 cable systems. These cables are laid on the seabed between landing stations in different countries. They are cheaper to deploy than terrestrial cables but are prone to damage from underwater earthquakes, dragging anchors, deep-sea trawling, and sabotage from hostile state actors. These outages can take between 5–15 days to repair, depending on the depth or location. While multiple cable systems can provide redundancy in the event of an outage, there are geographical “Choke Points” that can impact internet resilience due to the volume and proximity of cables on the seabed.
Choke Points include:
- The Suez Canal / Red Sea corridor, a prime route for East-West connectivity
- Singapore / Strait of Malacca, the key Southeast Asia hub
- Maltja, Marseille, and Gibraltar, for European and African access
Ironically, we are now more dependent on international fibre connectivity than ever before, but we’ve never been so vulnerable to disruption from something outside our control at a city, country, or regional level.
International fibre connectivity is no longer a utility but the foundation of the digital infrastructure age—enabling societies to thrive, innovate, and connect on an unprecedented scale. This has resulted in increasing access to high-speed internet worldwide, fueling economic growth, boosting employment, enabling innovation, and lowering barriers to trade.
If you wish to learn more about International Fibre Connectivity or view where all these cables are geographically, I recommend the websites below:
