Medusa Submarine Cable System

Medusa is a 8,760 km submarine cable system that lights up in 2026 and links the Mediterranean rim. It lands in a remarkable eighteen stations across thirteen countries — from Carcavelos and Sesimbra on the Portuguese Atlantic, along the Spanish Mediterranean coast at Barcelona, through Marseille in France and Mazara del Vallo in Sicily, to Algeria, Tunisia, Libya (Bizerte, Benghazi, Misuratah, Tripoli), Malta, Greece, Cyprus, Egypt (Port Said), and Morocco. The cable was designed specifically to bypass the chokepoints of the Red Sea route and to create a Mediterranean backbone that keeps its traffic on European and North African soil.

What makes Medusa notable — and what our data confirms — is its modern architecture. Twenty-four fibre pairs running through a single cable body, with a design capacity of 480 Tbps. For comparison, the 2011 Europe India Gateway along a similar route carries 3.84 Tbps on two pairs. Medusa's total capacity is more than 125 times its decade-older neighbour.

The asymmetry in our data

Our monitor has been measuring Medusa between its Port Said (Egypt) and Carcavelos (Portugal) landings — the cable's full east-to-west traversal. Over 30 days we collected 42 clean samples split across two directions:

A 152-millisecond gap between the two minimums. Not a transient — it has persisted for at least 30 days. The forward direction minimum of 77.8 ms is actually below the physics floor for Medusa's full 8,760 km length (theoretical: 85.7 ms). That is physically impossible if the packet traverses the entire cable. The real path is shorter — approximately 8,000 km — which corresponds to a direct Mediterranean crossing from the Eastern to Western Med without circling through all of Medusa's Libyan, Algerian, and Tunisian landings.

The return direction at 229.9 ms is another 150 ms on top, consistent with a more circuitous route that likely does traverse one or more intermediate landings and peering points. We cannot tell from outside which specific cable (possibly another Mediterranean cable like 2Africa or a pre-existing system) handles the return. What we can tell is that two directions of the same source-destination pair take notably different physical paths.

What 24 fibre pairs buys you

A fibre pair is the minimum unit in which a submarine cable can sell capacity — two strands, one for each direction of data. Two decades ago, cables went to sea with two to four pairs. Modern cables are pushing twenty-four. Why?

Fibre is cheap. Once you have already built and laid the physical cable body, added power feed equipment on each end, and placed submarine repeater amplifiers every ~80 km, the incremental cost of packing more fibres into the same bundle is small compared to the value of additional wavelengths. Medusa's twenty-four pairs mean twenty-four completely independent optical paths between any two landings — each capable of being lit with a set of coherent transponders carrying 200 or 400 Gbps per wavelength across roughly 80 wavelengths. The arithmetic: 24 pairs × 80 wavelengths × 200 Gbps = 384 Tbps, near the cable's rated 480 Tbps design.

Hyperscaler consortium cables — where Google, Meta, Amazon, and Microsoft own capacity — tend to justify fibre-pair counts in the teens. Carrier-consortium cables for national telecom operators have been quicker to adopt the highest counts. Medusa is the latter: its co-owners include operators from each landing country, who each take ownership of one or more dedicated pairs. The remaining capacity is sold into the wholesale market.

Eighteen landings, thirteen countries

The shape of this list matters. Medusa is not a point-to-point cable. It is a Mediterranean backbone, designed so any two landings can communicate over its fibre without going through a third country. Previously, most inter-Mediterranean traffic hopped through one or two large hub cities (Marseille, Palermo, Marseille again) on other cables. Medusa collapses this: a packet from Algiers to Athens can ride Medusa directly, without transit routing through a French or Italian peering point.

There is a strategic context too. The Red Sea route — the traditional Europe-to-Asia submarine path through Suez — has seen repeated cable faults in recent years, with a combination of anchor strikes and geopolitical tensions affecting traffic. Medusa is a pure Mediterranean cable, not a Europe-Asia system; it does not solve the Red Sea problem directly. But by strengthening the internal connectivity of the Mediterranean, it reduces the pressure on long-haul cables for regional hops and keeps North African and European traffic on a single shared fabric.

Libya, three landings

Medusa is unusual in giving Libya three separate landings: Tripoli in the west, Misuratah in the centre, and Benghazi in the east. Most cable systems that touch Libya land at one point only. The three-landing design is a response to practical realities: Libya has had no stable national telecom backbone for over a decade, and connectivity between its western and eastern regions is unreliable by land. Three separate cable landings, on the same cable body, provides redundancy and direct intra-Libyan connectivity that the terrestrial network cannot guarantee.

Whether Libyan carriers exploit this architecture in the near term is a separate question. The cable provides the capability; taking advantage of it requires stable commercial arrangements at each landing, which in turn depend on regulatory and political conditions. For now, the cable is lit and available; the market will follow.

What our data proves

Three concrete observations from 42 measurements:

• The cable works. Port Said → Carcavelos at 77.8 ms is near the physics floor for a direct Mediterranean crossing. Medusa is delivering fast, stable latency along its primary backbone route.
• Return path is different. 229.9 ms on the reverse corresponds to roughly 23,500 km of fibre — clearly using a different, longer route than Medusa itself, consistent with carriers preferring a specific direction for commercial reasons.
• Modern capacity unlocked. 24 fibre pairs × 480 Tbps rated design is a step change in Mediterranean connectivity. The value will manifest over years as carriers migrate traffic onto Medusa and away from legacy systems like EIG's Mediterranean leg.

Medusa is the opposite of a hyperscaler cable. Its sponsors are telecom operators, not cloud giants. Its mission is to densify regional connectivity, not to sprint across an ocean. But in the Mediterranean, where a dozen countries ring a sea that has been a commercial crossroads for three thousand years, that mission might be exactly right. Our 77.8 ms measurement is the first reliable latency number this cable will produce. Years of infrastructure on top of it will use that number as the baseline.

Try it yourself

Live data on the Medusa cable page. For context on Mediterranean and Europe-Asia cables, see EIG (2011 consortium trunk, 2 pairs) and Nuvem (2026 Google Atlantic cable with four landings). Our open measurement data refreshes every two hours.