Eastern Africa Submarine System (EASSy)

The Eastern Africa Submarine System (EASSy) is a 10,000 km undersea fibre cable that runs down the entire eastern coast of Africa, connecting nine countries with direct landings and, through landside backhaul, serving at least twelve more inland. The landing sequence reads like a map of the Indian Ocean rim: Port Sudan in Sudan, Haramous in Djibouti, Mogadishu in Somalia, Mombasa in Kenya, Dar es Salaam in Tanzania, Moroni in the Comoros, Toliara in Madagascar, Maputo in Mozambique, and finally Mtunzini in South Africa. Ready for service in 2010, EASSy was the first submarine cable in history to directly connect the entire eastern African seaboard — before 2010, most East African countries reached the global internet only through satellite or via lengthy detours through West Africa and the Mediterranean.

EASSy is a consortium-owned cable. Sixteen telecom operators and several development-finance institutions co-funded construction. Its design capacity was 3.84 Tbps at commissioning, upgraded over the years to 36 Tbps — a nearly tenfold increase achieved without laying any new fibre, just by replacing the transponders at each landing.

Two directions, two different internets

Our monitor measures EASSy between the two cable endpoints: Mtunzini in South Africa and Port Sudan in Sudan. Over 30 days we collected 36 measurements, split cleanly into two directions that tell very different stories.

The forward direction measures 64 ms minimum over 7 IP hops. The reverse direction measures 237 ms minimum over 10–16 hops. That is a 173 ms asymmetry — one of the largest we have documented on any cable, comparable to the 152 ms asymmetry we measured on Medusa and nearly matching the 200 ms gap on JUPITER.

Reading the forward direction

The 64 ms minimum is interesting for what it implies. EASSy's full length is 10,000 km; the theoretical physics floor for a round-trip across the full cable is 97.8 ms. We are measuring 64 ms — that is 0.65× the full-cable floor. Physically impossible if the packet is traversing the entire cable length.

What is actually happening: the forward path uses only part of EASSy. Our Mtunzini-side probe reaches a Sudanese target via roughly 6,500 km of fibre — which is the approximate length of EASSy from Mtunzini up the African coast to Mogadishu plus inland backhaul into Sudan. The packet may exit EASSy at an intermediate landing (Dar es Salaam or Mombasa, both of which have strong inland backhaul to Sudan) and finish the journey via terrestrial fibre rather than continuing through the full submarine path.

Seven hops corroborates this: a trans-continental path that stays on mostly one cable with minimal handoffs. EASSy's forward-direction route is clean, direct, and uses roughly two-thirds of the cable's length.

Reading the reverse direction

The 237 ms reverse minimum is another story. 237 ms round-trip corresponds to roughly 24,200 km of fibre — more than double EASSy's full length. This path does not use EASSy at all. The most likely route:

• Port Sudan → overland to Cairo or Jeddah via North African terrestrial fibre
• Cairo/Jeddah → across the Mediterranean or Red Sea on another cable (FLAG, IMEWE, SEA-ME-WE 5, or similar)
• Europe → down the West African coast on WACS or a similar system
• South Africa via a Cape Town landing and terrestrial fibre to Mtunzini

That is a roughly 24,000 km round-trip across three submarine cables and multiple terrestrial handoffs. It explains the 10–16 hop variation (different return paths on different days) and the high standard deviation (every handoff introduces queueing and routing uncertainty).

Why would the return path go the long way around? East African carriers historically sold their European transit capacity before their Indian Ocean capacity. A Sudanese ISP routing traffic southward to South Africa might have excellent European peering and weaker East-African backbone peering — the cheapest commercial option for Sudan → ZA traffic in 2026 may still be "via Europe," even a decade and a half after EASSy lit.

Nine countries, one backbone

EASSy's nine direct landings plus backhaul into the hinterland turned East Africa from an internet periphery into a connected region. Before 2010, Ethiopia, Uganda, Rwanda, Burundi, and Malawi reached the global internet only through satellite or via VSAT links to South Africa and Nairobi — adding hundreds of milliseconds of latency and tiny capacity. Each of those countries now sits behind EASSy via terrestrial fibre, along with Zimbabwe, Zambia, Botswana, and Lesotho.

The Horn of Africa landings (Port Sudan, Haramous in Djibouti, Mogadishu) were particularly notable. EASSy was the first cable to land in Somalia since the civil war disrupted the country in the early 1990s. Djibouti has since become one of the most-landed cable stations in Africa — a dozen cables now touch at Haramous — but in 2010 EASSy was one of its earliest direct fibre connections.

Consortium economics

EASSy was funded by a mix of commercial telecoms and development institutions. The African Development Bank, the International Finance Corporation (World Bank Group), and Germany's KfW all participated — unusual for a submarine cable, where traditional financing comes from the telecoms themselves. The development-finance participation reflected the cable's strategic importance for regional development, not just commercial return.

The ownership structure has evolved. WIOCC (West Indian Ocean Cable Company) was formed specifically to hold the African operator stakes, combining interests from eight East and Southern African carriers into a single entity with more purchasing power. WIOCC remains the largest owner-operator on EASSy today.

One consequence: EASSy is unusually well-integrated with inland African terrestrial fibre. Operators who own capacity on EASSy also own significant backhaul in their respective countries, making the cable a true "wet plus dry" backbone rather than just a coastal endpoint. Our Mtunzini → Port Sudan measurement at 7 hops reflects this integration — a packet rarely crosses more than two operator boundaries on the forward direction.

Fifteen years in service

EASSy lit in 2010. It has been in production for fifteen years. During that time it has seen multiple capacity upgrades (3.84 → 36 Tbps), one fault on the Port Sudan segment (repaired within two months in 2023), and no architectural changes to the cable body. The hardware at the bottom of the Indian Ocean is the same hardware the consortium laid in 2009–2010.

A 25-year design life is the industry standard for submarine cables. EASSy is past the halfway point. Its repeaters were rated for 25 years of continuous operation — titanium pressure vessels, sealed optical pumps, passive optical components designed to not drift more than a tiny fraction of a dB over decades. Our 2026 measurements show the cable delivering performance on the forward direction consistent with its 2010 design specifications.

What our data proves

• EASSy delivers 64 ms forward-direction latency from South Africa to Sudan. That is below the full-cable physics floor, which means the route uses about two-thirds of EASSy plus terrestrial backhaul — not the full submarine path.
• The reverse direction is 173 ms slower and uses a completely different route. 237 ms and 10–16 hops correspond to a Port Sudan → Europe → West Africa → South Africa detour, roughly 24,000 km of fibre across three cables.
• East African traffic continues to use European transit for some destinations. Fifteen years after EASSy commissioned, commercial peering on the north end of the cable still defaults to Europe for long-haul southbound traffic.

EASSy is the structural backbone that made East African internet possible. The asymmetry we measure is not a flaw in the cable but a mirror of how African internet traffic still flows: one direction uses the direct infrastructure, the other follows legacy commercial paths. Both observations are stable over 30 days — this is the 2026 steady state of East African peering.

Try it yourself

Live data on the EASSy cable page. For context on African submarine cables, see Equiano (West African Atlantic, hyperscaler, 2022) and SACS (first direct Africa-South America cable, 2018).