Africa Coast to Europe (ACE)

Africa Coast to Europe (ACE) is a 17,000 km submarine cable that runs along the entire Atlantic coast of West Africa and ends in France. It lands at 19 stations across 19 countries, which makes it one of the most widely-landed submarine cables in the world. The sequence, from north to south: Penmarch in France, Granadilla de Abona in the Spanish Canary Islands, Nouakchott in Mauritania, Dakar in Senegal, Banjul in the Gambia, Suro in Guinea-Bissau, Conakry in Guinea, Freetown in Sierra Leone, Monrovia in Liberia, Abidjan in Côte d'Ivoire, Accra in Ghana, Cotonou in Benin, Lagos in Nigeria, São Tomé in São Tomé and Príncipe, Bata in Equatorial Guinea, Libreville in Gabon, Duynefontein in South Africa, and Carcavelos in Portugal. Ready for service in 2012, ACE is operated by a consortium of 20 telecom operators, one per landing country plus lead partner Orange.

Every Atlantic-coast African country except Morocco has an ACE landing. That is the cable's distinguishing feature — it is the first submarine cable that gave every West African seaboard nation a direct fibre connection to Europe rather than transit through a neighbour.

174 ms France to South Africa

Our monitor measures ACE between its two terminal landings — Penmarch in Brittany, France, and Duynefontein on South Africa's Western Cape. Over 30 days we collected 44 samples, mostly in the reverse direction:

Light in submarine fibre has a theoretical round-trip minimum of 166.3 ms for a 17,000 km path. We measure 173.9 ms at our fastest sample — that is 1.045× the physics floor. For a cable that winds along the entire West African coast, handshaking at 18 landing stations between the endpoints, this is a remarkable near-theoretical result.

The 10 IP hops on the newer-target path are also clean for a trans-continental route. By comparison, WACS on a similar Portugal-to-South Africa journey measures 13 hops and 210 ms — 40 ms slower than ACE on similar architecture. Direct peering between the respective operators at each end of ACE makes the Penmarch-to-Duynefontein segment one of the tighter Europe-Africa paths we have documented.

19 landings, one-country-one-partner

The consortium structure is what makes ACE architecturally interesting. Each African landing country has a participating telecom that owns a fibre pair share in the cable, with Orange as the consortium lead and principal European partner. A packet from a Beninese ISP heading to Europe travels on capacity that is partly owned by Benin's national telecom, not leased from a European wholesale seller.

For small West African nations, ACE was often the first direct submarine cable they had ever touched. Before 2012, several of these countries — Gambia, Guinea-Bissau, Liberia, Sierra Leone, Equatorial Guinea, São Tomé — relied entirely on satellite or on terrestrial fibre into neighbouring countries with their own cable landings. ACE changed this overnight: each country got its own landing, its own capacity, and its own direct route to Europe.

Before and after ACE

Consider Freetown, Sierra Leone. In 2011, Sierra Leone's entire international internet capacity was approximately 155 Mbps, carried by satellite. Business applications requiring sub-500 ms latency were impractical; video conferencing was largely impossible; even basic web browsing was slow and expensive. One year later, ACE's Freetown landing delivered multi-gigabit capacity to Sierra Leone at European-comparable latency.

The same story repeated at Monrovia, Conakry, Bissau, Banjul, Nouakchott, and Malabo. ACE was not just another cable for these countries — it was the first modern fibre connection they had. Consumer internet pricing in these markets fell by 50–90% within two years of commissioning, and measurable latency improved by factors of 5–10× over what satellite could provide.

This is the pattern that distinguishes development-oriented consortium cables from commercial hyperscaler cables. Hyperscalers build where their own traffic volumes justify investment. Development consortia like ACE include landings that have small absolute demand but significant relative impact. The consortium's shared funding structure is what makes these landings economically viable.

Why so many landings matter technically

ACE's 19 landings cost physics-floor latency. A packet between Penmarch and Duynefontein, if routed along the straightest possible fibre path, would take approximately 130 ms round-trip (for the great-circle distance of ~9,300 km). ACE's actual fibre route is 17,000 km — 80% longer than the great-circle — because it hugs the West African coast rather than cutting across open Atlantic.

That cost buys 17 intermediate landing stations that would not otherwise exist. The trade-off is explicit in the design: every additional landing adds ~800 km of coastal cable and roughly 4 ms of latency, while providing terabit-scale capacity to one more national telecom. Over 17 additional stations, the total cost to latency is ~60 ms, which matches what our measurements show (173 ms measured vs ~130 ms theoretical for straight-line Penmarch-Duynefontein).

For most practical uses — web browsing, video streaming, file transfer — the 60 ms overhead is invisible. What matters is that every country along the route has capacity, not that the end-to-end path is maximally fast.

Four fibre pairs, 5.12 Tbps, still growing

ACE was commissioned with four fibre pairs and a total design capacity of 5.12 Tbps. Upgrades over the cable's operational life have raised usable capacity substantially through coherent-transponder refresh cycles. The cable has been in service for 14 years in 2026 — more than half of its 25-year nominal design life — and continues operating as the primary West African trunk while the newer Equiano and the forthcoming Africa-1 add capacity in parallel.

ACE's relationship with its successor cables is additive rather than substitutive. The economic life of a submarine cable extends well past the arrival of next-generation systems — older cables remain useful for backup traffic, regional peering, and markets where the newer cables do not land. West African telecoms often buy capacity on multiple cables simultaneously; ACE is the one that touches every coastal country.

What our data proves

• ACE delivers France → South Africa at 173.9 ms on the cleanest path, 1.045× the physics floor. Near-theoretical performance across 17,000 km of coastal fibre touching 19 stations.
• 10 IP hops on a trans-continental path is remarkable for a multi-landing cable. Direct peering between Orange (French endpoint) and the South African consortium member makes the Penmarch-Duynefontein segment unusually efficient.
• 14 years into a 25-year design life, ACE is still doing its job. Latency is consistent with the 2012 specifications; the cable shows no signs of repeater-chain degradation or capacity distress.

ACE is the kind of submarine cable that infrastructure journalists tend to overlook — it is not hyperscaler-backed, not hyper-high-capacity, and not geopolitically novel. But it is the cable that brought modern internet to the entire West African coast in a single deployment. Our 2026 measurements show it still carrying that traffic at close to the physics limit.

Try it yourself

Live data on the ACE cable page. For context on West African cables, see WACS (2012 consortium, 14 landings), Equiano (2022 Google, 6 landings, point-to-point), MTWA (2021 operator-owned, 6 landings), and the upcoming Africa-1 (2026 African-led consortium).