South American Crossing (SAC)

South American Crossing (SAC) is a 20,000 km submarine cable commissioned in 2000 that loops around the east and west coasts of South America, connecting twelve landings across eight countries and one US territory. From north to south along the Atlantic: St. Croix in the US Virgin Islands, Fortaleza, Rio de Janeiro, and Santos in Brazil, Las Toninas in Argentina. Around Cape Horn to the Pacific: Valparaíso in Chile, Lurín in Peru, Buenaventura in Colombia. Then north along the Pacific and through Panama: Fort Amador and Colón (Panama canal both ends), and Puerto Viejo in Venezuela back on the Atlantic. SAC is the sole submarine cable that laps the entire South American continent in one continuous fibre body.

SAC was built by Global Crossing and commissioned in 2000, in the last wave of late-1990s submarine cable construction that came before the dot-com telecom bust. Global Crossing filed for bankruptcy in 2002, and the cable passed through multiple ownership restructurings before ending up with Lumen (formerly CenturyLink). Like other survivors of its era, SAC continues to operate 25 years after commissioning, past the cable's original 25-year design life.

154 ms Panama to Argentina

Our monitor measures SAC between Colón in Panama and Las Toninas in Argentina — two landings on opposite ends of the cable's Atlantic leg. Over 30 days we collected 33 samples:

The forward-direction stability is the striking number: 3 ms spread across 17 samples over 30 days. Light through fibre has a round-trip minimum of 195.6 ms for a 20,000 km path; we measure 154.1 ms — below the full-cable floor, meaning the packet uses about 15,800 km of fibre rather than traversing the complete 20,000 km route. That makes sense: Panama to Argentina does not require traversing SAC's Pacific-coast segment. The direct path along Brazil's Atlantic coast is used.

The reverse direction is noisier. 187 ms average, with a single outlier spike to 316 ms. Hop count of 10 on the reverse is fewer than the forward's 17, which implies different routing on each direction — Argentine traffic reaching Panama via a shorter, more heavily-peered path, while Panama-to-Argentina takes the SAC-direct route.

Twelve landings around a continent

SAC's architecture is ambitious: a single cable body that connects every major South American country except the landlocked ones. In 2000, this was a novel thing to build. Previous South American submarine capacity came mainly from shorter point-to-point cables focused on Brazil-US traffic, leaving Argentina, Chile, Peru, and Colombia dependent on transit through Brazil or North American hubs.

The Panama landings deserve mention. SAC has two landings in Panama — Fort Amador on the Pacific coast and Colón on the Atlantic, connected by a terrestrial backhaul crossing the isthmus. This effectively treats Panama as a node where Atlantic and Pacific segments of SAC connect via land rather than through South America's southern tip. The same trick is used by several transoceanic cables to avoid the much longer (and colder, deeper) route around Cape Horn.

The Panama isthmus workaround

The terrestrial crossing at Panama is more than a routing convenience — it is the reason SAC can function as both an Atlantic and a Pacific cable. Going around Cape Horn would have added approximately 8,000 km to the cable route and required construction through some of the world's harshest submarine environments: deep water, strong currents, and ice exposure. The Panama workaround trades roughly 80 km of terrestrial cable crossing through the Panama isthmus for not having to traverse the Drake Passage. That is the same trick used by the Panama Canal for shipping; SAC applies it to fibre.

The practical result is that SAC behaves as two connected cables that share a continental termination. A packet from St. Croix to Valparaíso does not sail around Cape Horn — it enters SAC's Atlantic leg, lands in Colón, crosses Panama by land, re-enters SAC at Fort Amador, and continues down the Pacific. The terrestrial segment is short relative to the full cable length and its contribution to end-to-end latency is modest.

Survivor cable economics

By 2026 SAC is past its 25-year nominal design life. Like PC-1, FLAG-NAL, and the other cables commissioned around 2000, SAC is in the phase where physical infrastructure continues working despite being written off capital-wise. The cable has been through multiple electronics refresh cycles; its original 4.5 Tbps design capacity has been raised substantially through coherent-modulation upgrades. The glass itself has not aged.

What these 25-year-old cables collectively prove is that submarine fibre is a durable investment. Cables commissioned during the internet boom of 1998-2001 are still delivering physics-floor performance in 2026. Whatever the business and ownership turbulence, the physical cable body — steel armour, polymer insulation, sealed repeater vessels, glass strands — endures.

What our data proves

• SAC delivers Colón → Las Toninas at 154.1 ms, below the full-cable physics floor. The path uses about 15,800 km of fibre, not the entire 20,000 km system.
• Forward-direction standard deviation is under 1 ms. Twenty-five years of operation have not degraded the cable's baseline stability.
• Twelve-landing continental architecture remains commercially relevant. Smaller South American nations continue using SAC as their primary Atlantic or Pacific trunk; the cable's replacement value is enormous, its replacement cost prohibitive.

SAC is the kind of legacy cable that appears in no one's strategic roadmap but carries a nontrivial fraction of regional internet traffic every day. Our 2026 measurements confirm it still performs at specification. Whenever SAC is eventually retired, whatever replaces it will have to match a 25-year-old cable's integrated continental coverage — a harder design brief than building a single point-to-point trans-Atlantic.

Why legacy cables still matter commercially

A cable commissioned 25 years ago and written off capital-wise is in an odd commercial position. Its operating costs are modest compared to new cables (repeater power, landing station rent, coherent transponder refreshes), its capacity is smaller per pair than current-generation cables, but its ability to reach landings that newer cables do not cover is unique. For SAC, landings like Puerto Viejo (Venezuela), Fort Amador (Pacific Panama), and Buenaventura (Colombian Pacific) are served by few other cables. That uniqueness keeps the cable commercially viable long past its original amortisation window.

Try it yourself

Live data on the SAC cable page. For South American context see AMX-1 (2014 Caribbean-Atlantic coastal), ellaLink (2021 Brazil-Europe), and SACS (2018 Brazil-Angola). For other 25-year-old survivors see PC-1 and FLAG-NAL.