Meta files for 497 Tbps U.S.-Denmark subsea cable Aurora

Meta Platforms has filed for another subsea cable, this one named Aurora and running 7,268 kilometers between Manasquan, New Jersey, and Blaabjerg, Denmark. Edge Cable Holdings USA, a Meta entity, submitted the application to the FCC on Aug. 17. Data Center Dynamics reported the filing; Communications Daily was first with the name. Meta wants a landing license by November 2026, a prerequisite for the target launch date of Jan. 15, 2028.

Aurora will carry 24 fiber pairs. Each pair is designed for 20.7 Tbps. The design comes to roughly 497 Tbps in total. Those are design figures; the application says Meta has not decided how much capacity to light at launch. Both landing stations already exist: NJFX owns the New Jersey site, Arelion the Denmark site. That takes some land-side construction off the schedule.

The filing argues the route needs supply. Aurora, it says, will provide significant new capacity on routes where demand increases substantially each year. It would further strengthen the resilience of Edge USA's subsea connectivity. The application also leaves Edge USA the option to sell dark fiber into the secondary market. That option turns an internal network upgrade into something closer to a wholesale infrastructure business. How much capacity Meta keeps for itself is left open.

The Atlantic is already crowded with cable. The application's own roster lists MAREA, Grace Hopper, Dunant, Amitié, Apollo, AEC-1, EXA North and South, Atlantic Crossing-1, FLAG Atlantic-1, TGN Atlantic, and Yellow as existing or planned, with Nuvem, Fastnet, and Sol on the drawing board. Aurora joins a busy list.

Meta is no stranger to this business. In February 2025 it filed for Project Waterworth, a proposed 50,000-kilometer cable across five continents. It is also behind 2Africa, the 45,000-kilometer system whose partners include Vodafone, Orange, Telecom Egypt, and China Mobile International. That cable touches 33 countries. It has 46 landing points. Aurora is a smaller project aimed at a single, heavily used route.

Aurora is filed solely in the name of Edge Cable Holdings USA, while 2Africa is partially funded by a group of carriers and investors. That single-sponsor structure gives Meta direct control over deployment and, more importantly, over how capacity is sold. The dark-fiber clause is an option, with no commitment attached.

Meta's primary interest in Aurora is the same as its interest in Waterworth and 2Africa: cheap, reliable capacity for its own data centers. But the system carries no commitment on lit capacity, has an open architecture, and includes a dark-fiber sale option. That makes it a commercial asset as well. The 497 Tbps ceiling is unlikely to be lit on day one; the application's silence on that point suggests demand will determine the configuration. If Meta does not need all of it, it can sell pairs on the secondary market. That turns a capital project into a potential revenue stream and changes the competitive math for every other owner on the route.

The filing's reference to 'facilities-based competition' suggests Meta wants Aurora to be a credible alternative for third-party traffic, not just private plumbing. That reinforces the capital hierarchy this publication tracks: hyperscaler-anchored assets get infrastructure pricing, and Aurora extends the logic by adding a wholesale price tag to the asset itself. Meta is both owner and seller on the same cable.

The schedule is the real constraint. Meta wants the FCC to act by November 2026, roughly 15 months before launch. Any slip pushes Aurora toward an already crowded 2028 deployment window. The existing landing stations shorten construction, but regulatory timing is not entirely in Meta's hands. The FCC's calendar sets the schedule.