EXA Meridian and the return of operator-built transatlantic capacity

On 16 September, EXA Infrastructure announced EXA Meridian, a 6,552-kilometre submarine cable between New Jersey and a new landing station at Brean in the United Kingdom. The system is planned with 24 fibre pairs and more than 500 Tbps of scalable capacity, with a target ready-for-service date in the fourth quarter of 2029 (EXA Infrastructure). The headline is not just another large number in a subsea capacity table. EXA is building the route as a single operator, with a contract in force and anchor-customer commitments, at a moment when hyperscalers, neoclouds and financial institutions are moving from buying individual wavelengths to reserving spectrum and complete fibre pairs.

A cable announcement with a contract behind it

EXA Meridian will be EXA Infrastructure’s ninth transatlantic cable. The company says it is the first transatlantic system in a decade built by an operator rather than a hyperscaler, a claim that Light Reading presents as EXA’s position rather than an independently audited industry ranking. The more concrete fact is the engineering scale: 24 fibre pairs across 6,552 km, connecting the US East Coast directly to Somerset.

The project has passed a more meaningful threshold than a memorandum or a route announcement. EXA says it has reached Contract in Force with Xtera, which will supply the wet-plant technology, and that the investment has significant anchor-customer commitments from neocloud and hyperscale companies (EXA Infrastructure). Light Reading reports that EXA has pre-sold capacity to a broad customer set and has secured enough future business to de-risk the build, while retaining capacity for leases and future sales (Light Reading).

That distinction matters to carriers. A subsea system can be technically credible and still fail to become a service if it lacks financing, manufacturing slots, vessel availability or a sales plan. EXA has not disclosed the project’s investment amount or the identity of its customers. But Contract in Force and pre-sales are stronger evidence of execution than a capacity target alone.

The capacity unit is changing

From wavelengths to spectrum

EXA’s explanation of demand is more important than its marketing maximum. The company says its largest buyers are increasingly procuring spectrum and complete fibre pairs rather than individual wavelengths (TelecomTV). That is a change in the commercial unit of international connectivity. A wavelength is a managed service with a defined line rate and handoff. A spectrum slice or fibre pair gives a large customer more control over optical equipment, upgrade timing and traffic engineering.

The shift is not universal. Most carriers and enterprises will continue to buy managed wavelengths because they do not want to operate subsea line systems. But the largest cloud and AI customers have a different requirement: they need to move very large data sets among distributed compute sites, and they want capacity that can grow without negotiating a new wavelength every time the traffic model changes.

Light Reading quotes EXA CEO Jim Fagan saying that a fibre pair previously filled in three or four years, while recent demand has filled it in less than 24 months. He also says EXA is receiving requests for new builds with more capacity than its existing network by a factor of ten (Light Reading). These are operator observations, not an independent market forecast. They do, however, explain why a 2029 project is being contracted now. The long lead time is being pulled forward by customers that want physical capacity reserved before the next build cycle is sold out.

Capacity is not the same as throughput

The more than 500 Tbps figure is a system design and scaling statement, not a promise that all of that traffic will be lit on day one. It combines fibre-pair count, the optical design and future terminal equipment. The commercial question is how much of the system is committed to anchor customers, how much is available as spectrum or dark fibre, and how quickly the landing stations and terrestrial network can turn that physical resource into a bookable service.

That is why a responsible comparison should not place EXA Meridian’s headline number directly beside a carrier’s currently lit IP capacity. The relevant comparison is the combination of route control, upgrade rights, landing-station access, restoration options and time to provision.

Why single-operator ownership matters

EXA says the new Brean landing station will connect through diverse terrestrial routes to Slough and the company’s wider pan-European backbone (EXA Infrastructure). This is a different proposition from selling capacity into a consortium point and leaving the buyer to assemble backhaul, optical termination and cross-border transport separately.

The benefit is operational coherence. One network owner can align the subsea spectrum plan with terrestrial route engineering, data-centre handoffs and restoration policy. It can also offer customers a wider set of end-to-end combinations rather than one fixed cable landing. Submarine Networks describes EXA’s broader footprint as a 37-country European grid with more than 174,500 route kilometres and over 500 points of presence. Those figures are company-network claims, but the structural point is sound: Meridian’s value depends on the network behind Brean, not only on the wet plant offshore.

This does not make the consortium model obsolete. Consortium cables spread risk and can aggregate demand that no single operator could finance. A single-operator system instead trades shared governance for faster product decisions and tighter control of the customer interface. Its success will depend on whether EXA exposes enough flexibility to wholesale buyers without turning the cable into a closed internal asset.

What it means for the region

EXA Meridian does not create a new landing in the Balkans, and it does not replace the Adriatic, Eastern Mediterranean or Black Sea projects that are intended to diversify Europe’s southern and eastern approaches. Its relevance to Bulgaria is upstream. A high-capacity transatlantic route terminating inside a pan-European terrestrial network changes the supply options available to carriers that need resilient access to London, Frankfurt, Paris or Amsterdam before carrying traffic into Sofia and the wider region.

For a Bulgarian wholesale operator, the procurement questions are practical. Can a customer buy a protected wavelength, a spectrum slice or a full fibre pair? Where is the optical handoff? Is the route diverse from existing Atlantic paths all the way to the regional point of presence, or only across the ocean? What restoration rights apply when a landing station or terrestrial segment fails? And can the service be combined with an Adriatic or Black Sea route without creating a hidden common point of failure?

The lesson from EXA Meridian is therefore not that transatlantic capacity has become abundant. It is that the market is splitting. Consumer and ordinary enterprise traffic will continue to use managed IP transit and wavelengths. The largest AI, cloud and financial customers are reserving deeper layers of the optical stack. Balkan carriers should plan for both markets, and should treat the distinction between a headline Tbps figure and a controllable, restorable route as a commercial specification rather than a footnote.

Sources

• EXA Infrastructure: EXA Meridian announcement
• Light Reading: EXA to lay new transatlantic cable as AI gobbles capacity
• TelecomTV: Transatlantic networks, AI in Europe and EC rules
• Submarine Networks: EXA announces new transatlantic subsea cable Meridian