Lightera Nearly Triples Its Ocean Fibre

Submarine optical fibre is typically bought about a year before the cable that carries it is laid. Lightera says so in its own announcement, and that single piece of procurement arithmetic is what gives the announcement its urgency. On 21 September, from Málaga during ECOC week, the company said it would nearly triple its submarine optical fibre manufacturing capacity by 2029. Work back one year and the orders that fill that capacity are placed in 2028, which means the systems they serve are being planned now.

Lightera announced a major capital investment to expand its submarine optical fibre manufacturing operations in Denmark and the United States. The company named three facilities as its global manufacturing centres for submarine fibre — Brøndby in Denmark, Norcross in Georgia, and Sturbridge in Massachusetts — but did not say which of them receive the money, or in what proportion. Nor did it disclose the size of the investment. The usable figures in the release are the capacity multiple and the date: nearly triple, by 2029.

The name is newer than the factories. Lightera has existed as a brand only since 1 April 2025, when Furukawa Electric unified its Fiber Cable Division in Japan, OFS Fitel, LLC and Furukawa Electric LATAM S.A. under one identity, with operational headquarters in Norcross. Sturbridge came in with the OFS business; Brøndby operates today as Lightera Denmark ApS, listed by the company among its specialty solutions sites. This is an incumbent adding furnaces, not a new supplier arriving.

A transoceanic cable is a power-limited system. The repeaters that amplify the signal roughly every 50 to 100 kilometres are fed electrically from shore, across thousands of kilometres of conductor inside the cable itself, and the total power a system can deliver is fixed by the voltage that conductor will tolerate. Everything else follows from that budget. The less light a fibre loses per kilometre, the less amplification it needs. The larger the fibre's effective area, the more optical power can be launched into it before nonlinear distortion starts corrupting the signal rather than strengthening it.

This is why submarine fibre is a different product from terrestrial fibre rather than the same product at a premium. Lightera's ocean portfolio — the TeraWave SCUBA family in 80, 110, 125 and 150 variants, plus AllWave ZWP Ocean Fiber, AllWave ULL Ocean Single-mode Fiber and the TrueWave SRS, XL and HD ocean grades — is built around ultra-low attenuation and large effective area. The company's own milestones put those areas at 150 µm² for the first ultra-large-area ocean fibre and 153 µm² for qualified ultra-large-area fibre, against something closer to 80 µm² for standard terrestrial single-mode. Nearly doubling the effective area lets you raise launch power by a comparable factor before the same nonlinear penalty appears, and that headroom converts directly into either reach or capacity.

Drawing that fibre is not a matter of turning up a dial. Ultra-low-loss ocean fibre needs exceptionally pure preforms, tightly controlled draw conditions and proof-testing at high tension on every kilometre shipped, because a fibre that fails in mid-Atlantic costs a repair ship and a month. Capacity of this kind is measured in furnaces and years, which is why a 2029 target announced in 2026 is not a conservative one.

The economics shifted about a decade ago and the industry has not stopped adjusting. For most of the history of subsea, the goal was to get the maximum number of bits down each fibre pair, using the most spectrally efficient modulation the amplifier chain would support. Space-division multiplexing inverted that. Because the cable's binding limit is electrical power rather than spectrum, it turns out cheaper per bit to run many more fibre pairs at lower power each, accepting less spectral efficiency per pair, than to run a few pairs hard. Where a legacy transoceanic system carried four to eight pairs, modern designs carry twelve, and at the frontier up to sixteen or even twenty-four. Lightera's own ocean fibre history cites a twelve-fibre-pair transoceanic SDM system as a 2020 milestone.

Every extra pair is more fibre. An SDM cable of a given length consumes several times the fibre kilometres of the previous generation of the same length, at the same quality grade, from the same small set of qualified suppliers. Lightera describes its ocean portfolio as optimised for SDM-enabled ocean networks, which is the polite phrasing of a straightforward supply problem: the design change that made subsea capacity cheap for the buyer made fibre volume the binding constraint for the seller.

Layer onto that who is buying. Lightera says its direct relationships with submarine cable manufacturers and hyperscale companies give it early visibility into future demand — which is a fair description of a market in which a large share of new transoceanic capacity is now funded, and increasingly wholly owned, by hyperscale cloud and AI operators rather than telecom consortia. Their planning cycles are shorter and less patient than a carrier's, and a supplier that cannot commit volume three years out does not get designed in.

Hulse is explicit about what the next-generation technology is: expanding capacity today, in her words, while preparing for multicore fibre. The release is unusually specific about timing, putting initial submarine multicore deployments in the 2029 to 2030 window.

Multicore fibre puts several independent cores inside a single cladding — four, in the design Lightera illustrates — so one strand carries what four used to. In a cable where the physical space inside the pressure vessel is finite, that is the next lever after fibre count. Lightera markets multicore today for AI data centres, cloud and hyperscale applications and lists submarine cable systems among its target applications, while citing research and industry demonstrations rather than a shipping subsea product. That is an accurate description of where the technology sits: proven in trials, not yet qualified for a twenty-five-year design life under 8,000 metres of water.

The 2029–2030 date is the most interesting number in the announcement, because it collides with the end of the capacity expansion. Lightera is building conventional ocean fibre capacity that comes online exactly when its multicore product would begin displacing it. Either the demand curve is steep enough to absorb both, or the new capacity is deliberately flexible. The release does not say which, and the absence is conspicuous.

There is no public price on this, so the signal has to be read from the shape of the commitment rather than its size. A fibre maker does not plan to triple capacity on a five-year horizon against a market it expects to soften. And it does not expand submarine fibre production in Denmark and the United States — neither of them a low-cost manufacturing location — unless the qualification burden of moving ocean fibre production somewhere cheaper outweighs the labour saving. Submarine fibre is one of the few remaining photonics product lines where the customer's approval process, not the unit cost, decides where the furnace goes.

For anyone selling into subsea — amplifier pumps, coatings, splicing, measurement — the date to plan against is not 2029. It is 2028, when the fibre for those systems gets ordered, and the specifications that matter will be frozen before that.