Summary
- Prysmian signed an agreement to supply Molex with optical cables for use inside data centres for up to ten years and up to €5.5bn.
- Molex will make a €550m upfront payment, giving Prysmian a concrete demand commitment as it plans €1.25bn of fibre and cable investment through 2031.
- Prysmian says the expansion will more than double its US fibre capacity and create over 1,000 jobs, including 600 in the United States.
- The €5.5bn agreement ceiling, €550m payment, €1.25bn investment and wider €10bn revenue forecast are different quantities; none should be read as revenue already recognised.
The most useful number in Prysmian’s new agreement with Molex is not €5.5bn. It is €550m.
The larger figure is the maximum stated value of optical cables that may be supplied over a period of up to ten years. The smaller figure is money committed upfront. It gives Prysmian something more tangible than a forecast as the cable maker prepares to spend €1.25bn expanding fibre and optical-cable capacity through 2031.
That distinction matters because the artificial-intelligence infrastructure boom is producing increasingly large announcements with increasingly long delivery schedules. A contract ceiling shows the scale the parties contemplate. An advance payment begins to answer the harder question: who is willing to put capital at risk before the factories and data-centre halls consume the promised volume?
A buyer helps underwrite the factory decision
Prysmian says the agreement will take its products further inside the data centre. The company already sells power connections, long-haul optical systems and submarine telecom cable. Under the Molex arrangement, it will supply optical cables deployed within data-centre facilities, where rising fibre density is being driven by faster links between compute systems.
The €550m advance is equal to one tenth of the agreement’s €5.5bn maximum. That arithmetic is analytically useful, but it is not a disclosed contractual ratio. Prysmian and Molex have not published the payment schedule, refund conditions, security package, minimum purchase obligation or accounting treatment. It would therefore be wrong to call the advance non-refundable revenue or to assume it pays for a fixed share of the investment programme.
What can be said is narrower. Molex has moved part of its commitment ahead of physical delivery. Prysmian still has to build capacity, qualify output, meet schedules and turn an agreement ceiling into shipments that can be recognised as revenue.
The planned industrial response is substantial. Prysmian intends to expand production from glass preforming through finished fibre and optical cables in the United States and Europe. It says US fibre capacity will more than double. The €1.25bn programme is expected to create more than 1,000 jobs worldwide, 600 of them in the United States. Prysmian describes itself as one of only three US manufacturers of fibre and optical cable and the only major domestic producer in Europe.
Four numbers that should not be collapsed
The announcement contains four monetary measures with different meanings.
First, the Molex agreement is worth up to €5.5bn over up to ten years. Both qualifiers remain material. The release does not disclose a guaranteed minimum volume.
Second, Molex is making a €550m upfront payment. That is a financing and commitment signal, but the public documents do not establish when it becomes revenue or how it is protected if volumes or schedules change.
Third, Prysmian plans €1.25bn of capacity investment through 2031. That is capital the supplier expects to deploy, not the cost of the Molex contract alone and not an amount paid entirely by Molex.
Fourth, Prysmian forecasts more than €10bn of incremental cumulative revenue through 2035, compared with a 2025 baseline, from a wider set of agreements and commercial initiatives with hyperscalers and data-centre infrastructure providers. The company says that could include up to €1.1bn of annual revenue from 2031. This wider forecast overlaps the strategic programme around Molex; it must not simply be added to €5.5bn as if the two figures represented separate guaranteed orders.
The qualifiers do not make the announcement immaterial. They identify the work that remains between a commercial framework and an income statement.
The segment has momentum, but the base is smaller than the ambition
Prysmian’s first-quarter results show why management is prepared to invest. Digital Solutions generated adjusted EBITDA of €88m, up from €42m a year earlier, while its adjusted EBITDA margin reached 20.6%, an increase of 7.4 percentage points. The segment is already improving before the proposed capacity increase is complete.
But the same comparison puts the new figures in perspective. A €1.25bn industrial programme and a potential €5.5bn supply relationship are large relative to a segment that produced €88m of adjusted EBITDA in one quarter. Execution will be measured in plant commissioning, yields, customer qualification, delivery reliability and cash conversion—not in the size of the announcement alone.
Molex’s incentive is also practical. Paying in advance can help reserve supply in a market where AI clusters require more optical links at higher density. Yet the buyer accepts its own concentration and technology risk: architectures, link standards and deployment schedules can change over a decade. Neither party has disclosed enough to allocate those risks precisely.
The agreement is therefore best read as a form of shared conviction. Molex places cash before delivery. Prysmian places a larger capital programme behind future demand. The prepayment makes the forecast more credible; the investment makes the promise more demanding. Only subsequent orders, factory output and recognised sales will show how much of the headline ceiling becomes economic fact.