Fifteen years to a product: reading Aledia’s Champagnier fab against its own contradictions

I. The mechanism: a display company that had to become a semiconductor company

The usual way to read Aledia is as a display start-up. That reading misses what makes it structurally interesting to anyone working in the Grenoble basin. The company's bet was never mainly optical; it was dimensional. Conventional LED epitaxy runs on small, expensive sapphire or specialty wafers. Aledia's founding claim, carried out of CEA-Leti's laboratories in Grenoble after roughly six years of research before incorporation in 2011, is that three-dimensional GaN nanowires can be grown on standard silicon — and therefore on the wafer diameter the silicon industry already amortises.

That is why the single most consequential document in this file is not a financing release but an equipment order. In a release dated 20 October 2020, Veeco Instruments announced that Aledia had selected its Propel platform — described as the first 300 mm MOCVD system with an EFEM designed for advanced display applications — for microLED production. Aledia's own site carries the same release under the date 21 December 2020; the two-month gap between the two datelines on one announcement is the first of many date discrepancies in this record, and we leave it standing.

Read as a capability statement, the 300 mm choice tells a reader more than any roadmap slide. It means the company's process engineers work in the tooling conventions of mainstream silicon fabs, not of the LED industry — the same conventions used at Crolles and Bernin, twenty minutes up the valley. That is a labour-market fact before it is a technology fact, and Section VI of the companion skills dossier turns it into job families.

II. Champagnier: one site, three cost figures, four start dates

The site sits in the ZAC du Saut du Moine at Champagnier, which is also the company's registered address. Local reporting describes a phased build of three buildings planned through 2025, the first delivered at the end of its reporting year; the regional investment agency describes an announced 52,000 m² production site. Those two descriptions are compatible. The money and the calendar are not.

Three numbers, two currencies, no bridge between them. €140 million and $200 million are not the same claim restated: they differ in amount, in currency and in scope, and neither publisher states what the other counted. A dossier that averaged them, or converted one into the other, would be manufacturing a figure. We print all three and treat the fab's total cost as unpublished.

The calendar is worse, and more revealing. Four dated positions exist on when Champagnier started producing:

• 2022 — the plan-stage trade report, which said the line would start making microdisplays by then.
• "courant 2023" — the headline of regional business coverage of the first building's delivery, whose own body text describes three buildings planned through 2025.
• October 2024 — national business daily Les Echos, 7 October 2024, reporting that Aledia was completing the start-up phase of its production site.
• 2025–2026 — the company's own retrospective, republished on its site on 11 July 2025, describing more than ten years spent developing a proprietary microLED technology "without producing a concrete product or generating revenue", followed by the commercial launch of its first FlexiNova product announced on 26 January 2026.

The mechanism behind that four-year spread is not incompetence; it is the difference between four different events that press coverage compresses into one word. First silicon in a tool, first building handed over, qualified line, and first revenue-generating product are separated by years in any semiconductor ramp. The lesson for a reader assessing any deeptech factory announcement is precise: "production starts in 20XX" is not a date until the publisher says which of those four events it means. None of the four sources here says.

III. The €120 million that appears twice

Aledia's financing history is documented, dated, and internally ambiguous at exactly one point.

The first round is clean: €10 million, second tranche closed in March 2013, reported by Les Echos Capital Finance on 29 March 2013, which describes the company as driven by Giorgio Anania with associates Xavier Hugon and Philippe Gilet. (Reference material elsewhere names only Gilet and Hugon as founders. We report both accounts; we do not adjudicate a founder count.)

The second is where the ambiguity begins. On 7 October 2020, Aledia announced the closing of €80 million — explicitly the first of three tranches of a planned €120 million Series D — with Bpifrance's SPI fund among the investors; Bpifrance published its own account on 9 October 2020, and French coverage the same day framed it as "€80 million raised, €40 million still sought". Aledia's own site again carries this announcement under a later date, 19 December 2020, over body text datelined 7 October 2020.

Then, on 23 September 2023 (body dateline 26 September 2023), Aledia announced a raise of €120 million from its historic shareholders — CEA Investissement, Supernova Invest, Braemar Energy Ventures, Bpifrance — and from the French State via the French Tech Souveraineté instrument, alongside the arrival of Pierre Laboisse as President and CEO effective 26 September 2023. Bpifrance and regional business press corroborate the amount and the date.

The unresolved point: the 2020 release describes €120 million as the target size of a Series D whose first €80 million tranche had closed. The 2023 release describes €120 million as a raise being announced. Neither release cross-references the other, and no source we could open states whether the 2023 money completes the 2020 round or constitutes a new one. That distinction changes the company's cumulative funding by up to €120 million, so we refuse to compute a cumulative total at all. A reader who has seen "€200 million²+ raised" attributed to Aledia should know that the number depends on an assumption no publisher has made explicit.

One further discipline: the 2023 release names the State as a participant "via le dispositif French Tech Souveraineté". It does not disclose a euro amount for the public component, and we found no France 2030, IPCEI or European Commission document that itemises a grant to Aledia. "Backed by France 2030" is therefore reportable as a fact about the shareholder list, and not as a subsidy figure.

IV. The product, stated in volts

What Aledia now sells is described in its own datasheets with an unusual emphasis, and the emphasis is the strategic signal. The first commercial product, FlexiNova, is presented as a high-voltage architecture designed for mass production: datasheet ALD-FN1530F6 (Rev 1.0) describes a blue nanowire-based microLED of 15 × 30 µm in flip-chip configuration for direct transfer onto TFT backplanes, with a 6 V forward voltage; the company's flyer describes those emitters assembled on a 400 µm pitch backplane. On 3 April 2026 the company announced FlexiNova 9V, presented as the first microLED product operating at nine volts.

Why voltage, rather than brightness? Because at 6 V or 9 V a microLED can be driven directly by thin-film transistor backplanes and driver architectures built for higher-voltage operation, which changes the cost of the system around the emitter rather than the emitter itself. That is a manufacturability argument aimed at display makers' existing lines — the same argument, in a different industry, that 300 mm silicon made about epitaxy.

What is absent from that documentation matters as much. In this research pass we located no numeric external quantum efficiency figure, no luminous efficacy figure in lumens or watts per watt, and no colour-gamut specification in any openable dated source, company or third-party. For a technology whose entire commercial case rests on efficiency at small pixel sizes, the absence of a published efficiency number is itself a material finding, and we decline to import one from a vendor deck or an analyst note.

V. Context: a market with three incompatible sizes, and a cancelled anchor customer

Aledia's market is usually introduced with a forecast. That is precisely where a careful reader should slow down, because the forecasts do not agree with each other by an order of magnitude:

€50 billion, $9.94 billion and $1.23 billion for roughly the same decade cannot all describe the same object. They almost certainly count different things — panels, emitters, end devices — but no publisher states the boundary. Any strategic conclusion resting on "the microLED market will be worth X" is therefore resting on a choice of publisher, not on evidence.

The demand side has one dated, hard fact that cuts against the enthusiasm. On 22 March 2024, Bloomberg reported that Apple was winding down its in-house microLED display effort for future smartwatches; the same day, MacRumors, The Verge and 9to5Mac reported the cancellation, with earlier coverage on 1 March 2024 noting that ams OSRAM would reassess its microLED strategy after an unexpectedly cancelled cornerstone project. No source we could open ties Aledia commercially to that programme, and we assert no such link. The relevance is structural: the industry's most credible near-term volume anchor stepped back in 2024, in the same window Champagnier was finishing its start-up phase. Competitive comparison is likewise dated rather than editorialised — Taiwan's PlayNitride told Digitimes on 16 April 2024 that it would triple wafer capacity in 2024 and projected microLEDs undercutting OLED cost by 2030.

VI. What this record supports, and what it does not

Three conclusions survive every contradiction above. First, Champagnier is a 300 mm silicon-convention facility, which places it in the same tooling and skills universe as the Isère semiconductor base rather than in the LED industry's. Second, the company's commercial argument has migrated from photons to integration — voltage, transfer, backplane pitch — which is where its published documents now put their emphasis. Third, on the public record the company reached first commercial product in 2026, roughly fifteen years after incorporation, and its own republished retrospective is unusually candid that the preceding decade produced no revenue. That is the honest headline for a European deeptech factory story: the physics was never the schedule risk.

Refused in print

We do not publish, and did not estimate: any cumulative funding total for Aledia (the 2020 and 2023 €120 million figures cannot be summed without an assumption no publisher makes); any euro amount for the State's participation via French Tech Souveraineté, or any standalone France 2030 / IPCEI / European Commission grant to the company; any total cost for the Champagnier fab (€140 million and $200 million are both reported and are not the same claim); any single date for "start of production"; any EQE, lumens-per-watt, watts-per-watt, colour-gamut or lifetime figure; any pixel pitch other than the 400 µm backplane-assembly pitch the company's own flyer states; any wafer starts, yield, capacity or shipped-volume figure; any named customer, licensee or design win; any revenue figure; any salary band for any Aledia role; and any formal training partnership between Aledia and a Grenoble engineering school, none of which we could source.