Hello, everyone, and thank you for joining us today for Soitec's third quarter 2026 fiscal year sales conference call. I'm Pierre Barnabé, Chief Executive Officer of Soitec.

I'm joined today by Albin Jacquemont, our Chief Financial Officer, Steve Babureck, our EVP Strategy, and Alexandre Pétovari, Head of Investor Relations. First, I would like to step back and reflect for a moment on where the company stands today.

As you all know, I'm preparing to step down in just under 2 months when my successor, Laurent Raymond, takes over as Chief Executive Officer. This leadership handover comes at a time when Soitec is stabilized, successfully diversified into new growth areas, and well prepared for the new phase of development ahead.

The acceleration of our diversification has been enabled by our sustained yet targeted investment in R&D. Today, Soitec is focused on operational excellence and cash generation, supported by a very healthy balance sheet.

This gives the company the resilience to navigate diverse end-market dynamics and the flexibility to be well positioned for the next growth cycle. In light of this, the third quarter reflects a high level of rigor in execution in a challenging market environment.

It should be seen as confirmation that Soitec is operating with discipline, using the strategic and operational levers at its disposal. In this context, our priorities are very clear.

We are focused on execution, operational excellence, disciplined cost and cash management, and on aligning capacity with actual demand. At the same time, we continue to invest selectively in areas that are critical to our long-term technology road map without compromising financial rigor.

The overall picture remains broadly in line with our expectations. At a business level, this translated into the following trends: sustained activity in artificial intelligence for Edge and Cloud, where demand remains supported across data centers and edge environments.

At the same time, other parts of the business are still adjusting. In mobile communication, customer engagement remains active around advanced RF solutions and next-generation platforms, despite constrained volumes with RF-SOI demand still impacted by customer inventory correction.

In automotive and industrial, while the environment remains challenging, our technologies are well positioned to support future applications once demand recovers. I will now walk you through the third quarter sales figures and our outlook in more details.

Starting with the headline numbers, third quarter revenue reached €160 million. On a year-on-year basis, revenue remains down, reflecting a 22% decline at constant scope and foreign exchange and a negative currency impact of 7%.

This top-line performance is driven by different dynamics: strong trend in artificial intelligence offset by ongoing RF customers inventory correction in an uncertain smartphone market and weak automotive demand. Sequentially, revenue increased 18% organically compared with the second quarter, coming in above our guidance.

This improvement reflects the strong commitment of our teams. Looking at revenue dynamics by end market: Mobile Communications generated revenue at €90 million in the third quarter, down 36% year-on-year organically.

As expected, RF-SOI volumes remain impacted by customer inventory correction, which is improving, but not yet complete. POI activity was slightly down year-on-year while showing an improvement from Q2.

Higher demand from Tier 1 US fabless was offset by softer activity in Asia. FD-SOI adoption in 5G millimeter-wave is still ongoing, with a major recent design win for US flagship smartphone.

Edge & Cloud AI revenue amounted to €54 million, showing 27% organic growth year-on-year and improving from Q2 '26. Activity continues to be supported by AI-related demand across edge and cloud environments, with contributions from both Photonics-SOI and FD-SOI.

Specifically, Photonics-SOI remains very dynamic, as the technology continues to enable a growing number of high-speed, high-bandwidth optical interconnect applications, including pluggable transceivers and co-packaged optics, CPOs. We continue to strengthen our differentiation with new materials addressing growing interest for LNOI, lithium niobate-on-insulator, for ultra-broadband data communication.

These products will complement our silicon photonics offering to cover optical transceivers beyond 1.6 terabytes. Automotive & Industrial revenue reached €16 million, improving sequentially versus the second quarter, but remaining down 32% organically year-on-year.

Market conditions are still weak. Power-SOI volumes remained low during the quarter, reflecting both market weakness and delivery phasing that is skewed towards Q4 under a long-term agreement with a key customer.

Looking at the first 9 months of fiscal year '26, revenue reached €390 million, down 26% year-on-year organically. Performance over this period reflects significant RF-SOI undershipment, prolonged inventory corrections, and contrasted end-market dynamics.

Edge & Cloud AI showed strong resilience, especially excluding the anticipated phaseout of Imager-SOI, bringing the performance of this segment up 29% year-on-year. Mobile Communications and Automotive & Industrial remain more exposed to market adjustment, as expected.

For the last quarter of the year, we expect around 20% organic revenue growth versus Q3. We maintain our cautious stance on the market environment, and are closely monitoring the evolution of the smartphone market outlook and its potential impact on the ongoing customer inventory correction.

Mobile Communications should improve sequentially. Edge & Cloud AI should continue to show strong momentum, driven by demand for Photonics-SOI and FD-SOI.

Automotive & Industrial is expected to remain soft overall, but Q4 will benefit from seasonal deliveries for a specific customer. In this still challenging environment, we remain focused on disciplined cost management and cash generation, while continuing to execute our strategy through targeted R&D, technology leadership, and product diversification.

This concludes my comments on our Q3 '26 performance and Q4 '26 outlook. Then, after more than 4 years, this is my last call with you.

When I joined, Soitec was in a very fragile situation: unprecedented governance crisis, dependence on a single product, more than 2 years of RF-SOI inventories concentrated with a few customers with deeply damaged relationship, and a heavy cost structure. Year after year, we managed the depletion carefully while giving ourselves the time and the resources to accelerate 4 new product lines which have all reached critical mass, we have doubled R&D investment, and restored the balance sheet.

Today, the company is getting ready to turn a corner, driven by a successful diversification of product aligned with AI and energy efficiency megatrends. RF and Power-SOI demand will resume as customers clear their inventories.

Our innovation remains dynamic, with increased investment to strengthen our positioning in our current SOI market and to prepare our expansion into new markets. Our innovation capabilities will be critical to meet the industry's most promising developments, with key players now exploring SOI for advanced computing applications and memory.

Our relationships with customers and partners have been rebuilt and strengthened, from the US to China, where we executed a new strategic reset with NSIG. Looking back at these last 4 years, what I am most proud of is having recruited, developed, and retained outstanding professionals and teams.

We have made Soitec stronger and ready for a solid and sustainable rebound, because my extended team and the Board of Directors trusted me, understood the plan, and executed it with discipline. I'm deeply grateful.

I now hand over a solid platform to Laurent. I would like to pay a tribute to the ExCom members, to our diverse leaders across the organization, and to all employees for this successful journey.

With that, we are now happy with the team to answer any questions you may have. Thank you very much.

Yes, good morning, and thank you. Thank you, Pierre, and all the best wishes for your future endeavors.

Can you give us a little bit of a color on how you're doing with progress on rationalization of working capital, what kind of levers you have in place to drive a positive free cash flow into next year, and if you could give us an idea of how we should think about CapEx going into fiscal '27 as well? Thank you so much.

Thank you, Alex. Thank you very much.

Then I propose for everything regarding cash discipline and working capital containment to leave the mic to Albin, who could comment on the progress made so far and what's the trends for the end of the year and some colors for the next year to come.

Sure. Alex, look, we said in Q2 that we're working towards free cash flow positive in fiscal year 2026, and that remains true today.

We are making progress on that front. What we have been doing is managing the costs in a stringent manner while maintaining selective and strategic investment in R&D.

As for inventory reduction, which is another priority, we're aligning production with actual demand. Of course, the downside of this is that it brings a temporary hit on margins and profitability.

On the receivables front, we have renegotiated or we are in process of renegotiating some contracts to lower receivables that will yield results in this year and in fiscal year 2027. The way to think about our capital expenditure is moderation, I would say.

We are guiding towards CapEx of €114 million cash this year, and there is a little bit of flexibility built into this number. Capital expenditure will be much lower in fiscal year 2027, and that because we have already invested a lot to build capacity for the company, there is no need to expand our facilities.

Remember that we changed the free cash flow definition in Q2 this year, and this was to align with market practice, and now our free cash flow mirrors the debt variance in absence of buybacks, dividends, and M&A. We're making progress on working capital and free cash flow.

That's great. Thank you very much.

That's very clear. Can I just have a quick follow-up on your mobile business?

I have a couple of questions. The first one is, are you concerned about the health of the mobile phone market?

We've seen some forecasts now pointing to a decline in calendar '26, given the tensions we see in memory prices in particular. The second part of the question is, could you give us an idea by how much you're undoshipping into the channel?

What's the kind of order of magnitude there in your RF-SOI product line? Thank you so much.

Alex. On the smartphone market, we are observing, listening like you and reading like you that there are some tensions coming from memories prices, and availability that could weight on the dynamic of this market.

Then we're going to revise or not our forecast during the Mobile World Congress. We were expecting a growth around 2% of this market in volumes for this calendar '26 year.

We might revise maybe this trend looking at the different analyst view. That said, what is the weighting the most, of course, for us is the inventory depletion that is going in the right direction.

Then we continue to undership, and we are accelerating this undershipment over the last quarters. Then from you remember, the 2.5 million estimated inventories in 8-inch equivalent RF-SOI by our customers on September, we moved to 2.3 end of December, we were around 2 million.

Then if we continue to undership between 200 to 300k wafers per quarter, we could imagine that by end of this year, calendar '26, we're going to be in the range of 1 million. We don't know if this 1 million is the right momentum and turning point, but it does correspond to the volumes of our customers' inventories before the COVID events, then 5, 6 years ago.

That's what we see. Then we undership quite significantly since the last quarters and for the next 4 quarters to come to be sure that we're going to reach a level that is close to what we were observing 5, 6 years ago before the COVID.

That's what we can tell you.

That's very clear. Thank you so much, and good luck.

Thank you, Pierre.

Welcome. Welcome, Alex.

Hello, Pierre and the team. Of course, Pierre, all the best for the future.

I have three questions. Three questions.

First, Pierre, this is your last publication, your last call with us. Do you think you are leaving Soitec at the low point of the cycle?

Second, are you able to keep up with the strong demand for Photonics-SOI, or are you facing any constraints in terms of industrial capacities, your suppliers? My last question, maybe for Albin, to what extent are you hedged on the euro-dollar for the next financial year?

Thank you very much.

Thank you, Emmanuel. Then Soitec obviously is reaching step-by-step an inflection point.

What is clear is that we are building a platform that is ready to rebound in the near future. It's too early to tell you exactly when it's going to come.

It's clear that the growth machine has been totally reinforced with these five new products today, four new products on top of the RF-SOI that are critical masses and many of them are growing. Namely Photonics, I will come back on it, of course, FD-SOI, and POI.

The two others, RF and Power, are in inventory depletion mode. Then, of course, as we said, by end of this calendar year, we're going to have reached the pre-COVID level of inventories by our customer.

Then you can imagine that step-by-step, we're going to reach this inflection point to growing again. What is also very important to my point of view, and we need also to think on the future beyond the quarters for which we give guidance, is that we have doubled the R&D investment from around €60 million to €120 million today, also to prepare the future and to add new products on top of the five I was mentioning.

To underline what you said and what you asked on Photonics: Today, Photonics is in a very dynamic momentum. We have no specific limitation, but to be sure we organize properly our factories, because, of course, we are sharing the same factories for the 200 and particularly the 300.

Then it's a matter of anticipation. We have access to the right level of bulks.

As we diversified also our sources of silicon bulk in terms of quality and particularly the one able to correspond to Photonics requirements and demands. We have today more and more customers asking for Photonics and more and more in volumes, 200 and 300.

We are working on enhancing, thanks to the LNOI technology, the capabilities of our Photonics portfolio by extending the bandwidth and the capacity to reduce latency. Then now it's really a question of anticipation.

It's a question of signing LTAs with some of these customers to better anticipate what's going on and to manage the allocation. Today, there is no specific limits to take the ride and to accompany a market that is growing by around 25% to 30% per year.

Albin, do you have any point on the third question, please?

Absolutely. It's clear that foreign exchange continues to be a significant external headwind for Soitec, and the reason for that is that we have a significant US dollar exposure on our top line, this makes us sensitive to exchange rate fluctuations.

The one thing I would like to point out is that we guided for fiscal year 2026 on a 1.14 euro per dollar, and this number has remained constant throughout the year. We had hedged quite effectively our transactional exposure in fiscal year 2026.

For fiscal year 2027, we will give you the number you should factor at the next call. Nevertheless, what I can tell you at this point is that we have hedged 70% of our exposure at 1.

1.19 at this time. There is 30% which remains to be hedged.

Thank you very much. Thank you, Emmanuel.

Hi there. Most of my questions have been asked, but maybe you could just talk a bit more about the quantum computing opportunity.

It seems like a lot of the initial quantum is using a PIC and an EIC. How much of those PICs are using SOI today?

Interested in that recent SkyWater deal, for example, or are you at a sort of nascent stage? Thanks.

Hello, Robert. Then on on on quantum, of course, and and this part of some press release we we we we should recently.

It's a market we are, of course, working on. We are observing very cautiously, and SOI is is bringing some some capabilities very interesting for for for for qubit generations, including, of course, low power consumptions.

It it remains exploratory. We are working in R&D mode in this areas.

For the moment, the volumes expected, even midterm, are quite limited as long as, of course, this technology is not spreading out massively and becoming a mass-market application. We are not yet at that stage.

There are some hypothesis it could become, of course, in a hybridation with binary calculation the mass-market features, then the mass-market application. Then we are in.

SOI is is presenting some technological interest. We are working collaborations with some labs and a few emerging companies, but still not in our, let's say, radar screen for in a midterm point.

In the long term, we're going to see step-by-step, and of course, we're going to update you on the different progresses made.

Got it. Just a quick question on the loading.

Maybe you said this already, but I couldn't hear it. What is the plan for fiscal '27 from a loading point of view?

Are you just going to take the hit on the profit, run down inventory on hand, which is very high still, and basically, in a year's time start with a clean slate? Is that the goal?

What loading percentage are you targeting? Then

Albin, perhaps you can give some complement?

Yeah. As for the loading is concerned, I think one of the key action for us is to really match procurement with manufacturing, with delivery, with end-market needs.

That's a priority for us. We're not guiding on fiscal year 2027 revenue.

You understood from Pierre that we see a lot of tailwinds for some of our products. You also understand from Pierre that there is still some inventories on the RF market.

From that, you can infer that our loading will remain at a level which is below normal level in fiscal year 2027. The impact on the profitability would be pretty much the same as to what we indicated at the end of Q2.

Again, what is important for us is to act sequentially. Now we are concentrating on free cash flow, working cap, our own inventories, then at some point we'll see revenue stabilization and before a rebound.

When we see a rebound, the loading will go up, and obviously, hopefully, where we would be in this place, profitability will go up, but in a context where we have lower G&A and a working capital which will be closer to what we should have on our balance sheet.

Got it. Thanks a lot.

Welcome.

Hi, good morning. Thank you for taking my question.

Pierre, thank you again for everything, and wish you also all the best in your future endeavors. I would have two two question.

Oh, or maybe I'll just start off with RF-SOI. On the inventory digestion that you mentioned through 2027, is it fair to assume that the first half would be more on the higher end of this 200 to 300k range, or will it be more even throughout the year?

Then also just on overall goal levels, you mentioned the 1 billion, sorry, the yeah, 1 million level that would be aspirable. Just factoring in, let's say, the midpoint, 250k per quarter, that puts us at 12-month high inventory levels as basically expectation.

Pre-COVID, the levels at customers were roughly 6 to 9 months. I'm just wondering is there any reason why the aspirable goal is now higher than the previous level, or yeah, that would be just great to understand.

It's difficult to say. We give you points of comparisons compared to what we are experiencing and what we experienced in the past.

Then the 200k to 300k per quarter is an average. Then of course, it depends on quarter after quarter.

Over 1 year, we can say that depletion is around 800k to 1.2 million within the course of 12 months. It's an average.

That's the reason why from 2 million end of December '26, with the undershipment we are doing right now, we should be in the range of 1 million by end of this year. The 1 million correspond to what was in the inventories customers, in RF-SOI, before COVID.

Is it the right let's say turning point to see customers let's say investing again, buying again, and to see the RF-SOI business for us growing again? We don't know.

It's of course an assumption we share with you. Will it be before, will it be later?

That's really an observation, an assumption, and comparisons. We're going to see within the course of 1 year the needs for of course our customers.

We're going to see if some also breakthrough innovation might accelerate some let's say investment and some requirements and ask for RF-SOI specific volumes. That's the point of comparisons we give you and a trajectory.

That's very important and to give you figures that are as reliable as possible. Of course, we can imagine that calendar '27 should be a year of growth for RF-SOI, but we don't know exactly when.

Perfect. Then just quickly also on Edge and Cloud AI, just given now throughout the last quarter, there were a few photonics players mentioning different targets in terms of growth, which yes, fairly from a lower base sound more aggressive.

I just wanted to ask, are you still comfortable with the 20% to 30% CAGR range for this business? Are you slightly more bullish, and is there potential to come even more above this range, or are you let's say comfortable with the midpoint for right now?

The term comfortable doesn't exist in our world. Are we confident, and are we let's say working hard to beat the growth of this market, of course.

Today, we have positioned our silicon photonics solutions everywhere we can. We have more than five customers today engaged in our silicon photonics solutions, and more and more are coming, and each of these customers asking for more and more volumes.

Then clearly there is a dynamic. I'm confident that this dynamic should last because the need for AI are very important.

The demand for AI is higher than what the industry today can provide. Of course, we are a part of this components of this industry because this silicon photonics is today the only solution, viable solutions that is allowing first of all next-generation AI data centers transfer, control, transform electron into photons for high bandwidth, particularly if you move up to 1.6 giga, not only for the data centers, but tomorrow to interconnect the GPU and particularly with a co-package optics solutions.

On top of it, we are enhancing our photonics layers with lithium niobate let's say add-on that is helping the industry to move beyond the 1.6 giga, tera, sorry. That means that we have today a plan, a clear plan, and the industry have a clear plan to fuel the silicon photonics growth for the coming 4, 5 years without any doubt.

Our market shares, positionings, relationships with our customers is making us confident. Now we need to, as I said to Emmanuel, to anticipate this development, this growth, to prepare the industry to have enough rooms for delivering on time, on quality, and even to beat the growth of this market that, as I said, is around 25% to 30% per year.

Perfect. Thank you very much.

Very clear.

Welcome.

Hello, everyone, and thanks for taking my question. One on the competition in the SOI market.

Have you seen any kind of change in the competitive landscape with new players reinventing position, being a little bit more aggressive? I'm thinking about GlobalWafers plans that is building up in the US.

Do you see them gaining a little bit more market share in your key market? The second question is coming back to Photonics-SOI.

Can you remind us a little bit the difference in terms of content between pluggable optics and co-packaged optics? Is this a big step-up when we are moving to CPOs going forward?

Thank you.

Yes, Sébastien. Then on the overall SOI market, of course, we always say that there is competition, starting with our only licensee today with S.E.H., but we don't see big changes in the segment shares we are observing, and particularly for us regarding the 200 and the 300.

Of course, players are claiming to have SOI solutions, and some SOI solutions are not using our technology like Smart Cut. You can have also other ways to do SOI, silicon-on-insulator, namely BSOI, that is quite well adapted for power application in the cars, but even in that domains, competition and our positioning, it's quite strong.

Then today, we don't see segment shares significant evolution, and the competition is far away from what we are doing, all the more that we continue to invest in our SOI technology. As I said, we have doubled our gross R&D investment from €60 million to €120 million over the last 4 years, and of course, it is fueling our advance from innovation point of view.

We are filing more and more patents every year, more and more. This year, we're going to even beat what we did last year that was quite exceptional.

Then we continue to run ahead. Talking about GlobalWafers, as I just said, we have only today one licensees, only S.E.H., because, as we have decided in '24 to terminate the license contract with GlobalWafers.

Of course, that means that GlobalWafers is no longer authorized to use any Soitec IP in any form. They have a 2 years kind of transition time till July '27 to adapt themselves, because at this time, we will be in a position, of course, to infringe them for any copy or any used of our patents and IP.

Of course, it is coming years and years after they decided to build a factory in '21, then of course, they will have to manage this huge changes for us, but that's of course the name of the game. Then that's really the situation.

If we look at the Photonics-SOI contents, then if you look at the transceivers today that are used to interconnect servers to servers in the data centers are using a classical architecture, that is the first architecture that is allowing to use 400, 800 up to 1.2 1.2 tera. Now the industry is looking for more bandwidth and also to reduce latency.

If you want to reduce latency, you need to insert your let's say interconnections at the substrate level. This is co-packaged optic in a sum up architectures that's going to help the boards, the graphical and processing unit boards, to interconnect each other.

Then the co-packaged optics is really a new architecture for which we see some pull-forward, because it was expecting to come in '28, then '27, then today, we have heard that some first let's say design going to be commercially available by end of this year, then there is a clear accelerations. On top of it, we are developing and accelerating the development of LNOI that gonna enhance the capability of data throughput and data bandwidth.

We clearly see a very strong evolution in the architecture of the Photonics let's say capabilities. The transceivers are really at the beginning, and we are selling a lot of wafers particularly to serve these needs, but the next waves that going to come will serve the co-package optics first wave, and then we're going to have some evolutions, particularly with miniaturization and concentration of signal plus LNOI.

Then of course, it is part of the 25% to 30% growth we will observe in this market for the coming 5 years. Then of course, the volumes of wafers asked by the market gonna grow more and more, but it is also part and it's translations of the co-package optic adoption, then the evolution, and the LNOI adoption at the same time.

Then it's really correlated with what we see, volumes versus technological evolution and new design wins.

In terms of content in mm², do you have some data for pluggable then co-package optics, or this is not something that you have in mind?

We have, of course, some datas, but we prefer to speak in volumes of wafers. Because, of course, millimeter-square evolution is changing, particularly with new packaging, new 3D packagings, that is also modifying a bit the way to see the let's say the footprint in a planar way.

Of course, things are changing. It's better to think wafers volumes and type of wafers volumes.

That's more accurate.

Very clear. Thank you.

You're welcome, Sébastien.

Hey, guys, and thank you, Pierre, for your work, and wishing you all the very best for your future endeavors. A couple

of

questions from me. First question is, I believe you said by the end of calendar '26, we'll reach customers will reach about 1 million wafers in inventories.

I was wondering what kind of, in terms of the smartphone growth assumptions, what are you assuming for this? Are you assuming closer to your existing sort of plus 2% smartphone unit growth, or are you assuming closer to what some analysts are now banking in in terms of a decline for this year?

That will be my first question.

Then, Oliver, the impact of the growth rate is not significant. 1% correspond to few 10,000, a few thousands wafer.

Again, talking 8-inch equivalent, to be sure that it's very clear. Then the impact, it's quite limited.

What we are really working on is to continue undoshipping to reach this 1 million that correspond to what was the level of inventories before the COVID. Then the growth rate, whatever it is positive or negative in, of course, a single-digit mode, will have a limited impact in terms of volumes effect for us and might change for some weeks this inflection point if 1 million is an inflection point.

Got it. Then also just wondering how margins are trending for the second half of this fiscal year.

You guys did around 34% EBITDA margins in the first half. Just wondering from both the gross margins and opex how things are trending and what we should be aware of for the second half.

We don't communicate, of course, precisely on the margin evolution, but maybe Albin can give you some let's say trends and colors.

Absolutely. Just on gross margin, the thing to keep in mind is that there are several drivers.

There is the currency, there is the mix, the price, and there is the loading of the fab. The loading of the fab is by far the most potent driver of margins.

We don't guide on gross margin, but think of the fab loading as a drag on gross margin of somewhere around 600 bps. In addition to that, there is a foreign exchange, which should be a drag of approximately 300 bps.

This is half-on-half? Yeah.

Of course, it works both ways, huh? When fab would reload, we will see operational gearing.

Of course, particularly after the effort we made.

Got it. Very helpful.

Thank you.

Welcome, Oliver.

Thank you very much for your interest in Soitec and for the depth and quality of your questions. The next date in our agenda will be the release of our fiscal year '26 results on May 27th with Laurent after market close and a presentation held in Paris on May the 28th.

In the meantime, we will be very happy to meet with you on the road and at the Mobile World Congress in Barcelona very soon. This ends our call for today.

Thank you.