Aixtron SE

XETRA:AIXA

Earnings Call Transcript 2024-Q3

[Operator]

Good afternoon, ladies and gentlemen, and welcome to the conference call regarding the 9 months 2024 results of AIXTRON SE. [Operator Instructions] Let me now turn the floor over to your host, Christian Ludwig.

[Christian Ludwig, executive]

Thank you, Mark. A warm welcome to AIXTRON's Q3 2024 Results Call. My name is Christian Ludwig. I am the Head of Investor Relations at AIXTRON. With me in the room today are our CEO, Dr. Felix Grawert; and our CFO, Dr. Christian Danninger, who will guide you through today's presentation and then take your questions.

This call is being recorded by AIXTRON and is considered copyright material. As such, it cannot be recorded or rebroadcast without permission. Your participation in this call implies your consent to this recording. Please also take note of the disclaimer that you find on Page 2 of the presentation document as it applies throughout the conference call. This call is not being immediately presented via webcast or any other medium. However, we will place a transcript on our website at some point after the call.

I would now like to hand you over to our CEO, Felix, for his opening remarks. Felix, your turn.

[Felix Grawert, executive]

Thank you, Christian. Let me also welcome you all to our Q3 '24 results presentation. I will start with an overview of the highlights of the quarter and then hand over to our CFO, Christian, for more details on our financial figures. Finally, I will give you an update on the development of our business and our guidance.

Let me start by giving you an update on the key business developments of the second quarter on Slide 2. The important messages for Q3 are: We have delivered a robust Q3 '24 and recognized solid orders with EUR 144 million. We concluded the quarter with revenues of EUR 156 million.

We came out in the lower half of the guided range because a customer requested to push the delivery of one larger project into Q4. The gross margin came out at 43%, strongly improved versus H1 due to an improved product mix. Our equipment order backlog at the end of Q3 remains high at EUR 384 million. For the fiscal '24, we confirm our adjusted guidance published in July '24. For fiscal '25, visibility remains low. Revenue for the next year is likely to be flat or even slightly below fiscal '24.

But more importantly than the short-term numbers is for us that we could further expand our strong position in the silicon carbide and gallium nitride. As you know, our market position in silicon carbide has been strengthened throughout the last quarter. In particular, we have made further technical progress with the G10 silicon carbide.

Early October, at the International Conference on Silicon Carbide and Related Materials called ICSCRM in Raleigh, North Carolina, we presented these improvements on layer thickness and uniformity, which puts us in the leading position in the industry, not only in productivity and cost, but now also in terms of uniformity and run-to-run tool stability.

For gallium nitride, on the other hand, we see the next technology step on the horizon, the move to 300-millimeter wafer technology. AI and other large volume applications drive this opportunity. We at AIXTRON have anticipated this and we are well prepared. We leverage the multiyear experience and the core elements of our 200-millimeter technology.

In addition, we have built or we can build on 25-plus years of Showerhead technology in our 300-millimeter single-wafer reactor. With this, we are in a unique position and therefore, we have started work on 300-millimeter GaN early and built a dedicated cleanroom, our innovation center for our 300-millimeter technology. I will elaborate on that in more detail later.

Christian will now provide a detailed look into our financials on the following pages before I then take over with an update on our markets. Christian?

[Christian Danninger, executive]

Thanks, Felix, and hello to everyone. Let me start with the financial highlights of our income statement on Slide 4. We had a good quarter with revenues at EUR 156 million compared to EUR 165 million last year. We came out at the lower end of the quarterly guidance as the delivery of one larger project was pushed to Q4 as per customer's request. In the first 9 months of '24, we recorded revenues of EUR 406 million, almost in line with last year's 9-month revenue number.

Gross profit in Q3 '24 was at EUR 67 million and EBIT for the quarter was at EUR 38 million. Gross margin in Q3 was at 43%, showing strong improvement versus H1 due to a better product mix. After 9 months, gross margin was 39%, down 4 percentage points year-over-year, primarily due to a less favorable product mix, which included a high share of lower-margin traditional LED systems.

OpEx in the quarter went up to EUR 30 million, primarily driven by higher R&D spending compared to the previous year. As we said earlier, we have a double load in 2024 as we are currently finalizing the development work for the G10 series, while we have started the development of our next-generation systems like the 300-millimeter GaN platform already.

Now to our key balance sheet indicators on Slide 5. We have now seen the start of the decrease in inventories to EUR 427 million compared to EUR 448 million at the end of Q2 '24. And as stated in the Q2 call, even stronger inventory reductions will materialize in Q4 and throughout 2025. Trade receivables at the end of September were at EUR 116 million compared to EUR 158 million at the end of 2023 and flat versus the number end of June. The advance payments received from customers at quarter end were at EUR 119 million, representing about 31% of order backlog, a similar range as last quarter.

Moving to our key cash flow indicators. Our operating cash flow improved both in Q3 as well as in the first 9 months as the inventory build has stabilized and reduction has started. After 9 months, operating cash flow stood at EUR 28 million, an improvement of EUR 94 million versus last year's negative EUR 66 million. In the third quarter, we generated EUR 15 million of operating cash flow, an improvement of EUR 10 million versus last year's EUR 5 million.

Free cash flow also improved on the back of the improvement in operating cash flow. In the first 9 months, it came in at still negative EUR 58 million compared to negative EUR 82 million last year. The improvement was less pronounced than for the operating cash flow as our CapEx after 9 months was at EUR 86 million and therefore, significantly higher than the last year's number of EUR 17 million, this primarily due to the investment in the innovation center.

In summary, we will see further improvement of inventories in Q4 when the larger shipments are planned, which will result in an improved operating cash flow. At the same time, CapEx will come down as we approach the completion of the innovation center. These 2 effects should result in a strong free cash flow in the last quarter of the year.

Our cash balance, including other current financial assets as of September 30, '24, were stable at EUR 78 million compared to EUR 79 million at the end of June. Cash decreased from EUR 182 million at the end of 2023, which was mainly due to the mentioned CapEx projects and our dividend payment of EUR 45 million in May.

With that, let me hand you back over to Felix.

[Felix Grawert, executive]

Thank you, Christian. I would like to continue on Slide 6 and give you an update on key trends in our different markets.

We introduced our G10 tool family to the market since September '22. The goal was that the G10 family covers all our key end markets with exciting growth applications ahead of them. The market traction of our G10 series has been very successful. Already for the full year '24, it will be significantly contributed to our equipment revenue. This underlines that our strategy with a strong focus on technology and innovation is paying off.

Slide 7 gives an overview of the markets we serve. I will just briefly touch on key points of the overview and spend more time on the news for GaN and SiC that has occurred since July. As stated before, in 2024, we see a wave of investments into traditional red LED capacities, leading to expected high double-digit million euro revenues in fiscal '24 from this market.

Companies which have been only delivering to the blue LED market for lighting or backlighting are now also investing into red LED capacity. Our G4 is the tool of record in this segment due to its proven track record and low total cost of ownership.

Despite the news from Apple and ams OSRAM, the industry is continuing to work on micro LED technology, driven by orders from several customers building R&D and pilot production lines. Our micro LED revenues for fiscal '24 are also expected to be at high double-digit million euros. All this is well known.

New this quarter is that we can share with you that many customers no longer regard the mass transfer aspect from manufacturing the micro LEDs as particularly challenging. This used to be a key inhibitor for volume adoption. Now we are hearing from various sources that progress is being made and that the pipeline of first product is in preparation.

We believe initially smartwatches, AR/VR glasses like the one showcased by Meta the other day and TV automotive applications will be the first product to market. Still, timing and volume predictions are uncertain, but various customers are looking into '27 or '28 for their production ramp, which would correspond to AIXTRON tool shipments in '26 and '27.

We talked about the current silicon carbide market dynamics in our Q2 call. Not much has changed since then. We continue to observe that the global market has currently built more capacity than there is demand. Nevertheless, we and our customers remain very positive about the mid- to long-term outlook.

Driven by better product offerings, the EV market will continue to grow. The growth is particularly fueled by sharply decreasing costs for the expensive silicon carbide substrates, which makes silicon carbide more cost-competitive compared to silicon IGBTs. In fact, while forecast for the number of EVs have recently been reduced, the penetration of silicon carbide has been increased due to this change. For example, we now start seeing silicon carbide MOSFETs even in plug-in hybrids, where carmakers try to get more mileage out of the battery charge.

Another trend which is driving silicon carbide adoption in EVs is the new 800-volt battery technology that allows for faster charging. This also relies on silicon carbide and more and more cars with this battery technology are coming into the market.

We at AIXTRON had anticipated increasing cost pressure and commoditization in the silicon market -- the silicon carbide market right from the beginning and our expectations are now being confirmed. This trend increases the focus on cost per wafer, which plays to the strength of our product. The G10 silicon carbide has established itself as the most productive tool in the market, coming also with the lowest cost of consumables, which means cash cost for our customers. In addition to that, we have been able to improve our tool performance even further.

We presented our latest silicon carbide epilayer results early October at the International Conference on Silicon Carbide and Related Materials, in short ICSCRM, in Raleigh, North Carolina. With our 200-millimeter batch reactor, we have now taken the lead on both quality of layer thickness and uniformity, not only for a single wafer, but for all wafers in the batch and for many consecutive runs throughout the day, for example. Thus, now we not only offer the lowest cost per wafer, but also a better quality than single-wafer tools. As a result, we expect to maintain a clear #1 position with a growing market share in silicon carbide.

Let me continue in more detail with the gallium nitride power electronics market on Slide 9. In this slide, we have summarized the recent development of the GaN market. We have discussed this in depth in our Q2 call. We observed that GaN power switches are taking market share from silicon power switches, application-by-application and step-by-step.

Some applications have already been around since 2020 like the smartphone charger. Others are just starting now like high-voltage motor drives and onboard chargers. Several customers are working with high intensity on increasing the energy efficiency of AI by means of GaN power technology and others are looking to address the market for EV main inverters with highest voltage GaN at 1,200 volts plus. These are growth opportunities we expect to kick in around 2026 or '27 onwards.

With our G10-GaN, we have been addressing the 200-millimeter market, which covers all applications alike. Now in September, a major power electronics player has publicly announced to go towards 300-millimeter with GaN as well. We at AIXTRON have been working on 300-millimeter GaN technology for quite a while and are very happy that now a leading customer is providing the confidence to the market that this technology is actually materializing. Hence, let me take some time to share with you our perspective on 200-millimeter versus 300-millimeter GaN and introduce the AIXTRON 300-millimeter GaN effort to you.

First of all, it is a great sign that the market for GaN is expected to grow into a size where 300-millimeter economically makes sense. Niche markets are typically served from smaller wafer sizes. We see that with our tools for laser applications, which are still today shipping in 4-inch or 6-inch wafer size. A 300-millimeter wafer offers 2.25x the area and hence, 2.25x the number of chips compared to a 200-millimeter wafer. The move to 300-millimeter hence confirms the growth trajectory of GaN that we have anticipated.

Secondly, we believe that the markets for 200-millimeter GaN and 300-millimeter GaN will coexist for a very long period of time. The dash line in the figure on Page 9 shows our expectation on what applications can be addressed with 200-millimeter and with 300-millimeter technology. The overlap between both is about roughly the entire market.

From what we see today, 300-millimeter GaN initially starts at low-voltage applications below 200 volt. These are thin layers, which are easy to make on the large wafers. But we also expect 300-millimeter GaN to cover the important and large 650-volt segment. This requires a bit more development work, but we have shown in our AIXTRON lab that this is possible.

The market for high voltage at over 1,200 volts is very new and just emerging. Hence, we expect this market to be addressed and open up based on 200-millimeter technology first.

Now what drives customer decisions for 200-millimeter versus 300-millimeter? We believe this decision will be driven to the biggest part by consideration of fab equipment reuse. The GaN epi tools such as offered by AIXTRON need to be newly purchased in either case. Existing silicon tools don't reach the required temperatures and don't have the gas control mechanisms needed for GaN. But a big part of the equipment downstream after the epi deposition can be reused, for example, by converting a silicon MOSFET production line into a GaN power production line. This reuse of existing fab equipment may, in our view, become a dominant decision factor for our customers.

Customers with a large installed base of silicon power equipment in 300-millimeter may want to convert this to gallium nitride as they expect GaN to take away market share from silicon. On the other hand, customers with several existing 200-millimeter fabs doing the silicon power production on that today may want to convert these 200-millimeter fabs in the future to 200-millimeter GaN rather than making new investments into a completely new set of tools.

What does this mean to AIXTRON? We will continue to invest in our 200-millimeter GaN platform to further enhance its performance, such that it constitutes also the most attractive offering for our customers in the market. In parallel, we have developed and we continue to develop our 300-millimeter GaN platform. The goal is to take the leading position both in 200-millimeter and 300-millimeter GaN to make us agnostic of customer wafer size selection.

With this, let me come to our 300-millimeter GaN tool itself on Page 10. AIXTRON will address the 300-millimeter market with a single-wafer Showerhead reactor. We built on over 30 years of experience in the GaN material system with this reactor type. Initially developed at Thomas Swan, a start-up from Cambridge University, this technology was our platform for the blue LED market. And for over 15 years, we have applied this technology to 300-millimeter wafer size. The initial market Three-Five-On-Silicon, in short TFOS, did not materialize, but provided valuable learnings and experience.

We have now ported the most recent experience and our secret sauce developed for 200-millimeter technology to this 300-millimeter GaN platform. With this, the 300-millimeter tool starts with the performance on par or even better than our 200-millimeter G10-GaN tool right from the beginning. We call the 300-millimeter GaN tool Hyperion. You will see the name in future communications. As you can see, we build an experience and technology learnings of several decades in 300-millimeter GaN and also transfer our most recent technology building blocks into this market.

In September, at SEMICON Taiwan, we have presented the most recent process results to select customers with very positive feedback. This makes us confident that we can capture the additional growth opportunity that the 300-millimeter GaN market provides to us.

Today, multiple Hyperion units have shipped to customers who are working with this tool in their R&D and pilot lines. This will take some time and we expect volume production shipments from AIXTRON starting 2026, 2027.

The 300-millimeter opportunity is also the reason why we decided to invest in our additional 300-millimeter cleanroom, our innovation center on Page 11 of the slide deck. With the recent market news, we can talk about this openly now. The innovation center has been designed and built with 300-millimeter technology in mind to provide us the space and the right environment to capture this opportunity.

It will allow us a much deeper collaboration and co-development with our customers. The construction is well progressing and nearing completion. This will add 1,000 square meter cleanroom space to our R&D operation. We have moved our first system into the cleanroom already and the hookup is happening as we speak. We expect first wafers out in Q4 this year and first joint customer projects are already scheduled for Q1 '25.

With that, let me now give you the update on our full year guidance for 2024. As stated before, we confirm our guidance for '24 as published in July. We expect total revenues in the range between EUR 620 million and EUR 660 million. We expect a gross margin in the range of 43% to 45% and we expect an EBIT margin in the range of 22% to 25%.

Our revenue guidance for the last quarter is therefore as follows: For Q4 '24, we expect revenues between EUR 215 million and EUR 255 million. This includes the larger project that was pushed out from Q3 to Q4 upon the customer wish. This reflects the usual seasonal pattern, shipments being Q4-heavy based on customer plans to take delivery of the systems, which are mostly derived from customers at completion dates. We have seen a similar trend in the past years, so nothing really new here.

Let me also give you an indicative outlook for fiscal '25. The medium and long-term drivers for AIXTRON's revenue growth remain fully intact. Efficient power electronics, in particular, for applications in IT and AI, silicon carbide technology for efficient electromobility and micro LEDs for future display generation. In the short term, however, momentum in the end market remains slow.

Therefore, it is possible that 2025 will be characterized by moderate customer demand, while orders for '26 are already being placed today and some orders are being pushed out towards 2026. As things stand today, revenues for fiscal '25 are expected to be at the level of fiscal '24 or slightly lower.

With that, I will pass it back to Christian before we take questions.

[Christian Ludwig, executive]

Thank you very much, Felix. Thank you, Christian. Operator, we will now take the questions, please.

[Operator]

[Operator Instructions] The first question comes from Madeleine Jenkins, UBS.

[Madeleine Jenkins, analyst]

I just have one on your new single-wafer GaN tool. I was just wondering kind of how the transition was going from batch to single, kind of what difficulties you're facing and also what feedback you've been getting from customers in your kind of initial pilot line phases?

[Felix Grawert, executive]

Thank you very much. Good question. It's going very well. And the tool is giving us until this moment, fortunately, only positive surprises. It's been a joyful ride, I can say. So to put more color on that one. So what I mentioned before, we take the elements, the building blocks, the modules, so to say, that we've developed on 200-millimeter technology and we have brought them into the 300-millimeter tool now, so to say, all these ingredients are in there.

And based on that, we've been able to achieve really fantastic results in terms of the uniformity of films, the crystalline quality and everything that you can imagine. And what I mentioned before, in some cases, we have achieved, in fact, results on a much larger wafer, which, of course, is more difficult to handle in a better quality than we have seen it in our 200-millimeter tools. So it's going very well.

[Madeleine Jenkins, analyst]

And then just a follow-up. I guess, given the dynamics now in gallium nitride, also in silicon, are you considering a tool to develop a single-wafer silicon carbide epi tool kind of for the -- I realize it will be long term, but for the future? Or are you comfortable with your batch proposition for now?

[Felix Grawert, executive]

Yes. We are very happy with our current tool. As I mentioned, we've been able to further boost the productivity and especially the performance of this tool. We are able now with the batch tool silicon carbide to achieve results, which are ahead of what is currently seen in the market from single-wafer tools. So there is absolutely no rationale to develop another tool.

[Madeleine Jenkins, analyst]

Okay. And just in the transition, I know it's really far out, but into 300-millimeter in silicon carbide, would your tool be able to be used for that?

[Felix Grawert, executive]

I'm not seeing 300-millimeter silicon carbide, honestly. I don't expect to.

[Operator]

Next question comes from Olivia Honychurch from Jefferies.

[Olivia Honychurch, analyst]

A couple from me, please. So firstly, on your 2025 outlook statement, you've guided for sales to be flat or slightly lower versus this year. If I press you which of the 2 outcomes do you think is more likely between flat and slightly lower? And I guess if they're slightly down, what sort of percentage decline do you think that could end up looking like? And then I've got a follow-up.

[Felix Grawert, executive]

So I think with flat or slightly down, with slightly down, I mean, something like 5% or 10%, let's say, midpoint of the guidance, which is probably the most realistic, I mean, the '24 guidance, yes. So that's for the percentage.

And now to your question, what is more likely? Honestly, it's really all about the market dynamics and how it further unfolds. Yes. I mean, in the end, so to say, it's a question, how long is the current downturn going to hold? Yes. And I think nobody knows that, right? Is the downturn continuing all the way through '25, like orders are only coming back at the end of the year, meaning shipments are pulling back and increasing again in '26? Or is it that whenever towards the middle of the year, the market returns with increased shipments at the end of '25 already, yes, I think nobody in the industry can predict that. And if so the person probably can make a lot of money out of that.

[Olivia Honychurch, analyst]

Okay. That makes sense. And then secondly, on the Q3 pushed out delivery, can you give a little bit more color there? What exactly was the size of the revenue delta? And what was the reason for that delay? And I guess following on from that, has that tool or that series of tools now been delivered since the start of Q4?

[Felix Grawert, executive]

It was a customer in the micro LED space who wanted his tools in December, yes, rather than October. So I think that pretty much answers the question.

[Operator]

And the next question comes from Martin Marandon-Carlhian from ODDO BHF.

[Martin Marandon-Carlhian, analyst]

The first one is on silicon carbide. Could you maybe give us a bit more color on which player drove the silicon carbide business order intake in the last 2 quarters? Is it about new customers or existing ones? And...

[Felix Grawert, executive]

Sorry, I didn't get the question. Let's stay on this one. Let's clarify. Can you repeat? I didn't fully understand.

[Martin Marandon-Carlhian, analyst]

Yes. I was wondering if you can give a little bit more color on what drove the silicon carbide order intake in Q3? Was it more about new customers that you won this year? Or was it more about existing customers?

[Felix Grawert, executive]

Well, it's both the mix of existing customers and new customers. So I think it's getting very broad. I think for silicon carbide, it's quite interesting. We see across our customer landscape, a very diverse landscape. We do see some customers who is running up to 50% of their capacity being idle.

And on those ones, the good news is if they are only using half of their tools, the tools which they continue to use is the AIXTRON tools and we love that because it shows us that our tools have the best productivity and lowest cost, yes. So they continue to run even if other tools get turned off.

And what's also interesting is we do see that some customers are fully loaded and asking us to pull in and to even accelerate their ramp. So we have some customers who have a very strong position in the Chinese EV space. It's not Chinese customers. This is outside of China customers, but they've made design-ins and some of the models, which is currently hot in the press about making competition to the European carmakers, yes.

So this is made on AIXTRON with customers -- AIXTRON customers outside of China. So we have a pretty -- and these customers who have made it to get designed into the China silicon carbide supply chain and car industry, you can imagine they are asking us to pull in to accelerate even further, yes, while some of those customers whose factories is half empty, of course, are pushing out orders and don't accelerate. So it's an extremely diverse mix.

Just to shine color on that. Of course, as always, we preserve the confidentiality of the customers. So I cannot give further detailed indications to that. But that gives you an idea why our silicon carbide continues to be -- yes, despite the overall market slowdown, yes, better than you would expect if you just take the average market.

[Martin Marandon-Carlhian, analyst]

Okay. That's helpful. And still on silicon carbide, do you see now other applications outside automotive like [indiscernible] ramping up in volume? Or is it too soon?

[Felix Grawert, executive]

Honestly, it's a bit difficult to say which ones. But yes, I can give you a qualitative answer. I don't have a quantitative answer. I'll leave that to the market analysts of [indiscernible] and Gartner and these guys.

So what we do see is that, of course, the EV main inverter is the main application that is there for sure. We get more mileage out of the battery. That's for sure. We do see that both wide bandgap, GaN and SiC are fighting with each other in the onboard chargers, which also is a decent volume. And I think they make a hard time for silicon, pushing that out.

Silicon carbide having a big benefit from the falling wafer prices, yes, especially now due to China pushing hard on the wafer prices, yes, this makes silicon carbide cheaper. And with that getting cheaper, silicon carbide gets more attractive and wins market share, very clear. But also gallium nitride, I think, is doing a good push on that one.

The other market where we do silicon carbide currently gaining momentum and again, attractiveness coming from advanced technology, but also decreasing prices is the market -- the industrial market of higher voltage classes. So here, you speak about devices in the 1.7 or 3.3 or even 10 kilovolt range, where silicon carbide is taking away market share from silicon IGBTs and the applications you find in large wind turbines, large solar plants and overall energy conversion as well as in large super high-power motor drives, whether this is stationary motor drives or whether you find that, for example, in rail traction like high-speed rail and trains and so on, yes.

And yes, we do know this market is increasing. We see also increasing customer demand coming to our laboratory and saying, "Hey, AIXTRON, can you help us?" And we work with customers. Now what part of the market that makes, I cannot give you the details. But I do see a strong momentum from that segment.

[Martin Marandon-Carlhian, analyst]

Okay. And maybe the last question, if I may, maybe more of a question for Christian. On the gross margin, I mean, if we look at the guidance for 2024 today, it kind of implies a very strong gross margin in Q4, around 50% or slightly above, which would be a record. Can you help us maybe reconcile how we get to that number? Is it linked mostly to mix to the new G10 systems? Yes.

[Christian Danninger, executive]

Yes, Martin, very clear that this question is on the mind of a lot of people. Yes, I mean, just doing the math, this requires a very strong Q4. This will be supported on the one side by the pushed out, I mean, these systems are pretty much ready for shipment, they are the pushed out systems. And then at the end, we're talking about the similar level like last year.

So the main contributors here to achieve that will be the improved mix that will benefit to the gross margin. On the other side, also better fixed cost aggression, so some operating leverage in there and of course, the volume effect of the just pure high volume, yes.

So with all of that, we believe that we will come out within the guidance range. Where exactly time will show as it depends on which tools really ship them at the end of the year. But overall, we're on a good track here.

[Operator]

Next question comes from Gustav Froberg, Berenberg.

[Gustav Froberg, analyst]

My margin question was just answered, but I have 2 other ones on silicon carbide and on 2025.

Starting with silicon carbide, you talked about broadening your customer base. How many customers do you estimate that you have in silicon carbide right now? And have you started the process of selling into Chinese either wafer makers or chip makers as well? I know this was talked about in the past.

And then a question on phasing in '25. You say flat to slightly down versus midpoint of guidance year-on-year next year. Does this require a lot of growth in the second half of '25? Or do you see this being more evenly spread throughout the year next year?

[Felix Grawert, executive]

Gustav, thanks a lot for your question, but they're difficult to answer because it's all about numbers, which honestly, I don't really have, yes. So let me try nevertheless and fill in qualitatively where I don't have the exact figures, yes, if you allow me.

Honestly, I don't know how many silicon carbide customers we have. The customer base has broadened significantly. I can really say that. I can say it's a truly global footprint. If I think we all know, it started off pretty much in Europe and the U.S., yes. But by now our customer base also has broadened very well, I would say, in every space where the semiconductor industry is located.

We have a decent number, multiple in Taiwan. We have multiple in Korea, and not only 1 or 2, but multiple. And we have multiple, if not even many in China. And to your question about China, yes, we also have quite significant orders from Chinese customers, both in a number of customers, but also a few high-volume customers, meaning large customers, in terms of multi-tool orders and with multi, I mean like a dozen or more than a dozen, yes, so not just 5 or 6, yes.

So I would say you can consider this by now a broad global and diverse customer base in terms of footprint, size, geography and applications, yes. And allow me that I don't have precise details. I can just say it's all over the place, yes. Maybe that helps you already.

Now with respect to the exact revenue distribution in '25, honestly, I don't have the numbers. But I mean, you know our seasonal patterns. Typically, you can, for sure, expect a strong Q4, yes. I mean you've seen it in the last year. So you see this year, you've seen it last year. And you can definitely repeat that to come. Deriving from that, for sure, I would expect the second half to be a bit stronger than the first half, but I don't have the exact numbers and I think a lot is still in the flux, yes. So let's see how it comes.

[Operator]

Next question comes from Michael Kuhn, Deutsche Bank.

[Michael Kuhn, analyst]

A few follow-ups from my side. Firstly, let's say, maybe on the end markets that were not so intensively covered so far. Obviously, we have seen a little comeback of the old LED business over recent quarters. Is that continuing for now? Or are we coming down again here?

[Felix Grawert, executive]

The LED business is typically lumpy. We still have some in the order pipeline, but you never know when exactly it's coming, yes. What I mentioned already in my speech is that we have seen now quite chunky orders from a few large customers who in the past have in China only been addressing the blue LED market.

And now as signage, these large displays are coming. And then, of course, the road map is to shrink them eventually then calling it micro LED, yes. So some of these LED players have now been completing their portfolio. In the past, they only used to have gallium nitride-based LEDs, which is blue. And now they complete that also by green and red, yes, and red is based on our G4 tools, so to say, yes. So this is what's currently happening.

Yes, there is some more to come. We expect that, yes. But again, given that this market is lumpy and always was lumpy, yes, if you pull back to the last time we had red LEDs, we were always talking about lumpy big orders. So typically, you get a PO and the customer wants 20 tools in a shot, yes. And then you have a few quarters nothing and the next customer comes and the minimum order size from these guys typically is a 10 or a dozen, sometimes even more. So there is more to come. But when exactly and how it is distributed, I cannot say.

[Michael Kuhn, analyst]

Given, let's say, the implied Q4 gross margin guidance, I would assume there's hardly anything from the LED side involved?

[Christian Danninger, executive]

Of course, the mix -- the end market mix and also the generation mix in Q4 will be significantly better. Yes, I cannot exclude a few LED shipments in there as well. I mean, we are talking about EUR 250 million to EUR 255 million. But -- so overall, the mix will be significantly better.

[Michael Kuhn, analyst]

Understood. And then one more. I mean, obviously, let's say, the sluggishness in some of the end markets is now taking a little longer than you had initially anticipated. I mean, we all know you're a very focused company still. Is there, let's say, any opportunities outside your traditional core markets that you would probably take a look at or any, let's say, other diversification opportunity just to, let's say, drive growth and get a little more diversified maybe?

[Felix Grawert, executive]

There's always something in the pipeline. I think by now, you know us, yes, hence, also your question. However, we all know in our market, which is highly technology, highly innovation-driven, some things materialize, others don't materialize. And I don't like to speak about things before they are really there because it's very much fluent. So let's take it when it comes.

I think the key message I would like to you guys to take home is I think we have a pretty cornering around '25, yes, at least based on what we see today is our best guesstimate now. The real guidance will come in February of '25. And I think we've provided you quite some indicators that we really see '26 and then '27 with many of our addressed core markets today kicking back in.

Silicon carbide is our expectation to come back, micro LED to materialize in the next momentum and GaN to materialize both in 200 and also in 300, yes. So I would like to leave it at this point for today.

[Operator]

Next question comes from Martin Jungfleisch, BNP Paribas.

[Martin Jungfleisch, analyst]

I have 2, please. The first one is on the orders. I mean, you mentioned that there was a pushout for revenues into the fourth quarter. Was there a pushout for orders into the fourth quarter as well? And maybe could you provide any color on your expectations for Q4 orders? Should they be relatively on par with Q3, slightly better, slightly worse? That's the first question.

[Felix Grawert, executive]

In terms of orders, let me try to take your first part. So I understood your first part, but let me really make sure that I got it. I'm not sure. I took your question whether there was a pushout of orders or only a pushout of revenues, right? Did I understand right? Yes, there was only a pushout of revenue. It was a pushout from a customer shipment. The customer said, I don't want your units in October, I want your units in December. But that has nothing to do with an order. The order had already long placed.

[Martin Jungfleisch, analyst]

Okay. And then just any color you can give on the fourth quarter orders? Would it be relatively on par with [indiscernible]?

[Felix Grawert, executive]

In this market environment, I think everybody is just putting their numbers up and listen to what our customers are communicating to the market also, right? So to say, right now, I just read all their news. And I think the key message they provide to their shareholders is we are pulling down on our CapEx. We are very disciplined. And I hate no message more than that, because my customers' CapEx is my order. We probably get the same. So let's really see what comes.

[Martin Jungfleisch, analyst]

Cool. And then second question is on this Italy site that you acquired earlier this year. Can you provide some color on the ramp of the production there? And if you would incur any idle cost for this site in the beginning?

[Felix Grawert, executive]

Well, what we mentioned before is, I mean, we were able to strike simply a pretty good deal, yes, which is we're able to buy a brownfield site. We looked all over Europe. We're able to buy a brownfield site for a single million-digit amount, yes. And we are planning to have first shipments out of Italy in Q4 of this year as part of our revenue guidance and then gradually ramp it up in Q1 and Q2.

And due to the fact that this was a brownfield site and we are just starting now, we are able to adjust our, let me call it, renovation investments. The facility is there and we can make some renovation for fitting it out in cleanroom style. We are able to do that now step-by-step as the revenues are coming. And of course, given the currently slightly slower market momentum, we are able to do that in -- just as we are filling this, we built a team up there, but we've not made any big large investments, which come back to us now as idle cost.

[Operator]

Next question comes from Nigel van Putten, Morgan Stanley.

[Nigel van Putten, analyst]

Just a follow-up on the gross margin, but more into next year. Perhaps you can play it off the current quarter, which seems to be the average quarter you need to reach your revenue guide for next year, 43%. Would it be fair to also model or put at least 40% into our models for next year as a start? Or would that not be correct? That's my first question.

[Christian Danninger, executive]

We are hesitating to give any indication here. I mean, it's the first time really we have already, at this point in time, given a revenue indication for next year. With margins, be it now gross margin, EBIT margin, we should really wait until Feb when we come out with the real guidance for next year. We need to see what the mix would look like and so on. I mean these dynamics that we have discussed in the past about the end markets, lower traditional red LED shipments, more next-generation shipments. But to be honest, we don't really know yet. So I would be hesitant to give more detail at this point in time.

[Nigel van Putten, analyst]

Okay. I guess similar question on OpEx, perhaps that's more under your control. As sort of sales are not going to show growth next year potentially coming down, should we expect OpEx down as well? Or are there elements like higher depreciation upon completion of current investments like the headquarters and the refurbishments in Italy you just referred to that would sort of skew that number higher into next year?

[Felix Grawert, executive]

You asked a question that we haven't even figured our mind out. So you are ahead of the curve, yes. I would clearly say, given that revenue will be flat or reduced, you should not model an increase in OpEx, yes. Whether you should model a down or reduction and how much of that, let's wait like Christian also indicated for our guidance that we provide in February. But I can tell you, definitely, there will not be an increase in OpEx.

[Operator]

And the last question for now is from Malte Schaumann, Warburg Research.

[Malte Schaumann, analyst]

First one is also on silicon carbide. Maybe you can elaborate on your confidence on winning further the other larger customers out of the top 5 players? And what's your take on the potential time frame when this might happen?

[Felix Grawert, executive]

Our confidence is high. And for a time frame, I'm expecting 2026.

[Malte Schaumann, analyst]

'26 sounds pretty late.

[Felix Grawert, executive]

Well, do you know anybody who's ordering from the last 5 in '25? Tell me the guy.

[Malte Schaumann, analyst]

Okay. Good. And then 2 tiny ones. First is on the mid-'24 guidance, you seem to be shooting for the -- rather for the midpoint of the guidance than for the lower end. Is that right? And then another tiny one, how much of the order backlog from the current order backlog you have, do you expect will finally be for '26?

[Felix Grawert, executive]

So I think we've left the guidance range open because both the low end and the high end is possible, yes. So the orders are there. And as always, there's some external factors we cannot influence, yes, which is some parts from suppliers, which need to come in on time in order to be able to ship and finish the tools. On some point, there's also some export licenses open, which is normal at this point in time, still 2 months ahead of the end of the quarter, yes.

So these are the 2 factors. If they're not coming out so good, it's the lower end. If all of that comes in, then we will be at the higher end, and that's the reason why we have the full range. So we don't know it, you can toss a coin.

[Malte Schaumann, analyst]

Okay. And on the order backlog for '26, can you share a number how much of your current backlog is probably for '26?

[Felix Grawert, executive]

Well, I mean, for this year, we are still -- for '24, we are planning to ship about 50% to 60% of the backlog, yes. Into '26, I don't know if I have that number for '26...

[Christian Danninger, executive]

I think it's more than we usually have because we had some pushouts I indicated already going into '26. I think it's a low double-digit million amount, which we see already for '26.

[Operator]

Thank you very much. Right now we have no questions anymore. Back to Christian Ludwig.

[Christian Ludwig, executive]

Well, then thank you very much, all, for your questions. If there are still questions open, the IR Department is at your disposal. Please give us a call. And otherwise, if we don't see you before year-end to talk to you, then we'll see you or talk to you at the latest with our full year release end of February next year. Have a great day. Thank you and goodbye.