Yes, good morning, ladies and gentlemen. Thank you for holding and joining our earnings call. Everything live, so please excuse the short delay. You have on our call our CEO, Jochen Hanebeck, our CFO, Sven Schneider, our CMO, Andreas Urschitz. The Electronica Trade Fair currently makes Munich the epicenter of electronics, and what better place to broadcast this call from than the vibrant Infineon booth. Different location, established procedure. Jochen and Sven will provide an overview on the market situation and divisional performance, key financials, and of course, our awaited outlook for fiscal 2027. Our prepared remarks will also cover projected further AI growth, our sustainability achievements, and the dividend proposal. After that, we will start our Q&A session.
The accompanying slideshow for the call is available at infineon.com/slides, we will provide a PDF with Jochen's and Sven's introductory remarks in the course of the call on our website, namely infineon.com investor. There, you also will find a recording of this conference call, our press release, as well as our investor presentation. Now, Jochen, over to you.
Thank you, Alexander, and good morning, everyone. Let's begin with a little retrospective of our very eventful 2026 fiscal year. Twelve months ago, we predicted only a gradual, uneven recovery. In hindsight, cyclical and structural growth momentum has strengthened and broadened strongly. The difference in revenue between our weakest and our strongest quarter amounted to more than EUR 1 billion. AI drove unprecedented structural growth, and with EUR 1.7 billion, we again more than doubled our revenue in this area. The convergence of structural growth and upcycle is continuing, driven by AI and in its many shapes and forms, power infrastructure, as well as Automotive, where electrification and software-defined vehicles are strong secular drivers. To prepare Infineon optimally for the coming high-growth phase, we executed successfully on major pillars of our strategy in 2026.
We opened our new smart power fab in Dresden at the best possible point in time. We shaped our business portfolio further through acquisitions and divestitures, the same way as we continued to streamline our manufacturing footprint further. Since 2022, we went from 20 sites down to 13. To further strengthen our application focus, as well as to improve speed and agility of our organization, we adopted a new, more application-centric divisional setup. Overall, we feel very well positioned for significant profitable growth in our just-started 2027 fiscal year. But before we come to our outlook, let's look back at the recent past. The fourth and final quarter of our 2026 fiscal year saw record revenues of EUR 4,706 million, up 13% sequentially. The year-over-year growth rate of 19% is further illustrating the acceleration of our business.
For the full fiscal year, revenues amounted to EUR 16. 353 billion, an annual increase of 12%, which at constant U.S. dollar-euro exchange rate would have been even 15%. This growth rate was twice as high as the one we had originally anticipated. Applying our professional cycle management, our teams handled the rapid turn in the cycle very well. In lockstep with revenue, also our profitability expanded throughout the 2026 fiscal year, with Q4 clearly being the strongest with a segment result of EUR 1. 119 billion. This corresponds to a segment result margin of 23.8%, 470 basis points up from the quarter before and slightly better than anticipated, mainly driven by positive volume and mix effects. For the full 2026 fiscal year, the segment result margin was 19.7%, up 220 basis points year over year.
Regarding our order backlog, we have decided to provide qualitative color going forward and cease to disclose specific numbers in line with the majority of peers. That being said, we did see a further increase in incoming orders continuing over the course of the September quarter, relating to widening supply constraints while distribution and customer inventories are at or below target levels, thus supporting visibility into the new fiscal year. Now to our divisional review, which we are doing one last time along our former four divisional setup, commencing as usual with Automotive. In the final quarter of our 2026 fiscal year, revenues of ATV increased by 8% quarter over quarter to EUR 2. 89 billion. Growth was driven mainly by microcontrollers, smart power components, and MOSFETs.
The segment result of ATV increased to EUR 468 million, corresponding to a segment result margin of 22.4% compared to 18.4% in the previous quarter. The improvement was driven mainly by higher volumes, supported by some positive mix and pricing effects. Looking at the last fiscal year, our mainland China auto business grew twice as fast as our global ATV revenue, as we could fortify our quality lead in an otherwise supply-constrained local automotive-grade semiconductor market. Currently, we see a rising order intake from automotive customers as unchanged long-term trends driving content growth are complemented by some healthy restocking. In an overall muted car market, fuel cost increases and the introduction of more affordable models are supporting rising xEV adoption in Europe, India, and Southeast Asia. At the same time, OEMs are starting to roll out more advanced software-defined vehicles architectures across all major regions.
The ATV segment is the prime example of our stringent portfolio management. We strengthen our strengths, as in the case of Ethernet solutions, which plays out very successfully in combination with our microcontroller position. We are fixing core areas in need of fixing, as the refocusing of our high-voltage drivetrain business is showing, and we exit when non-core areas have a better owner. Therefore, we announced in September an agreement to sell our NOR Flash and F-RAM business to Winbond in an all-cash transaction valued at $1.12 billion. Subject to required approvals, we target closing in the second half of calendar year 2027. Our constantly evolving portfolio is rewarded in the marketplace. We have recently secured a major design win for a dynamic domain control system at a leading Japanese car manufacturer.
The solution is based on our latest generation AURIX TC4 microcontrollers and underlines our pole position in software-defined vehicle architectures also in Japan. Another example is a design win with a North American electric vehicle manufacturer. Our HybridPACK Drive silicon carbide modules will be used in the traction inverter alongside AURIX microcontrollers, power management ICs, transceivers, and gate drivers, demonstrating once more the breadth of our profitable system offering. Let's now look at Green Industrial Power. GIP's revenue increased by 13% quarter-over-quarter to EUR 503 million, marking the September quarter the third consecutive one of double-digit growth. All application areas contributed to the sequential increase. The segment result improved to EUR 69 million, corresponding to a segment result margin of 13.7%. The improvement was driven by higher volumes and a favorable product mix.
The recovery in industrial markets continues to broaden, while power infrastructure remains the strongest structural growth engine. Custom orders continued their upward trajectory, accompanied by emerging supply tightness in high-performance silicon carbide products and high-power modules. Investments in grid modernization, renewable energy integration, and AI-related power infrastructure are supporting demand for energy storage systems, solid-state transformers, and solid-state circuit breakers. We note a strongly growing design win pipeline in these applications. Our technological and reliability differentiation and our system understanding positions us well to benefit from continued growth in power infrastructure, with additional opportunities emerging from factory automation and broader industrial markets. A word on silicon carbide, which in Infineon's case is driven jointly by industrial data center and automotive applications. In our 2026 fiscal year, we achieved revenues of EUR 700 million and annual growth of 10%.
For the coming year, growth is expected to significantly accelerate, and we project silicon carbide revenues of well above EUR 1 billion, driven by our broad portfolio targeting a multitude of applications. We are determined to keep leading in innovative wide-bandgap semiconductor technologies, as demonstrated also by our first processing of functional silicon carbide power devices on 300 mm wafers, announced yesterday. Now to Power and Sensor Systems, which had a stellar quarter and accelerated further on its profitable growth trajectory. PSS achieved revenues of EUR 1.796 billion in the September quarter, a full 25% up sequentially. Compared to the same quarter one year ago, the growth rate was a staggering 48%. Similarly, the margin of PSS expanded significantly. The segment result increased to EUR 554 million, leading to a segment result margin of 30.8%, approximately 6 percentage points more than in the previous quarter.
Besides the significant volume increase, also mix and pricing had a positive impact. The very positive growth and profitability developments relates first and foremost to our leadership position in AI power solutions. We achieved EUR 1.7 billion of dedicated AI power revenues in the concluded fiscal year 2026, equivalent to a bit above 10% of group revenue. Together with power solutions for non-AI racks and servers, our overall data center business amounted to around EUR 2.2 billion in the last fiscal year, of which around EUR 800 million pertained to the September quarter. Going forward, we will report and forecast one unified number for the overall data center business. New AI processor generations and higher compute density go hand-in-hand with higher power density and novel architectures for central data centers. At the same time, evolving AI workloads like inference and task coordination translate into significantly rising CPU growth rates.
Our key focus is to increase our dollar per kilowatt content by addressing the needs of our customer for even more sophisticated vertical power delivery and other modules, the transition to 800 V architectures, as well as high-power parts reaching into the grid, like solid-state transformers. Said differently, AI is broadening and continues to be a substantial tailwind for us in terms of growth and profitability. Our just started FY 2027 fiscal year revenue growth will again be supply limited, in other words, a function of successfully executing manufacturing capacity expansions. We strongly benefit from the fact that we will be ramping our new smart power fab in Dresden, as well as scaling up manufacturing partners to serve unabated customer demand. In our last earnings call, we already flagged that leading customers across the data center ecosystems are engaging through multi-year delivery and capacity reservation agreements with us.
The total cumulative sales volume of these agreements, both signed and in negotiation, has surpassed the EUR 10 billion mark. Through volume commitments and prepayments, investments, risks are being shared. For our FY 2027 fiscal year, we project data center revenues of close to EUR 4 billion, adding another year of tremendous growth. A key ingredient to our unrivaled leadership, besides a highly differentiated portfolio system-level expertise and in-house manufacturing capacity, is innovation, which helps us to drive our average dollar content per kilowatt. As a prime example, we are advancing our collaboration with NVIDIA to include gallium nitride-based intermediate bus converters. Our GaN offering supports the full range of conversion ratios from 800 V to 50 V, 12 V, and 6 V , giving data center operators the flexibility to optimize future rack architectures without compromising power density or efficiency.
Accelerating innovation to customer value in AI was also the motive behind our acquisition of C2i Semiconductors, a Bengaluru-based technology company specializing in software-defined multiphase controllers and smart power stages for vertical power delivery, including substrate-integrated voltage regulators. All in all, Infineon remains the preeminent provider of solutions across the full power delivery chain from grid to core. Let's complete the divisional review with Connected Secure Systems. CSS recorded revenues of EUR 318 million in the September quarter. The sequential decline of 9% related to temporary fluctuating project business and supply constraints affecting embedded security and identification solutions, which more than offset positive momentum from the ramp of consumer eSIM and improving microcontroller and connectivity demand. The segment result of CSS amounted to EUR 25 million, corresponding to a segment result margin of 7.9%, reflecting the decline in revenue.
Looking ahead, recovery signals are broadening across industrial smart home and medical applications. Edge AI is increasing the need for secure and energy-efficient processing, sensing, and connectivity. With the expansion of our PSOC control portfolio targeting motor control and power conversion, we are strengthening our ability to bring differentiated system solutions to industrial, including infrastructure, AI, and consumer applications. Physical AI is developing into a major new growth driver for Infineon, not in a distant future, but beginning here and now as our cooperation with NVIDIA is demonstrating. Our integrated motor control solution combining Ethernet connectivity, microcontrollers, power, and sensing is available on NVIDIA's robotics developer platform. Digital twins of our reference designs let developers perform tests in NVIDIA simulation environment before hardware is built. Our certified trusted platform modules integrated into NVIDIA's robotics computing platform provide security across the product lifecycle.
Robot developers can thus move robots into deployment quickly and securely. Here at Electronica, we show a use case in semiconductor manufacturing with a partner. This concludes our divisional review for today. Before handing over to Sven for the financials, let me briefly comment on our changed organizational setup and its link to Infineon's target operating model. As you know, we are operating for a bit over four months now in these three divisions: Automotive, Edge Systems, and Power Systems. The key guiding principle for this streamlined setup is clear application ownership, including but not limited to our main secular growth drivers. ATV remains responsible for the structural automotive trends of software-defined vehicles and e-mobility. The focus of ES is on applications at the edge between the physical and digital world, in particular edge AI and robotics, the latter being jointly supported by all divisions.
PS is in charge of powering data centers from grid to core, as well as power infrastructure. The transaction into the new structure has worked really well. Early customer and employee feedback is very supportive. To further explain the individual value proposition and how each division drives profitable growth and contributes to Infineon's overall target operating model, the three division heads will each do a capital markets update throughout the course of our 2027 fiscal year. ATV will start on 7th of December. Our targets for Infineon stay in place, including growing the company by more than 10% through the cycle, and we are fully confident of our ability to capture structural opportunities. For the foreseeable future, we expect the PS division to grow significantly faster than 10%, whereas for Automotive and ES, we anticipate growth of around 10%. Now over to Sven for the key financial figures.
Thank you, Jochen, and good morning, everyone. In my part, I will focus on the September quarter, occasionally adding comments about full-year numbers. The fourth and final quarter of our 2026 fiscal year saw continued favorable gross margin development. The reported gross margin increased to 41.8% from 40.8% in the quarter before. The adjusted gross margin reached 45.1%, up 230 basis points from the previous quarter's 42.8%. Higher volumes in conjunction with positive mix and pricing effects drove this margin expansion more than compensating rising input costs for items like precious metals and freight. As communicated, idle costs stayed on the same level as one quarter before, given our continued focus on bringing down inventory reach. For the full 2026 fiscal year, the adjusted gross margin increased to 43.1% from 41.4% in the year before.
The main factors at play on the positive side were top-line expansion, lower idle costs, they declined as predicted from close to EUR 1 billion to just under EUR 650 million, and further additional benefits from our structural Step Up program. Main margin burdens were annual price downs of a low single-digit percentage on group level and a significantly negative currency impact. Additionally, the high-voltage drivetrain business within Automotive constituted a margin drag of about 1 percentage point on group level. Now to our operating expenses. Research and development expenses amounted to EUR 671 million in the September quarter, sequentially constant. Selling, general, and administrative expenses went from EUR 433 million to EUR 470 million. On an annual basis and net of non-segment result charges, R&D as a percentage of revenues was just under 15%, SG&A close to 9%.
Further positive contributions from the Step Up program on our OpEx were partly offset by higher incentives linked to business performance. Non-segment result charges for the fourth quarter went up to EUR 344 million, bringing the total for the 2026 fiscal year to EUR 1,008 million, practically the same level as one year earlier. The biggest components of the non-segment result are acquisition-related amortization and stock-based compensation. The financial result amounted to EUR -88 million after EUR -63 million in the prior quarter and contained adjustments to the carrying value of participations held at equity. Income tax expense for the September quarter was EUR 163 million, equivalent to an effective tax rate of 23%. Cash taxes amounted to EUR 93 million. The quarterly cash tax rate stood at 13% only, unusually low due to tax refunds related to prior years.
For the full 2026 fiscal year, the effective tax rate was 23%, the cash tax rate 20%. Going forward, for modeling purposes, a tax rate between 20% and 25% remains a reasonable assumption. Our investments in property, plant, and equipment, other intangible assets, and capitalized development costs increased to EUR 931 million in the final quarter of the 2026 fiscal year, bringing the annual total to EUR 2.6 billion, a good EUR 100 million below our projection. Depreciation and amortization expenses, including acquisition-related non-segment result effects, amounted to EUR 499 million. The annual total for D&A came in at EUR 1.9 billion. Now to our free cash flow from continuing operations. In the September quarter, the free cash flow increased to EUR 657 million after EUR 599 million in the quarter before.
The closing of the acquisition of the sensor portfolio from ams OSRAM at the beginning of the quarter was overcompensated by a higher operating result and positive trade working capital effects. The reach of our on-books inventory went down quarter-over-quarter from 165 to 136 days. This number is influenced by reclassifying inventories belonging to the memory business to be sold to Winbond as assets held for sale. Adjusting for this effect, the DIO at the end of September would have been 144 days, slightly better than this year's target. On a full-year basis, our reported free cash flow amounted to EUR 993 million.
The adjusted free cash flow, considering acquisitions and expenditures for large front-end buildings, in particular the mentioned ams OSRAM transaction and expenditures for our smart power fab in Dresden, came in at EUR 1.9 billion, equivalent to 11.8% of sales, in line with our target operating model. Our gross cash position stood at around EUR 2.1 billion at the end of September, benefiting from the positive free cash flow in Q4, less payments made for share buybacks in connection with our equity-based compensation plans. Gross debt amounted to around EUR 6.8 billion as per the end of our 2026 fiscal year, net debt to around EUR 4.7 billion. Gross leverage of 1.7x is below our maximum target level of two times. Net leverage sits at 1.1x .
Our after-tax reported return on capital employed continued to develop positively in the final quarter of our FY 2026 and came in at 9.6%. That being said, and despite the positive trend, the value of 7.2% for the full fiscal year is falling short of our aspirations, underscoring the imperative to pursue further profitable growth. Jochen will comment on the positive further ROCE trend in our current fiscal year later in the outlook. As the FY 2026 lies behind us, I would like to briefly comment on our dividend proposal. At our annual shareholders meeting in February, we will propose a dividend of EUR 0.38 per share, EUR 0.03 up from the prior year. This would result in a payout of about EUR 500 million.
Following two downturn years in which we kept the dividend constant, a prudent raise strikes a fair balance between shareholder remuneration and sufficient financial flexibility for the profitable growth of the company in the coming years. Now back to Jochen, who will present and explain our outlook for Q1 and the entire FY 2027. He will use our new setup for divisional comments. To enable comparisons for you, please see our quarterly press release or the analyst presentation for adjusted historical data for Automotive, ES, and PS. An electronic version is available on our IR website.
Thank you, Sven. The market environment for our outlook is clearly favorable, as structural tailwinds are supported by ongoing cyclical recovery. Broadly speaking, demand is developing very well, and supply conditions are tightening. In particular, secular opportunities such as AI, power infrastructure, and software-defined vehicles offer additional growth potential. Lead times continue to stretch further. Customer and channel inventories are at or slightly below target levels. Customers increasingly focus on securing supply further out. Rising backlog levels are improving visibility and lead us to expect very positive business performance into and throughout our FY 2027. For our outlook for the first quarter as well as the full fiscal year, we are assuming a U.S. dollar exchange rate of 1.15. Our rule of thumb for currency sensitivity going forward is being adjusted in line with our increasing business volume and higher relative U.S. dollar share.
For every EUR 0.01 change in the exchange rate, quarterly revenue moves by around EUR 30 million and quarterly segment result by around EUR 15 million. For the December quarter, we expect revenues of around EUR 4.6 billion. The very slightly sequential decline corresponds to a much better-than-usual seasonality, which is also apparent in a projected year-over-year growth rate of 27% for our first fiscal quarter. By segment, for ATV and PS, we expect revenues to hold up quarter over quarter, whereas for ES, we predict revenues decline moderately. We expect the first fiscal quarter segment result margin to come in at around 23%, reflecting the slight volume decline. For the full 2027 fiscal year, we expect revenues to come in at around EUR 20 billion, equivalent to very high annual growth of more than 20%.
From a segment perspective, PS is poised to even grow materially faster than group average on the back of highly dynamic demand for AI power solutions, as already indicated, and increasing momentum in power infrastructure and other industrial markets. For ATV, we are expecting mid-teens growth, driven by strong content growth from the broader adoption of software-defined vehicles and rising xEV penetration in some car markets. Specifically on the high-voltage drivetrain business, which we are refocusing, I can report that we are seeing first signs that our efforts are bearing fruit. The revenue run rate has inflected, and profitable design win activity for future years is picking up again. For the 2027 fiscal year, however, it will still slightly dilute growth and margin for ATV.
Lastly, for ES, we expect revenue to grow at low teens percentage, and as gradual recovery in most of the segments and markets is complemented by structural growth in applications like robotics, smart home, and medical devices. Strong top-line growth on company level will go also hand-in-hand with noticeable margin expansion. We expect the full-year adjusted gross margin to go up to a mid-to-high 40s level and the segment result margin to increase to a mid-20s level. Apart from the positive fall-through from significantly higher volumes, we project idle cost to go down to an annual level of around EUR 350 million, the decrease being equal to around 1.5 percentage points of margin. Furthermore, we expect pricing to have slightly positive impact over the year on group level, as well as the full run rate of contributions from our Step Up program, as commented by Sven.
To some extent, offsetting these factors will be rising input costs for various materials and energy, higher depreciation and amortization for manufacturing ramps, as well as rising absolute operational expenses, primarily on the R&D side, to protect and deepen our competitive moat through innovation. In order to support our profitable growth, we expect to spend between EUR 3 billion and EUR 3.5 billion on investments in the 2027 fiscal year. Key projects in the current and coming years will be the ramp of Dresden 4 for power and analog mixed signal technologies, expanding Kulim 3 for silicon carbide, and starting to build the world's first gallium nitride production on 300 mm wafers in Villach, as commented two weeks ago. On the backend side, we will focus on building out our high-volume sites in Malacca and, in the future, Bangkok.
The exact magnitude of overall investments within the projected corridor will be a function of tangible customer demand for 2027 and beyond. By doing so, we drive our large-scale manufacturing sites into further improving economies of scale, supporting our competitive differentiation. For , we anticipate a level of EUR 2.2 billion, including amortization of around EUR 300 million resulting from purchase price allocations, which will be recognized in our non-segment results. Based on our favorable business outlook, we are expecting the reported free cash flow to increase to around EUR 1.8 billion for the adjusted free cash flow, which is net of investment into major front-end buildings and M&A transactions. We expect a level of around EUR 2.2 billion, which would correspond to around 11% of group revenues. For our return on capital employed, we project an increase to low teens level.
Two housekeeping remarks on our outlook figures. First, we have not baked in the planned divestiture of our NOR Flash and F-RAM memory business. Assuming a closing of the transaction within our 2027 fiscal year, our free cash flow would rise by close to EUR 1 billion pre-tax. The impact of a sale on group revenue, segment result, adjusted free cash flow, and ROCE would be immaterial due to the closing assumed to occur in the second half of calendar 2027. Second, our forecast implicitly assumes that geopolitical conflicts or trade tensions do not significantly escalate further. Complementing the business and financial aspects, I would like to briefly touch on our sustainability metrics. We are very well on track to reduce our CO2 emissions related to scope one and two.
Notwithstanding significantly rising production volumes, we have reduced such emissions by 81% by end of September compared to the base year 2019. Also, for our scope three target, which we have set last year, we are very well on track. By 2029, 72.5% of our suppliers should have a science-based target. Currently, we are already close and at 65%. In addition, we have given ourselves a water recycling target this year. By 2030, we want to recycle 50% of our process water. Currently, we are at 28%. Ladies and gentlemen, let me summarize. The September quarter was a strong finish at our fiscal year 2026 with a record EUR 4.7 billion of revenue and 45% gross and 23.8% segment result margin. Secular growth drivers are proving to be very strong. At the same time, the upcycle is fully on track, and market strength is broadening, and supply continues to tighten.
AI is highly dynamic, and Infineon remains at the forefront of powering data centers from grid to core. Our unrivaled portfolio, coupled with a significant additional clean room space, will propel growth. We expect revenue in this area to significantly increase to close to EUR 4 billion in our fiscal year 2027. We are growing strongly also beyond AI. Our great business diversification is letting us benefit from further key growth drivers such as power infrastructure and software-defined vehicles. Based on our bright business outlook, we expect significant profitable growth in fiscal 2027, around EUR 20 billion of revenue, mid to high 40s gross, and mid-20s segment result margin, and around EUR 2.2 billion of adjusted free cash flow. Innovation speed, operational excellence, and active portfolio management are the basis for our continued value creation.
Thank you very much. Ladies and gentlemen, this concludes our introductory remarks, and we are now opening the call for your questions. As a reminder, this call is being broadcast live from Electronica. As we would like our board members to go out and continue doing business with customers soon, we kindly ask you to limit yourself really to one question and one follow-up today. Thank you so much. Operator, please start the Q&A session now.
Thank you. Our question and answer session will be conducted electronically. If you would like to ask a question, simply press the star key followed by one on your telephone. If you are joining us-
Operator, please take the first question.
If you are joining us today using a speakerphone, please ensure that your mute function is turned off. We will take our first question from Sandeep Deshpande from JPMorgan. Please go ahead.
Yeah. Hi. Thanks for letting me on. My question is regarding your guidance of the EUR 4 billion in AI revenue in 2027. How do you look at that revenue with regard to the orders? Do you have substantially more orders today in AI revenue versus what you have guided the market to, and how you are positioned in terms of your capacity if there were to be even much more orders through the rest of this year?
Thank you, Mr. Deshpande. I am afraid we still have some technical issues. The speakers could not hear you. Please stay connected. The Q&A session will continue any moment. Thank you. We have returned now. Mr. Deshpande, please repeat your questions. Thank you.
Yeah. Hi. Thanks. The question I have is, you have guided the market to about EUR 4 billion in AI sales in FY 2027. How do you see your orders in AI in the coming year, in the FY 2027? If the orders are bigger than the EUR 4 billion that you have guided the market, do you have the capacity to deliver much more than EUR 4 billion in sales?
Yeah. Hi, Sandeep. This is Jochen. I can confirm that we have more than that in terms of orders. The final number will depend on output of our factories, of course, mix, and pricing effects. I think those are the dominant factors. If they all play to the upside, then also the EUR 4 billion will be crossed.
Understood. I have a question on pricing, Jochen. This year you have had, or rather FY 2026, you had multiple increases in price on various products that Infineon sells. How do you look at pricing into FY 2027? Do you expect to be raising pricing again in FY 2027? You previously said Automotive pricing for the OEMs and tier 1s will be raised in the new negotiations, but what about the rest of the products?
This is Andreas speaking. Hello, Sandeep. We see very strong demand and a broadening of the given allocation situation, mostly driven by the AI boom. That one has now started to spill over to most of the other business segments. Having said that, we, as Infineon, do benefit right in time from our manufacturing capacity ramped in Dresden. This is definitely good news. Coming back to your question on pricing, as a consequence of what I said, pricing has already been increased two times in 2026, as you know pretty well, mostly for power. Currently, we see this to radiate into other segments. Thus, while we are having our volume purchase agreements that happen while we speak with our largest OEM, we find ourselves in pretty positive pricing environment all around.
I am not yet in a position to give you figures on that and where this will result, but all in all, pricing environment towards 2029 is positive, and we will report on latest achievements in due time.
Thank you.
To reiterate, the AI business is clearly supply limited at Infineon.
Then the next question comes from Didier Scemama from Bank of America. Please go ahead.
Good morning, gentlemen. Thank you for taking my question. I have got a question on the Automotive ATV division, because clearly you are guiding to something that is surprisingly better than expected, considering some of your German OEM customers are struggling. I guess my question to begin with, and I have got a follow-up, is to what extent are you exposed or overexposed to the China xEVs coming into Europe? Is your market share in those models better than it is in the domestic models? And I have got a follow-up. Thank you.
Yeah, I take this one. Indeed, we are guiding ATV to grow mid-teens. Let me clearly make here also the statement, given our capacity expansion, we are clearly able to serve good business on the automotive side as well as the AI side. We do not have to make choices as competitors have to do. And of course, that gives us already the first opportunity that certain competitors are not able to follow. When asking about China, I said that we grew significantly more in mainland China last year for our automotive business than our overall global automotive sales. That gives you a good indication also on our competitive positioning.
When it comes now to the OEM exposure, I may remind us all that we have very similar market shares in China as compared to global, and your assumption that Chinese OEMs are typically, especially if it goes into export, choosing reliable high-quality, proven sources is a correct one. All of that helps us, besides, of course, also the trends. Let us not forget, the Chinese OEMs are at the forefront of software-defined vehicle, which is one of the strongest growth drivers for the industry, but also for Infineon.
Very clear. Thank you, Jochen. I think that point is often missed by investors, so thank you for clarifying. My second question is on AI, and I am afraid it is on two fronts. One, on the LTAs that you have announced, like EUR 10 billion or above EUR 10 billion, can you give us a sense of the maturity profile of those LTAs? Is it two years? Is it beyond that? Then secondly, on the VRM business, clearly an incredibly strong position from Infineon, but competition is increasing with some of your U.S. competitors making acquisitions. We are also hearing, obviously, risk of, let us say, diversification from some of your customers on that front. So how do you feel about your competitive position in VRM over the course of 2027? Thank you.
I answer first the second one, and I have to ask, VRM is now vertical module in your case, or is it IVR?
No, the vertical module.
Vertical module.
voltage regulation module, sorry.
Okay. You were talking about the significant acquisitions of other competitors. I think everybody who wants to participate in that race, and in this case, that's why I'm asking, it's IVR, which is at the horizon, so really the power stage in the substrate. Everybody who wants to play in that needs to have certain competencies. If you're starting on a base, you may need big acquisitions. In our case, we only needed a bandwidth increase, and that's why we bought this company, C2i Semiconductors in India, which is exactly what we need. The purchase price was less than EUR 100 million. So I think we are building on a strong base and just amend our competencies for the IVR space.
If you talk about current voltage vertical power modules, here we are also at the forefront and are ramping the latest generations of vertical power modules with various hyperscaler and GPU customers, by the way, across the globe, not only in the U.S. The first question, I hand over to Andreas.
Yeah, Andreas here. Talking about the profile or maturity profile of the CRAs and the EUR 10 billion+ that we have been reporting. All in all, I would say on average, these agreements comprise or have a duration of two to three years. First point. Second point is that in terms of customers and also having skin in the game, we are asking those guys to be, so to say, also putting deposits on the table so that it is a shared risk going forward as far as additional investment is concerned. The other issue that comes again and again is the question on pricing. We do not have fixed pricing over the whole duration. Typically, pricing for the devices is negotiated for the next 12 months. That's pretty much how these agreements are structured.
So with that, we believe we are pretty flexible, then also in terms of detailed market developments in the medium to more far-term future.
Brilliant. Thank you so much.
Then the next question comes from Johannes Schaller from Deutsche Bank. Please go ahead.
Yeah, good morning. Two questions on the silicon carbide side. I mean, with 300 mm SiC wafers, that is quite a technology breakthrough, I would say. Jochen, could you maybe give us a bit more detail kind of where the IP resides here between you and your wafer-producing partners? I guess the question I am asking is how easy is it for some of your competitors to get access to 300 mm wafers and process them as well? And maybe give us a bit of an idea kind of when you would go into volume production. And then as a follow-up on the silicon carbide business, you are obviously back to really nice growth this year.
Could you help us understand kind of how much of that is AI versus auto, and talk a little bit more about these auto wins that you have helping your growth as we go into next year? Thank you.
Yeah, Johannes, thanks for your question. I take the first one. So the 300 mm topic, there are very few, but there are some suppliers who offer now the doped 300 mm silicon carbide substrate. The key challenge is really in the processing. Obviously, as always, 300 mm is much more prone to [audio distorted].
Sorry, we seem to have some technical issues. Seems to be a feedback.
Sorry, we have an echo here in the line.
Yes.
Sorry. Can you hear us again?
Yes, we can hear you.
[Non-English content ] Okay. 300 mm, the real issue is the processing, as the wafers are obviously much bigger. There are also no commercial epitaxial tools as of now available, but with our announcement, we expect suppliers to work on that one now. In terms of industrialization, well, we do not have a fixed timeline, but the gallium nitride story can give you sort of reference, where I announced in September 2024 the availability of our first gallium nitride process wafers, and now we announce the industrialization for end of 2027. Whether it's the same line, we don't know yet, but it will take several years to materialize. Right now, we are in full swing of converting towards 200 mm. That should give you an idea.
Johannes, Sven speaking, I think your question was in regard to the silicon carbide growth. 10% growth last year, now we said we grow significantly from around EUR 700 million to significantly above EUR 1 billion. The growth is coming from both divisions, from Automotive and from PS, including, of course, the previous GIP and infrastructure and industrial business. Both are growing and contributing equally to that very steep increase in the fiscal year.
The auto design wins you mentioned, is it fair to assume that some of those are with Chinese OEMs as well?
The one I reported about is a U.S. company. I said a U.S. electric vehicle maker, but of course, we are also active in China, but we are only taking on the business that is really achieving the profitability goals we have set ourselves. So we are selective with respect to Chinese high-voltage business.
Very clear. Thank you.
The next question comes from Janardan Menon from Jefferies. Please go ahead.
Yeah. Hi, good morning. Thanks for taking the question. I just want to go back to the margin guidance for FY 2027. When I look at previous years where things are accelerating, you've always ended up being quite conservative in your guidance, and your final results have significantly exceeded initial guidance. Given that you're saying that the pricing is still being negotiated, and there could be more price increases coming down the line, should we assume that this mid-20s segment margin is a base case, and there could be some upside to that? Just within that, the Automotive high-voltage exit, when exactly does that complete, and from which quarter would you get the jump in margins in Automotive from that exit? Thanks.
Okay, Janardan, I take your questions on the high voltage. As we said, it will be a headwind, although a smaller headwind for the full fiscal year. From next fiscal year onwards, we expect that to be no meaningful negative contribution, neither to Automotive nor to the group. Now to your margin question, which of course we have expected. I think I could spare you the intro, but I do it regardless. We are the first to guide for the full fiscal year. You know how Infineon is guiding. We have changed from a qualitative guidance to a number-based guidance as much as possible back again, given the visibility we are having. If you look at the contributors, puts and takes for the full year profitability level, I can run you through.
Of course, the biggest one, and you asked about upside, that's also linked, is the growth. There is a nice fall-through from the growth. There is a mix effect. If we do more AI business, you can look at the profitability development in the PS division. You see it, so that's a positive one. Andreas talked about positive pricing contribution, although it's not yet concluded. We have idle reduction, and we have the full run rate of the Step Up contributions. All that is contributing. On the flip side, we see significant cost increases and more and more foundry price increases, given that they see the same positive demand trend and are also fully booked. The other one is, you may have seen the D&A number. This one increases from EUR 1.9 billion to EUR 2.2 billion, the biggest contributor, of course, being the Dresden Module 4 ramp.
These are the factors, to be very explicit, which drive the margin. And of course, there could be upside from, as I said, revenue. There could be upside from pricing. There could be upside from the currency. Just maybe to preempt the next question on the 23% segment result margin for the current quarter, just to assume the dollar would not be at 115, but at 112, this would give us for the quarter already 50 basis points of margin tailwind.
Thank you. That's very clear. And just a brief follow-up is on IVR. From which year would you expect? You previously, I think, suggested 2027 in one of your charts, end of 2027. Is that a possibility? And previously, you've talked about two to three times the value when you go from lateral to vertical. When you go from vertical to IVR, is there still another significant jump in content?
IVR is realized in different flavors. Sometimes the conversion inside the substrate is rather a small step, let's say, from 1.56 V to core voltage. That can be handled by FinFETs, so CMOS, two-stack CMOS devices. However, the real benefit of IVR only comes into play if you increase that voltage significantly, and that's where our MOSFETs will play a role. So, there is one customer that is going into IVR, presumably next year. But before we see a wave into IVR, I think a little bit more time will pass by. But the logic behind it, getting closer to the processor with a higher voltage, is the same. And I can confirm that from lateral to vertical, the factor you mentioned is a correct one. Yeah.
IVR will also be a step up, though you may not be able to tell us.
Yeah. If it is FinFET-based, then this last conversion step is done by a, let's say, CMOS technology. We will have to see how it exactly plays out, whether customers will only go, let's say, for a small step in voltage, or whether they will go for a higher voltage, let's say, 3 V and more . In any case, we are very close in this case to the processor, so reliability and quality topics will be very high on the agenda because obviously, you are in danger of throwing away the processor if you do not handle it correctly.
Understood. Thank you.
The next question comes from Joshua Buchalter from TD Cowen. Please go ahead.
Hey, guys. Thank you for taking my question, and congrats on the results. I was hoping to get a little bit more of a breakdown in the EUR 4 billion data center outlook. Any help you can give us on how much is going to be driven by the traditional applications versus AI, or a split between power stage versus stage one or stage two you are willing to give us? Thank you.
Yeah. We expect strong growth definitely in the CPU arena, right? Remember, the agentic is calling for more CPUs, so that I can share. I also would like to remind you of one statement I made before. The outgoing run rate of our AI business, or sorry, of our data center business in Q4 was already close to EUR 800 million. Now, we are predicting EUR 4 billion for the year. So if I keep the run rate, I would be already at EUR 3.2 billion. Again, the total number is subject to output, mix, and pricing. So let's see where we are going to end. With respect to your question on stage two versus stage one, I would expect a similar split like last year. But again, CPU is a strong growth driver as we speak. We have a long, long tradition in that market.
We are in all relevant reference designs, which is important because here, the customer or the supply chain is not necessarily from us to the CPU makers. You have to be on the reference design, but then other companies are assembling that. But that is a traditional stronghold of Infineon, and therefore, clearly a strong contributor to this. In any way, the boundaries will become blurry, as we see also not only agentic CPUs, but the inference part, and that's why we put everything into one number, because it doesn't make really sense to separate.
Got it. Thank you for the color there. Then as we think about your mid-teens ATV revenue growth outlook for the year, any help you can give us on how much of that out-growth versus SAAR is from continued microcontroller share gains? I think you're well above 30% at this point. Is there much more runway you have there? Or is it more from trends you're seeing on the EV side, whether via silicon or silicon carbide, as there's been some renewed optimism in Europe, and China continues to move forward with EVs? Thank you.
So the mid-teens is basically a good number across the board, of course, particularly in compute and connectivity, right? The microcontroller success story continues, and it's now structurally even enforced by Ethernet. Ethernet, of course, has even higher growth rates than the mid-teens, naturally. That's why we bought it. But it also goes all the way to MOSFETs, where again, we are in the position to supply the market with high-quality grade MOSFETs, which is critical for many OEMs. So, really broad growth momentum, not isolated to individual product groups. The HV story, high-voltage story, is a bit of a specific. Here, we will also grow, but also because we moved the onboard charger business from GIP and PS partially, which was there, into Automotive. So that's more of a baseline topic.
Thank you for the color there.
The next question comes from Simon Coles from Barclays. Please go ahead.
Hi, guys. Thanks for taking the question. It is just on constraints. You talk about, I think you said Q1 is a bit supply constrained, and we know AI power is supply constrained, but how do we think about those constraints capping growth through the year? Is the issue with your outsource partners, or is it just specifically how quickly you can ramp Dresden? Then have a follow-up. Thank you.
Yeah, Dresden, we will ramp at double the speed as compared to the past. We said that theoretically, if the demand holds up, we are able to ramp Dresden Module 4 to full capacity within two and a half, three years. This would be a new benchmark, but of course, we also depend on OSAT partners to ramp. We also have the possibility to not only grow ourselves in Malacca, but now also in Bangkok. I think, again, a good decision to keep the construction project going through the downturn, and now we have the benefit out of these options. But there are, as we speak, of course, also quite some bottlenecks which lead to the overall supply limitation situation. But I think compared to the rest of the industry, we are in a very good position, but we are still supply limited.
Okay, thank you. I know you are not giving the backlog anymore, and you said it increased, but is it fair to assume it increased at a similar rate as it did last quarter, or was it a little bit stronger this time? You sort of alluded to it around the order momentum picking up in September, but just wondering for a bit more color there. Thank you.
Yeah, look, the problem with this backlog is we would have to go in much, much, much more detail to make it meaningful for you, meaning, what is the backlog for the first 12, 24, 36 months? In that sense, we would like to stick to this qualitative statement, because to really interpret the backlog, we would have to give you a transparency which we feel is not what we want to do.
But Simon, maybe let me add one more thing to be here very clear. I said it one hour ago in the press conference as well. We are not stopping the disclosure because the number is negative. I can tell you qualitatively, it went up further, so that is not the reason why we stopped communicating the exact number.
No, no, I am sure it did. Thank you. Understand.
Then the next question comes from Lee Simpson from Morgan Stanley. Please go ahead.
Great. Thanks for fitting me in. I am just trying to get a couple of clarification points, if I could. It looked as though Dresden 4 is ramping this year, so there will be some associated startup costs with this. And I think if your comments are to be understood, you are going to see this in D&A, and we have seen this gap up of about EUR 300 million. So, can we say at this point that the startup costs or ramp-up costs at Dresden 4 are of the order of EUR 300 million, and that is associated with a couple of hundred tool sets coming in? And maybe as a follow-up, it looks on the slide deck as though you have increased the BOM for humanoids, I think, to $1,000 .
Just trying to understand what the gap must have been from, I think you had $600 prior, up to that $1,000 mark. Thank you.
Yeah. Hi, Lee. Sven here. I will take the first one. So, the D&A increase year-over-year is EUR 300 million. Probably one half of that is Dresden Module 4 related. The rest comes from other CapEx increases, for example, Kulim, Villach, and others. So half of the EUR 300 million is attributable to Dresden Module 4.
I guess maybe just as follow on to that, I mean, if you are installing, let's say, about 200 or 300 pieces of equipment, you are testing them, you are fitting that in, qualifying, and signing off, there will be associated costs with that beyond just D&A. Does that materialize in SG&A, or where would that come in?
No, you are absolutely right. This is materializing, and this is one of the reasons why we are putting that into our Q1 and Q2 numbers. But it is in the gross margin, of course. It is not SG&A.
Got you. Super. Thank you.
And other robots?
Yeah, Andreas is speaking as far as robots are concerned. So we increased the average number of BOM drivers for this relative to what we said before are pretty much in the area of connectivity. As you know, Ethernet is something where Infineon has moved in strongly with the acquisition of former Marvell business. And with that, we also proliferate into robotics related to us building what we call the nerve system going forward. Within these robots, second element I want to mention is the area of tactile sensing. So around these robots, there's a lot of discussion ongoing on this, quote-unquote, "human-like feeling" at the fingertips that requires quite a sophisticated solution in the area of sensing and control, where we are also in a very good position to be on the frontier of things.
And that, combined with a couple of other minor impacts, led us to increasing the number that we have been given to you. Yeah, and I hand over to Sven.
I think we answered it already. Last question. Next and last question. Okay.
Then we are now taking our last question, which comes from François Bouvignies from UBS. Please go ahead.
Thank you very much. I just wanted to come back on the capacity side quickly. I mean, you guided EUR 20 billion revenues for the full year 2027, which is EUR 3.7 billion incremental revenues. If we assume pricing is up 3%, that is EUR 500 million. That means that you have EUR 3.1 billion of revenues incremental, excluding pricing. If I put Dresden EUR 2 billion roughly, which is your two to three years cadence, that means that you have EUR 1.2 billion incremental revenues, excluding pricing in Dresden, which, of course, you have a bit of inventories you can sell, so a couple of hundreds. That means close to EUR 1 billion incremental revenues outside Dresden of incremental revenues. That seems kind of a big number. I was wondering where the upside you think could come from the capacity point of view.
Because when I look at those numbers, it seems that you are really running at full steam, if you see what I mean.
Yeah. Thanks for the question, François. So your analysis is correct. Please remember, we just finished to equip Villach last year. So you have a full year effect out of the Villach now fully ramped. Then, of course, our share of OSAT and foundry is by now significant. Remember, all the micros, all the connectivity products are externally produced. Then, of course, as Sven mentioned, we reduced our idle costs internally. So we have also more factories like Kulim and related backends of modules, which are now ramping up production. We still have EUR 350 million idle, which I am not happy about. I think in a real peak year, which is probably not this one, but we should have structural idle significantly lower than the EUR 350 million. So we are ramping up through various sources.
Okay, thank you. Maybe on robots, I mean, since you have the BOM increasing, maybe that is a sign of better visibility. When do you think it could become material for Infineon, these humanoid robots? Do you have any visibility on the timing at all?
That is still difficult. I think the number of robots produced in the world are still not that high. Please also remember, the $ 1,000 relates to a humanoid, right? Here at the Electronica, I joined stage with Agile Robots, which is focusing on industrial robots. For example, industrial robots do not have legs. They have a solid ground with a battery. That reduces then somewhat again the semiconductor content. So I think we are still in the early innings of this market and how it plays out. But we are working with all relevant players from east to west, also here in Europe. We are also testing these robots. Again, I talked about that one this morning, Agile Robots, robots in our manufacturing environment. So we see what is possible.
But I think before it moves really the headline of Infineon at a level of EUR 20 billion, I think a couple of years will pass by. I think it is going to be the next wave. The current wave is AI, is power infrastructure, is software-defined vehicle, and very likely then the next growth wave is going to be around robotics.
Thank you.
All right, ladies and gentlemen, that closes our call for today. It is time to wrap up a memorable call in more than one respect. Apologies for the technical hiccups and glitches, but memorable hopefully also refers to the content. We are aware that there is a list of analysts which have not been able to ask a question, so the IR team has its work cut out for this afternoon. For everyone else, take care and have a good time. Speak to you again next time. Bye-bye.