Good morning, everybody. Thank you all for joining. My name is Jakob Bluestone. I run the European Tech Hardware team at BNP. Obviously great to see so many people join, and particularly also thank you very much to Jean-Marc and Remi for joining us today from ST. Probably don't need any introduction, but anyway, Jean-Marc's obviously CEO. Remi runs particularly the MCU side of the business as well. Thank you very much both for joining today. I should say as well, thank you for doing a guidance upgrade this morning, which is extremely well-timed. I think maybe we'll start off with the guidance upgrade. Can you maybe just talk us through what prompted it, and maybe also a little bit more broadly, where do you see the big opportunities for ST when it comes to AI data centers? Well, I will start to explain why this morning we have given an update about the guidance moving from nicely above $500 million for this year to about $1 billion. Why? Simply because, first of all, it's absolutely not a demand limitation factor. Demand is really increasing like hell on data center, especially for ST, related to what we call the network flow, the connectivity flow. Means the data transfer inside the rack or rack to rack or outside the rack. Remi will explain later. Not limited by the demand, but limited at a certain moment by our confidence level to execute the manufacturing ramp-up because this demand came pretty fast, okay, end of last year in Q1. Now I can confirm to you that our confidence level to deliver this revenue is very high. Our manufacturing is performing exactly at the expected level. We have removed bottleneck, that the reason why I confirm $1 billion, okay, about $ billion this year. Taking into account the dynamic of the engagement of the customer, and the order entry we have, we will double this revenue at least next year. That the reason why we have updated now. Well, about data center overall, before Remi go little bit more in detail. Basically, you have three opportunity for ST to compete in data center and AI factory. Well, the first one is a classic one, I have to say. It is what is related to the thermal flow infrastructure of data center. Well, it is an industrial market, I have to say, where you have power devices, analog devices, and microcontroller. Here we compete exactly, okay, at the same level that our main competitor, Infineon, on the power guy, I have to say. The second flow where we can compete is what we call the power flows. To make it simple, you have to move from the grid, 20,000 V, your electronic stage to be capable to supply the GPU at 1 V, 1,000 amps, very stable. If you can do it on various step where you have a pretty huge bit of material in electronics. Here ST compete very well everywhere from 20,000 V up to 50 V, I have to say. Later on, okay, we are not yet present. Why? Because this is simply product we have not developed in the past, focusing more on automotive, but we have the capability to do it. This will come. This will come, and especially we have decided this will come when the architecture of the power supply will change, moving from the traditional 20,000 V, 450 V, still alternative, to 20,000 V to 800 V direct current. When we will move this, okay, ST will be there. The third mainstream is what we call the network flow, so the data transfer. Here, clearly there is a major disruption, using technology which are optical cable, and Remi will give more accuracy. At ST, we have decided to invest few years ago in the critical technology to enable this optical cable. This is paying back now and with a trend which is huge. This is the three mainstream where the company is playing. Clearly this year, if we represent again $1 billion revenue, double next year, we have already the visibility for 2028. Of course, this will be a key growth driver to put ST at $18 billion for 2028 and to go to our ambition to $20 billion. Maybe, Remi, you can give a little bit more accuracy, okay, on maybe optical cable, if you want. Sure. Today, the market for optical is centered around pluggable optics. It's used three ways. One, to connect data centers between each other. The other way is to connect racks within a data center between each other. The third one that is going to come in the back end of 2027 will be to connect GPU together. Each time around, it has been happening as a replacement of copper. At this stage, you are going to see and expect non-linear growth when it comes to optics as a function of gigawatts being deployed. It's non-linear because it's way more than what I call scale across data center to data center or scale-out rack to rack. It is now going to go and become scale up, which is about connecting GPU to each other. The reason why when we look at the, you see a lot of excitement around optical and photonics because of those dynamics. In the past, anything that was optical was not based on silicon photonics. It was based on EML technology or even VCSEL technology. Laser, yeah. Now it's actually changing because of the throughput we are talking about. The industry has moved over years from 25 Gb to 50 Gb to 100 Gb, 200 Gb, 400 Gb, 800 Gb, 1.6 Tb ps now, and it will go even faster moving forward. As it grows and it goes faster and faster, silicon photonics is becoming the predominant technology to be used. All of that combined is what's creating this huge acceleration when it comes to photonics, which like Jean-Marc described, is really the essence of who we are in ST. It is pure material science. It's coming with a lot of tricks on how to get to appropriate performance and to yield it at pretty high yield. What we should add as well is ST is the first 300 mm wafer native silicon photonics player. We don't have any legacy. We deploy that straight into a 300 mm fab in Crolles, France. That give us a huge advantage, not only in term of capacity, but also in term of yield and quality of our products. Maybe if I can just pick up on that last point. As is very clear, there's massive growth potential from silicon photonics, and maybe if I can get you just to expand a little bit around the particular competitive advantages you have, how you see the current state of competition for silicon photonics, how you see your market share evolving within silicon photonics, and maybe just sort of help understand a little bit more what is it that makes ST's position so unique around the cost and the sort of yields. Yeah. I'm looking at my team because they ask me that I need to be clear and simple and not getting into technology or details. I will work backwards from your question. The construct we have, the way we build our technology is giving a lot of advantage on the way we manage the waveguide, where the photons are passing through. We have also the ability to do integration between what is being transmitted versus what is being received, in a fashion that is unique and only us can do. We have also the ability, to Jakob's question, to do advanced integration, because when you look, when you double-click on those optical modules, there is really three key components. A microcontroller that is managing different aspects of the optical connection, what we call the EIC, which electronic IC, that take care of some part of the problem, and then a PIC, which stands for photonics IC, that takes care of the photonics, the optical transmission or reception. As you move more and more towards integration and what we call near-package optics, where the optical element or engine is very close to the processor, then you need a lot of tight integration because space is becoming a constraint. Here, the fact that we mastering all the elements, that we acting nearly as a one-stop shop for this market is giving us a lot of capability when it comes to advanced packaging to put things together in a very compact fashion for our customer. That is, I think, one of the big advantage. Now, to Jakob's question on market share, and this is what I was saying earlier in different session. In all humility, we are going to be the growth story of that market simply because of the fact that we are starting this year with 5% market share. Our ability to get to 30% is really in sight. We see how to get there. We know of the engagements to get there. We have the backing of the biggest hyperscaler to get there. We really believe that actually you will see a lot of growth in that business from ST moving forward, which will be backed up by also capacity investment that we can do in a very efficient manner because we can continue to extend our Crolles factory, and keep adding capacity without asking our customers to re-qualify because we are forcing them to go into another fab. That's also a huge advantage for us. We started the discussion talking about you raising guidance as you got more comfortable around supply constraints. Can you maybe just help us understand, given there are so many bottlenecks right now, where is it you have incremental capacity across the world, across your different facilities, in terms of being able to deliver more? No, the capacity limitation for wafer fab, we will have not. Because, again, we are playing with our main 300 mm fab Crolles. Where today we have already built an infrastructure that can support the business for Remi, for let's say 2026, 2027, and we start with 2028. Now what we have to decide, if what he explained, the change in technology going to what they call near-package optics. To have the optical engine closer the processor, this will be a boom because it will boost the addressable market of ST. Here we have to decide between now and the end of the year, additional expansion in Crolles. Well, this is most likely that what we will do. The other limitation factor is more on the packaging assembly, because in ST, we have decided to be 60/40, 60 internal, 40 external. When we are external is because our dimension of scale is not big enough to be competitive. Here we have to pay attention. It is a technology, what we call a wafer level chip scale package. Here we have long-term agreement with supplier, and this is something where we have to pay attention. Last but not the least, it is what we call a 3D integration, where you stack some chips and so on and so forth. First of all, we have the experience with another big customer, so on the optical solution. It is our main customer. Second, we have decided to close the 200 mm fab in Crolles to make space available for this 3D integration. The company has the capacity really to support this goal. ST will not be limited by our capacity in the next three years, on data center. Okay. We have to pay attention on subcontractor. Let's turn to another hot topic, which is space. There's obviously an IPO going on. I guess if you could maybe just help us, first of all, just if you can maybe just talk us through what is it ST is providing. How do you feed into the supply chain for LEO satellite? Then we can maybe sort of dive a little bit into the outlook for that business. Well, Remi will give the detail, but what I would like to say is that first, this is a business in which we are investing since 2014. It's not something coming from the sky by opportunity. Okay, in 2015, Mr. Musk visited Crolles, and we decided to go together to make this business. At this point of time, okay, no satellite, low Earth, no terminal, no gateway, nothing. Now it's million of user, thousand of satellite. It is a legacy for ST, and we have a great experience, in the technology, both the chips but the package that are enabling this kind of activity. I repeat, for us, it will be between above $3 billion cumulatively in the next three years. Today, we have 90% of market share in this business. With a growing volume, naturally, you will have other source coming because, okay, the implication are very important, but it is a main growth driver of ST. Now I give Remi to comment more in detail. In terms of products coverage, like Jean-Marc was explaining, we have products in the user terminal, in the satellite, in the gateways. It's both on the digital side as much as it is on the RF side. The way those things work is that, you have in the user terminal many small antenna elements that essentially gang together to create a very strong signal because you need to be able to reach something that is 500 k up in the sky, okay. Which is more or less a distance to a LEO satellite. To do that, there is some phenomenon called beamforming, and in the beamforming in the user terminal, you end up having a modem that is doing the modulation or demodulation. You are having an element that is creating the beamforming, and more importantly, you have all those antenna elements, which are active elements that is essentially are tuned together. We'll skip the details, but tuned together to essentially create that strong signal. Those elements needs to be able to create that strong signal, but as you can imagine, as importantly, they need to be able to receive the strong signal from the same satellite situated at 500 km away. There is a technology that has been proven to be best in class for that. It's actually a BiCMOS technology, which is essentially a mix between bipolar technology, silicon germanium and CMOS technology. That technology end up having the best RF performance in term of noise figure, to be able to receive those signals, but also transmit them very efficiently. This has been the basis of the foundation of our business. The other foundation of our business has been our FD-SOI process. Because when you're in space, especially for the satellite parts, when you're in space, you are subject to alpha particles radiation. Heat cannot be evaporated by convection, because it's a vacuum out there. Okay? Actually our technology, our FD-SOI technology, has been proven to be super efficient for space-based application. To give you a little bit of the lay of the land, we look at this market in 2028 to be a $ 2 billion SAM, and in 2030, a $3 billion SAM, and I want to make sure I'm qualifying what is behind the SAM. When we talk about those numbers, it's only for the broadband part, so internet access part. Keep in mind, there is 3 billion people in the planet that are not within high broadband fiber optics reach today. It's 3 billion users that could be targeted like this. What we are not including in that SAM, which is most likely in the next question of Jakob, is everything around orbital data centers, because this is something that we are very much involved with, but have not been able to scope properly yet. My next question was indeed going to be data centers in space. Instead, maybe just to sort of help us think about this business, is the driver predominantly the terminals on the ground, or is it the satellites being launched? I guess from what you're saying, it's a mixture of both, right? It's particularly the terminals, the sort of key driver? I think that actually, just to give the rough numbers, looking at our flagship customer, they have deployed 10,000 satellites. As you have seen from their S-1 IPO filing, they have more than 10 million active users. The ratios are not the same. It is actually technically three orders of magnitude. Even though it is true that the content is also actually far more dollar rich in a satellite than it is actually in a user terminal. Overall, if we look at our business, it's overly dimensioned by user terminals. Mm-hmm. Jean-Marc, I think you mentioned that we are seeing some second source, and I guess the fact that it's 90% share, there's already some second source in there. Just can you maybe help us, how do we think about that as a headwind? Is it something people should be worried about or not, in terms of the limitations on the growth of your satellite business? No, again, when you have spoken about the magnitude of this market, to have other source coming is business as usual. At ST, we have been a pathfinder company, thanks to the vision of the team and our technology differentiator. This is a major growth driver for ST in the future. Competition is healthy. What I can say more? We don't intend to keep 90% of market share in this business. What we target, generally speaking, okay, with the team, is to have 30% and above market share. Here, yes, you have pricing power, you have the scale to finance the R&D, you have the relationship customer, you can drive your innovation. This is what we are targeting. 90% market share, okay, we don't target this. For us, it's absolutely a no-brainer. Are there opportunities for additional content growth in space? Well, as Remi said, the first layer is broadband. The second layer will be the direct-to-cell telecommunication, and the third layer will be the data center in the space. Here, the data center in the space, you have many paths, okay, to enable it. We are playing, okay, on the value space. Now it's too early to say. ST, we will have material market share in the first layer. Thanks to the relation we have with our main customer, we do believe we will have important market share in the second layer. Now we are deploying a data center on Earth. silicon photonics is really important on data center on Earth. Will be more important in the space. Why? Because of the weight. You cannot put copper into space, okay? 1 k costs a fortune. We have the ingredient to continue this journey. At the end, okay, our strategy is to have a business which will well balance with infrastructure and device. Whatever is the device, consumer, automotive, industrial. This is where we have adjust the trajectory of ST since two years. Excellent. Let's maybe turn to some of the other parts of the business as well. Analog stocks generally have seen a massive rerating. Some of it's about AI, some of it's also about the recovery in industrial. Maybe if we can sort of turn the discussion towards some of these other parts of the business. Can you maybe first of all just share how do you see the sort of non-AI industrial outlook currently? Then maybe we can touch on autos after as well. Well, consumer markets or digital personal electronics, it is clear that the market is pretty soft. Thanks to our exposure to our main customer that everybody knows that is doing well, we expect to grow in 2026 low to mid-single digit. The visibility we have is, as always said, is three to five years. Clearly I can confirm personal electronic for ST will be a growth driver for the company and activity. Industrial market, it's the same. I have to say today what is related, infrastructure, energy, renewable energy, power conversion, grids is really doing well. What is really for us a great opportunity, because decision we have taken in China. Why? Because China is moving their economy more to industrial than buildings and, let's say, this kind of infrastructure. They are great in robotics, they are great in drone, they are great in car industry, ST is there with infrastructure. Industrial is very promising in China, in Asia, coming in Europe as well. This year we expect to grow a solid 30%. Automotive. Well, first of all, I like automotive, you need to like your customer to make business. The good news in automotive is Europe. Europe is one of the good driver despite the difficulties this industry and the challenge this industry is facing. We have good news in Europe on the electrification of the mobility. There is a car maker I cannot mention that are really doing very well with model of car that has really triggering a good appetite for customers. Europe is doing very well. China, despite, okay, some cool down of what happened during the past two, three years, is doing well for us. Overall okay, we can expect to grow high single-digit, low double-digit on automotive. Also, you know that we have acquired recently the MEMS that is mainly an automotive market for us from NXP. The forecast we receive from the Tier 1 is pretty stable now, so is a good sign that inventory has been cleared and digested. Clearly, the industry has not completed its overall reshaping between hybrid car, electrical car, thermal combustion engine. Thanks to our capability to address this market, thanks to address it in a competitive manner, having completed our reshaping plan, it will be a growth driver for us. Let's start by this year, high single-digit, low double-digit. Very clear. Can you maybe also just provide an update on your JV in China? Which I think you're sort of ramping this year, and what are some of the benefits that this brings? The JV in China is participating, well, first of all, to our strategy, one product and capability for ST to offer to supply chain. Why? Because you will have, in the near future, Western companies wanting to make business in the Western world, but in China. In China, they want to have an infrastructure to address China. You will have big player in China that want to play in Western world, so they want to have a Western infrastructure. ST is preparing to address it, and the JV is contributing to that. Second reason, we have to acknowledge that with the change of strategy of the economy in China focusing on digital industry and electrification and so on, there is a very strong innovation in China on car, on robotics, on humanoid, on drone, on automation, on power conversion. We want to drive our own innovation in Europe to be present in China. That is the reason why we have set up this JV. More short term, it is clear that on silicon carbide, which is the first target for the JV, the market in China is really challenging because you have a competitor that are vertically integrated, like a BYD or Huawei. Well, it is clear that the price has been put under pressure in China on silicon carbide, but we are very happy with the JV because [further] are growing very fast. Now we have the state-of-the-art technology in term of SiC, and we will start to make the ramp-up on time, starting this year, second half. Very clear. I've got one more question, then I'll start turning to some of the audience questions that have been coming in. Maybe if we can just turn to your gross margin. You've got a medium-term gross margin target of 45%. Can you maybe just sort of talk us through how do we think about the evolution of gross margin from here as revenues ramp up? The product mix changes. You've obviously had FX change as well since your CMD back in 2024. Just how do we think about gross margin trajectory from here? No, I think a high-level milestone for us is, first of all, to come back as soon as we can, $ 4 billion run rate per quarter. This could happen in the next few quarters, I have to say. The second milestone is to complete our reshaping plan. Doing this at the current exchange rate, we will go back above 40%. 40%-41% gross margin is step number one. The step number two, we confirm that at $18 billion, our model that was 44%-45%, at exchange rate, which was more favorable, 1.09, we confirm the model with the current exchange rate. We have to watch carefully the dynamic of the business on data center because it's really accretive on our gross margin. I confirm the model $18 billion, 40%-45% at exchange rate 1.09. More 43.5%-44.5% at 1.16, 1.17. An intermediate milestone is $16 billion yearly run rate and the reshaping plan completed. That will bring ST back to above 40%. This is what I confirm. I'll start turning to some of the audience questions. You should have the instructions if we maybe put them back on the screen. There are already some questions coming in. Maybe if I can turn back to data center. The question, which I think is an important one, is what gives you conviction that data center demand remains structural and that we're not approaching a peak in AI-driven CapEx? First of all, okay, the way we manage, I repeat, okay, today we have three years engagement. It's related to the capital expenditure that has been announced by hyperscaler, by LLM maker. Here we are well-covered, okay, clearly. We are installing the capacity that can support this business at this stage. Again, I repeat, if it go above a b illion, our manufacturing site can support because we have the flexibility. Then after, okay, to explain you that this time the semiconductor market is entering in an hyper cycle, I will not do it. Okay? There is so many complex factor, okay. You know that the challenge that many company has now is to transform the AI deployment in term of breakthrough, in term of capability to design faster product, capability to improve our productivity in business processes, our capability to make much more predictable our quality, our reliability in manufacturing. You have many opportunity to create value, and this is what is important now. Short term, medium term, what we know is the CapEx spend by our customers, and ST is present to support this CapEx. ST has a strategic position to support this CapEx, and ST has a roadmap to continue to offer differentiation to our customer. That matter. Then, okay, mid, long term, okay, the journey of AI is just the start. Would the semiconductor be a cycling industry? The answer is yes. When you think around your own capital allocation in that sort of context, how do you think about things like CapEx? What's the right level of CapEx to be deploying from your side? By today, we are disciplined, okay, to invest what we have as a depreciation. Until we will reach $18 billion and until we will deliver our model of gross margin, okay, we put under strict control our depreciation to control our costs, okay? Clearly, this is a discipline that we have in the company. We can execute this discipline because we have the flexibility to move out some technology to make capacity available for the growing technology. Is point one. The point two is integrated design manufacturer like us. We know that in average, 15% as a CapEx to sales ratio, in average, across the cycle is something we can sustain. Maybe if we can also just touch on some other topics. Robotics. How do you see that as a growth opportunity for ST? Where do you sort of fit in? Is it particularly on the sensor side? Is it more about power? How should we think about that as a sort of medium or term opportunity for you? Overall, after I will let Remi. Overall, when you see the humanoid robotics architecture, you have different system. Of course, you have the central CPU, but then you have zonal architecture. What we are doing, we are assessing all the subsystem you have, and everywhere, you have a motor control, MCU, power switch, driver. This is not a surprise. Where I want to repeat, ST is a unique semiconductor company having all the technology level is a sensing. Okay, the sensing in a humanoid robots will be the key differentiating factor. The way you take a glass is not with the imaging sensor. It's really with the sensing you have. This system is a very complex system, and we have the knowhow to develop the right product. There is no other company that will be capable to offer sensor, microcontroller, driver, switches to customer. That is the reason why we have this agreement with NVIDIA to try to modelize the humanoid in their data center, giving our own model to train the humanoid because we have not real data to train them. ST will be present everywhere. Is it for tomorrow? No. It is more for the head of the current decade, but clearly it is what we call the physical AI will be the augmented business opportunity for ST after, let's say, the agentic AI that we are addressing now. Remi, you want to add something? No. Maybe the only thing I would add is you have to keep in mind that between us engaging into a new class of products or technology, the time between we start and the technologies that scale on the market is within three years. Today we are having actually really structuring discussions that are dimensioning what an MCU should look like because it's not really an industrial MCU, it's not really an automotive MCU. It's kind of a hybrid of them. What also our gate driver should look like, what type of performance our gate drivers needs to deliver for the switch also, what will be the key criteria for optimized switch. As you can imagine, if you look at our hands, and if each of these section here in your finger end up being actually an MCU with a gate driver and a switch, and together with a motor, integration will become super important. We are having a lot of discussion now in terms of anything that will make those products a reality that I think will place us in a great shape. Jean-Marc, Remi, we could go on for ages, but I'm afraid we've reached the end of our session. A big thank you to both of you for a really interesting presentation.