Morning. Why do not we go ahead and get started? Welcome to JPMorgan's 53rd Annual Technology Media and Communications Conference. My name is Harlan Sur. I am the Semiconductor and Semiconductor Capital Equipment Analyst for the firm. I am very pleased to have the NXP Semiconductors team here with us today: Kurt Sievers, Chief Executive Officer; Rafael Sotomayor, President and who will assume the role of CEO in late October, as Kurt has recently announced his retirement. Of course, we have Bill Betz, Executive Vice President and Chief Financial Officer. Here in the front, we have Jeff Palmer, Senior Vice President of Investor Relations, all here with us this morning. Gentlemen, thank you for joining us this morning. Kurt, thank you very much for all the support over the years.
Yeah, maybe, you know, that is a great place to start, Kurt, is maybe you can kick us off with reflecting on your 30-year career at NXP, last five years as CEO, your views on the significant and positive transformation, right, of the company under your leadership. Maybe, Rafael, a brief overview of yourself, the 10 years thus far at NXP, and then we'll go ahead and kick off the Q&A.
Yeah, thanks, Harlan, and thanks for having us. Good morning, everybody. A long journey. Thirty years. When I started thirty years ago, we were a division of Philips Electronics. We took it private in 2006. We IPO'd the company in 2010, 15 years ago.
Yes.
Since that moment, I've been on the executive management team of the company. I think we applied a lot of strategic focus and learned financial discipline, which is not what you have as a division of a large corporate conglomerate, which we used to be before. Maybe to put it in a few simple numbers, since we IPO'd in 2010, we have tripled the revenue of the company. We have more than doubled the operating margin performance of the company, solidly above 30%. Most importantly, we've built a very strong platform to be the number one in bringing intelligence systems to the edge in automotive and industrial going forward, which is where Rafael will take a very, very strong lead.
Over that whole period, I think working on financial discipline and focusing the portfolio for that future in edge compute, in automotive, in industrial, that has been the major transformation which we have been undertaking over that period. You might ask, so why leaving the ship now? I want to be very sure that I can communicate this in all clarity. It's a purely personal decision. I'm at good health. My family is at good health. It is actually for that reason that I want to catch the moment and just recalibrate my life in favor of my family going forward, who has helped me so far very much to do what I could do for NXP. It's not a decision against NXP or any of its stakeholders. It is clearly a decision for me personally.
I'm super proud and super glad that we have Rafael, who indeed was announced as President last week. He will own all of the operations of the company, which means all of the business lines, sales, operations, supply chain, and R&D already for the next six months, which is the transition period before he takes over from me as CEO at the end of October with the Q3 earnings. The board had a very careful and thoughtful consideration about internal versus external succession. Since my retirement is purely my decision, there is no disconnect whatsoever about the strategy, about the capital allocation policy, or the financial model, which we launched in November last year. The board wanted to have a successor who accelerates the execution of that strategy. It's not about disrupting it.
It's not about changing it, but actually accelerating it going forward, which is why we felt an internal successor is the best thing we can do, because there is no time wasted. Since Rafael has co-created that strategy, he has been owning over the past years our industrial business and our mobile business. As such, he's been running, say, 50% of the company. It is going to be very seamless. I am glad we have today another opportunity to face all of you with me and Rafael, which is why I would also like to hand over to you, Rafael, and maybe give a little bit of your perspective for how we're going to double EPS over the next couple of years into 2030 with high single-digit revenue growth and moving the margin solidly above 60%. Over to you.
Let me start by introducing myself, just who I am. I started my career as an R&D in R&D, worked for Motorola and Intel. One thing led to another. I ended up at Broadcom when Broadcom barely started the wireless connectivity in the early 2000s. Over the 10 years of my tenure there, we created an enormous franchise for connectivity for Broadcom. Joining NXP in 2014, I saw a similar kind of trend with contactless transactions, piggybacking on the technology of NFC and security. That is exactly what we have done in the last 10 years since I joined NXP. Created a very big franchise with our mobile wallet and anything associated with contactless. It was a small business when I joined NXP. It is quite a large, really good franchise for NXP. Later and within NXP, I took over the industrial and IoT segment for NXP.
This includes the microcontrollers, microprocessors, connectivity, security, anything associated with delivering a solution to our core markets. I am now very proud and very anxious and excited to kind of continue the legacy that Kurt has created with NXP.
Great. Again, Kurt, thanks for all the support over the past few years. Rafael, I look forward to the leadership at NXP going forward. I'm going to start off with just a couple of the sort of near to midterm focus questions, and we're going to focus on strategy and product execution. The team had their earnings call two weeks ago, the NXP team. Some of your peers in the analog and embedded MCU space have talked about current trends that point to the ending of a cyclical recovery after an industry downturn that started in 2022, turns orders increasing, higher backlog carry of distant customers, stabilization of orders from your direct customers. Many investors argue that this uptick in orders and near-term demand activity is no different than a demand step up due to pull-ins ahead of the expiration of thef 90-day reciprocal tariff reprieve.
We've actually heard from large OEMs during this earnings season that they are indeed pulling in some component inventories near term. We've had some of the compute names like Intel and some of the memory companies that have also talked about seeing tariff-related pull-ins. Maybe take us through the process by which the NXP team goes through to decipher pull-in or potential cyclical-related improvements.
Yeah, so first of all, Harlan, I would absolutely confirm that we are in this crosshair situation of finally, and you know that we called it wrong a year ago, of finally seeing the early innings of what we definitely consider cycle recovery. You mentioned it, that stabilization of the long-term order book from the direct customers. It is the backlog increase of our distribution customers. It is actually some supply escalations, which, by the way, if you would just pull in to increase inventory, you would not necessarily escalate to the CEO for not getting a part. Therefore, I think it is different than pull-ins.
Crosshair on the one side, this, which is something we had been waiting for too long of a time, and at the same time, definitely a significant level of uncertainty created by the tariffs, where I should clearly say the current tariffs, and they changed again over the weekend, they have had and do have no material impact on our P&L. Our guidance is not impacted by that. When we speak about uncertainty from the tariffs, it is the impact on our customers and on the macro. It is not directly about us. That uncertainty, unfortunately, in my view, has also not gone away, not even by the recent announcements, because, mind you, it is a 90-day kind of holding period.
That's right.
Yes, it's been reduced now, which is great between China and the U.S. The only thing it says is that it will be reviewed through the next 90 days, and let's see where it stands then. That inherent uncertainty for our large industrial and automotive customers, that unfortunately hasn't really changed. Those are two competing dynamics where Bill and I said if we had been in the absence of these tariffs, I think now finally the moment would have come where we would have said we increased general inventory. You know, we still spend at nine weeks versus our long-term target of 11 weeks. That was the moment we had been waiting for. Now we decided to not do so because of the uncertainty.
On the pull-in question, I do not think there is any indication that what we called as a positive trend comes from pull-ins because of the tariffs, which is A, because it started mid of Q1, which was before Liberation Day, before all of this. Admittedly, we have not even seen any change either through Liberation Day or through the parking of it for 90 days, which was a week or whatever after Liberation Day. This has not reflected at all in the order patterns and the behavior of our customers. That is just an analytical consideration. There is also a consideration which comes from factual discussions with our customers. I mean, we are sitting down with customers. We try to understand what they do, because you know that we are highly allergic if customers want to increase inventory.
I mean, we have tried to avoid this for a long period, which was our soft landing approach in order to minimize the peak to draft D elta through the past couple of years. We certainly would not want to spoil all of this now by easily giving in to pull-ins or anything like that. We have very distinct discussions with our customers. There are a few who would say, yeah, maybe we have this a little later or have this a little later or earlier, but nothing material at all. That has been the case when we had the earnings call. I would today, two weeks later, make the absolute same statement. The improvement which we have talked about in the order patterns and some early supply escalations, we really think is a sign of a cycle recovery.
Yet you might have seen we've been cautious. I mean, we didn't call the cycle. We didn't say, this is it, and now everything is hooray, because we still believe that that uncertainty from the tariffs is such a massive cross current that we tried to be cautious. We got it wrong last year. We don't want to get it wrong once again.
Perfect. If we look at the value chain side of tariff and trade, geopolitical dynamics, looking at your U.S. wafer manufacturing footprint, three out of your four internal 200-millimeter fabs are based in the U.S. Given the push by the Trump administration for more domestic manufacturing, has that motivated the NXP team to maybe rethink or modify your 200-millimeter consolidation strategy?
A clear no. Absolutely not, because our hybrid manufacturing strategy, which indeed involves consolidating the 200-millimeter assets which we have in favor of moving into 300-millimeter JVs, which we are building with Vanguard in Singapore and with TSMC in Germany, they are so value-creating by reduction of fixed cost, by accessing the lower cost base of 300-millimeter manufacturing in a modern place and in what we consider geographically resilient place when you think about geopolitics. To be very specific, in the case of VSMC, which is the main trigger for the consolidation of the 200-millimeter facilities, this is like a $4 billion revenue opportunity for NXP. If we load it to the extent we can, it will be a $4 billion annual revenue for NXP, which delivers 200 basis points higher gross margin for the whole company.
You will understand that this is top of mind for us. We would rather go earlier out of the 200-millimeter facilities than later, because for us, it is just a significant part of the commitment which we gave at last Investor Day to increase gross margins into a bracket of 57%-63% and eventually higher later on.
Part of the other, I appreciate the answer there. Part of the other dynamic on supply chain diversification strategy that you outlined at the annual stay in November, particularly your China for China strategy, right, Kurt? You mentioned that on the last earnings call that roughly 35% of your revenues that you ship to China, of that amount, roughly 30% is covered by wafers that are manufactured in China. And that number should expand, right? You've got strong foundry partnerships in China with TSMC, SMIC, Grace. What's the team's longer-term plan? China revenue manufacturing as a percent of your total China revenues?
Yeah. First of all, China, very clearly for NXP, is a very important opportunity which we cannot miss. With our focus on our automotive and industrial, we just have to acknowledge the fact that China has become innovation leader, both in industrial electronics as well as in automotive. Not playing in China for us would be a significant issue. Secondly, that induces then the China for China strategy, which in itself has two legs. One is manufacturing. That is the local for local manufacturing, which I will explain in a minute. The other one, by the way, which is more strategic and in the long run even more important in my view, is actually reallocating R&D funds to Chinese lead customers, not treating them as second or third-class customers, but actually take a solid part of the NXP R&D to develop product and solutions which are dedicated to China.
Why? Because they are faster and they have the lead now in the transformation into SDVs, so software-defined vehicles, EVs, but also in industrial electronics when it's about automation and robotics in factories. China is leading in the world. If we have now a priority call where would we spend a dollar R&D for a Western automotive company or for a Chinese one, we spend it for China. That's the strategic. Now back to the question on manufacturing. It is a matter of fact, and that was way before the tariffs. Tariffs are just accelerating this trend. It is a matter of fact that in almost all of the requests for quotation which we get from Chinese customers, the first line in the questionnaire is, can you manufacture the product in China? It's about the front-end manufacturing.
It is not about back-end, which would be easy for us. We have a large facility there anyway. It is about the wafer manufacturing. If you cannot say yes, you are out of the race already. You do not even get further into negotiation about a new product. That is the first line. That is why this is so important for us. Now you quoted some numbers. I want to make it very simple. It is about 17%-18% of NXP's total revenue, which is China for China, which is half of the famous 36%-37%.
That's right.
That 17%, 18%, a third of that is already manufactured in China today. A third of the business, which falls in the category which we are discussing, we have already sourced from Chinese wafer foundries. Those are three. It is SMIC. It is TSMC Nanshan, which is a 16-28 nanometer facility, which is very good for us. That is our sweet spot for microcontrollers and microprocessors. It is increasingly, it is still very small, HC Grace for the future for analog mix signal. We want to increase the number. I will not give you a definitive number, but we want to get it as high as ever possible, because it is very obvious. The little tariff dance over the past couple of weeks just gave good evidence of this. This is a big differentiator for us.
We actually consider this whole situation in the meantime not as a challenge for NXP, but it's actually a competitive differentiator, because our manufacturing strategy, our supply strategy allows us the flexibility to do so, because we are not having huge factories in the home country, in the U.S. or in Europe, which we have to fill. I mean, this whole strategy of hybrid manufacturing helps us to be flexible, to do local for local manufacturing. We are very busy to lift up the third to a higher number. I want to mention again, this isn't that much about cost. I mean, it should be also cost competitive, of course. We are, and Rafael is driving this very hard now with me. We are deeply convinced that sustainable success for a Western semiconductor company in China is about innovation leadership.
It will all be about differentiation and innovation leadership to give them a product which they cannot get from local competitors.
Yeah. On that note, differentiation, innovation, let's focus on your end markets, product leadership, mid to long-term strategy. The team, I thought, delivered a very strong message at your annual stay last year, right, focusing on bringing more system-level solutions to the edge, semiconductors, hardware systems, software services, right, helping customers drive faster time to market, lower cost of ownership, and helping them to deal with the complexity of integrating new features like artificial intelligence and machine language. Particularly within your industrial and IoT segments, I mean, I see this as a pretty significant opportunity for NXP. We all know NXP has a very strong number one position in automotive semiconductors. In industrial, you're a top 10 supplier, number five, number six globally, but still definitely large opportunity to capture more share, right?
Rafael, at annual stay, you took us through the system-level strategy for industrial and IoT, leveraging, I think, something like 3,000 plus MCU and microprocessor SKUs, broad portfolio of analog power management, connectivity capabilities, combined that with strong software stack targeted for large markets like factory automation, building automation, healthcare. Can you just give us an update on the scorecard? What percentage of your current design wins today are more system-level solutions integrating a whole bunch of your connectivity analog products attached to your processor products with integrated software from your partners or developed by NXP?
Yeah. Great question. Thank you for the question. Let's start with the premise, right, that today or in the past, the industrial market was, for the most part, really, I will say, most of the market was covered by analog companies, right? The premise is now, in the future now, starting now, the complexity of the solutions needed in industrial is getting higher and higher to a point that goes beyond the technical capabilities of the OEMs in industrial, beyond their capabilities, despite the fact that they actually need these technologies. The premise is that by delivering system solutions to our customers, we solve that complexity, the complexity of integrating technologies, whether it's connectivity you mentioned, artificial intelligence, security, different types of processing that is required, and then creating an ecosystem around it, creating systems of systems.
That complexity, that's the key by delivering system solutions. We have seen a significant shift since we have adopted this strategy with respect to how we address our customers. Just the conversations with our customers are significantly changing, right? The moment we start talking about solutions to the customers versus a component, the conversation is extremely different. We're starting to see now that our ability to solve a customer problem and our chances to win in the socket go up significantly. Give me some examples that will go into this direction. In the megatrends of electrification, for instance, in which the electrification of everything, we develop a few system solutions in emerging product categories. This is in battery energy storage systems and EV chargers, for instance. In this, we saw an immediate return on business.
We saw our market share in these nascent categories just go up to 80%, close to even 90% in some of these cases, because this first-mover advantage of delivering complete systems to our customers really made a difference. We can see now in these examples that system solution did make an impact. It continues to make an impact in the way we continue to sell these products. We are getting inspiration for what we did there now, and we started developing now. I think we have today about 25 different system solutions on the fly, working with different customers in different areas, whether it's healthcare is one of them, obviously continue to do power and energy, and I will say factory automation.
One of the key metrics that we were going to track, because this is already the case in power and energy, our design wins in a system solution today in the power and energy segment are bigger than component design wins. Think of that as very, very powerful. Design wins in systems today in this particular subsegment of industrial is higher than component design wins. That to me is a sign that, yeah, we're getting some things right. I think we're just going to have to continue with this.
One of your other strategies in industrial and IoT here is to expand your presence in the distribution channel focused on broad market, mass market focus, right? More focus here will definitely help, I think, to drive a higher gross margin profile. Team has been adding more to its family of general purpose high-performance MCUs like your MCX A series family of microcontrollers. In addition to expanding the mass market portfolio, what is the team doing go-to-market-wise to drive solutions to that long tail of small to medium-sized customers?
All right. We'll start with the fact that delivering a system solution to a tier one and not able to take that system solution to the mass market, to the long tail, I would consider that a failure. Because about 75% of the market today is in the long tail. Delivering a system solution for tier one does not really move the needle from a business perspective, does not make an impact in the market, okay? Not doing it, it just kind of basically is not even part of the game plan. It has to be fundamental to actually take every system solution to the long tail and to the mass market. Start with that. Of course, delivering the technical solution is a big part of it, making sure that you pick the right system solution.
You just do not want to address a problem that only 0.0001% of the market really needs. Selecting the right system solution that you can take to the mass market is a key area. At the same time, there are other elements that need to be put in place, right? We start with how do you make your system solution accessible by the market, right? Creating the right collaterals. For instance, the hardware collaterals need to be cost-effective. They need to be modular. Think about snapping into these standard kind of connectors like Legos. Do I snap a connectivity? Do I snap security? The ability to actually make this very modular design is very cost-effective from a hardware perspective is key as well. Training your sales force. What I mean by sales force also includes your distributors, right?
Making sure that they understand how they position our solutions in terms of value proposition and the roadmap for it. You have to have a very strong presence in the digital world, right? You need to have a very strong digital presence with e-tailers. You have to train your ecosystem, design houses, software partners, EBS or embedded board solutions partners. That whole infrastructure in order to kind of take your solutions to the channel is, and that investment, by the way, may take just as much effort and time that it took you technically to develop the system solution. That is part of what the transformation we are doing in terms of the go-to-market.
Yeah. I want to add that the channel position which NXP has is extremely strong. It's more than 50% of our business has been and is running through the channel. We have a head start to what Rafael is saying. I mean, of course, a lot of hard work to be done. I think as compared to many other players who want to get into the same game now, we have a very, very strong starting position because globally we are so deeply embedded with our franchise distribution partners.
Yeah. You have a very strong breadth of portfolio, which is a strong differentiator. Lots of focus on machine learning and AI. At the annual stay, you did talk about more of these machine learning capabilities in the edge and endpoints.
You've been offering integrated neural processing engines, what we call NPUs, for a number of years, well before the pickup and focus on AI, right? These are integrated into your processors, your crossover family, your MCU families. What type of design win traction have you been seeing with embedded NPUs? Are these organically developed NPU engines? Are you licensing them from ARM via their Ethos family of NPUs? What is the strategy here? Of your embedded business, what % is AI and machine learning sort of focused these days?
No. It's interesting because this is now a topic that is kind of being covered. I mean, it's basically obsessed. Our customers are obsessed with where the direction that we're taking. Just to give you a bit on the history, we started early licensing cores. We licensed cores at the time from two different companies. And this was mostly on the microprocessor side. And this is our i.MX family, by the way. We quickly realized that in order for us to actually really have control of our own destiny, we need to develop our own NPUs. We quickly shifted to an internal development. We also, because we wanted to also take this capability into not only microprocessors, but our low-end MCUs. I think the breakthrough will be also in this tiny ML type of applications.
By developing our own NPU, we were quickly able to really scale down the capabilities of AI into product categories that licensing would have never allowed us to do so. At the same time, we started realizing software enablement was key. The ability to actually make algorithms and machine learning algorithms kind of easy to put into the hands of our customers became also clearly apparent to us. We started investing on the entire software ecosystem of enabling these. Think about an AI SDK. The market shifts quickly from CNN models to transformer models that is GenAI. There is this huge investment in the cloud to develop all sorts of kind of AI models, large language models, which if they're properly sized and properly adjusted, they could make an enormous impact into edge devices.
We quickly kind of pivoted to say, OK, we go from machine learning to GenAI. We evaluated this buy versus build. We thought we evaluated, and this is how we ended up with doing the acquisition of soon-to-be finished acquisition of Kinara, which this is a leading company developing GenAI for the edge, where I guess the ratio of power versus performance, which is ultimate what we want to do. That tells you that NXP is positioning ourselves, we're positioning ourselves as to be the premier provider of edge AI solutions for sure, from MCUs all the way to high-end microprocessors.
Perfect.
Edge AI is so needed by especially industrial, but also automotive customers for privacy and latency reasons. They do not want to get their data into the cloud because their whole asset is, for example, the manufacturing data of a factory. In order to optimize that, which is a big deal for them from a profitability perspective, they want to do it on the premise, not in the cloud. That is why edge AI is so important. Secondly, most of these edge applications are real-time applications. You cannot achieve real-time performance if you have the latency of going forth and back to the cloud. Those two reasons, latency and privacy, are the strong driver for edge AI, which is where the Kinara NPU acquisition Rafael talked about falls into. We think this is redefining the future of industrial electronics.
That is also the reason why we have a high conviction to claim a much bigger space. You talked about our number five or six position today, a much bigger space from a market share perspective in industry going forward. Not against analog competitors. They are very strong. We cannot displace them where they are, but tapping into the opportunity of the higher edge compute requirements in industrial.
Perfect. Before I move on to automotive, does anybody have any questions in the audience? If you do, raise your hands and we will get a mic over to you. Anyone? OK, let's turn over to the automotive segment of your business. Automotive down 4% in calendar 2024, slightly lower than total auto production at flat.
The team was, like many of your peers, under shipping consumption for the better part of last year as the tier one auto OEMs continued to drive their internal levels of inventory lower. Your accelerated growth drivers, S32 software-defined vehicle, radar, electrification, connectivity as a group grew year- over- year last year. If you hit your guidance for the June quarter on auto revenues, your auto business will be down 3%-4% year- over- year through the first half of this year. Are the four accelerated product categories still growing in the first half of this year? How do you see that playing out for full calendar year 2025?
Do you want to go?
Yeah.
No. Actually, the thesis of the growth drivers continues to be untouched, right? We continue to see a tremendous adoption of the technology that we're developing, these growth drivers. Software-defined vehicle is driving the adoption of our new NPUs and new MCUs in automotive, our K5s and our S32Ns. These are unique products that NXP has developed and processed, know that nobody today, any of our competitors have in the market. Obviously, the uniqueness of the products and the shift towards the software-defined vehicle is truly driving. These software-defined vehicles also are very well connected. The connectivity that we have at NXP, whether it's Wi-Fi, whether it's UWB, whether it's BLE, all these kind of connectivities are being adopted in these new architectures. We are quite bullish on what's happening there. The growth drivers are exactly matching what we said we're going to do.
That's the case across the company. All the growth drivers are on track to growth. I think within automotive, it's important to state the trend change. Quarter two guide, for the first time after a long period, is flat year on year. We've been declining year on year for many, many quarters. Q2 guide automotive for the first time is now flat year on year, which really marks a change in direction.
Yeah. Within automotive, I mean, the exemplification of intelligent systems to the edge, right? A big focus at the annual stay in terms of driving that was the focus on your CoreR ide software-defined vehicle platform, leveraging strong platform across MCU, crossover MCU, full-blown embedded processors with partner NXP developed software, middleware library of application software. Can you quantify the traction that you've had so far with CoreR ide SDV reference platforms?
When do you expect CoreR ide-based wins to start to ramp?
I think we're seeing it's very similar to the discussion we have about systems for industrial, right? The conversations that we're now having with our customers are very different. I'll tell you that the S32 CoreR ide in terms of a platform is the vehicle in which we engage in terms of what is truly the need of the different OEMs. That allows us to have a more conversation so you can tailor the value proposition of the platform for each one of these OEMs. It disconnects our S32N, which is a high-performance central compute platform, our Sonos products like K5 and above, and some of the switches at the end, right? The ability to connect really the entire think about the nervous system of the car, what makes the car run.
We expect probably design wins to start actually having a commercial impact into NXP revenue sometime in 2028. This is when I think that's what it's going to take us between now and then to actually be able to deploy some of the things that we're doing. One way to accelerate this was the acquisition of TTTech. This is over 1,000 engineers with middleware, functional safety at the application level, deep customer intimacy in several OEMs, right? This kind of puts NXP in a completely new level with respect to system solutions.
So many more questions. Not enough time. Really appreciate the team's participation today. Thank you very much.
Thank you.
Thanks, Harlan.
Thank you very much.
Thank you.
Yes. Thank you.