NXP Semiconductors N.V. (NXPI) Earnings Call Transcript & Summary

So I'm Joe Moore, Morgan Stanley semiconductor team. I'm very happy to have with us today the CEO of NXP, Rafael Sotomayor and CFO, Bill Betz. Thank you, guys. So maybe just talk generally, we've had the CEO transition. How you're thinking about the business any differently than the way you thought about it in the past, your priorities for the next year, start with that.

No, I think that we get very excited about the current trajectory that we have of bringing intelligent systems to the edge. I think it's -- it gets even -- now is even the things that are happening in the industry, it gives us more conviction that we're on the right path and just accelerating what we have already on the pipe.

Yes, yes. Okay. You had a pretty strong Q4 result, pretty good Q1 outlook. You talked about things being better than expected 90 days earlier. Can you just talk generally about the industry environment that you're seeing now?

No, indeed, I think Q1, the outlook that we gave in Q1 is -- the performance of Q1 is definitely more robust than we thought about 90 days ago. And that's kind of driven by visibility, right? The current backlog that we have in direct and the channel strengthens. And if you look at the recovery that we gave from a Q1 perspective, it was broad-based. All segments are improving year-on-year. And so that's quite healthy for us. Now if you look at under the hood, I guess there are 3 main drivers you would think of in auto, you can see that the normalization of inventory is healthy, is good. So now we're starting to ship to end demand and now the story of content growth, which is a real story in auto is starting to kind of drive top line growth. Industrial, you see broad-based recovery, and that is truly broad-based. And what we see in industrial is a bit more of a, I would say, kind of a flight to safety -- flight to quality, I would say, in industrial. You can see the leadership that we have in processing is starting to take -- kind of take share of some of the recovery demand that we have in the segment. And the last one is company-specific product -- project ramps. We have NXP-specific ramps in auto and industrial and mobile, and that we're executing to that. So backlog has strengthened and some of the company-specific drivers starting to kind of really perform.

Yes. That makes sense. You sort of talked about not seeing or anticipating any kind of restocking in customer inventories. And I see the same thing, but I'm sort of surprised by it. You've had this kind of next period situation, which had people wind down fairly quickly, kind of resolved pretty quickly, too, but like it sort of showed how little inventory there is. There's like angst on DDR4 memory going into cars. Don't those kinds of supply situations trigger the feeling like, hey, we probably need a little bit more inventory to deal with all of this.

It may and it should, but we're operating on the assumptions that the current demand environment is the new normal, right? So just to be clear, we have no broad-based restocking in any of our outlooks. And by the way, we don't depend on it either, okay? So it's the demand environment, I mean we -- it's also important to also acknowledge that the auto supply chain is kind of difficult to really kind of analyze, it's complex, and it varies by Tier 1, by OEMs. But we are -- at the aggregate level, we are operating, I think we consider deemed the inventory levels to be below our manufacturing cycle time. And so that kind of makes it very tight and potentially sensitive to disruption in demand. But right now -- I mean, this is how companies are managing their working capital. And the good thing is that we're now shipping to end demand. And so that gives us transparency. It gives predictability. But yes, we don't see broad-based recovery. We will see it if it happens. And if it happens, we will see it as a consequence of improving fundamentals. But the way we see it, we will see it in our orders direct. We will see it in our channel backlog. We will see it on sell-through in the channel. We will see it in spike in short orders. But regardless, we kind of remain disciplined. We basically plan for what we see. And if the signal says something else, we will adjust accordingly.

Great. So within autos, you're shipping to end demand and you're starting to see content increases matter. Can you talk generally about the state of that market? And I guess you talked about the complexity. It seems like regionally, there are a lot of differences. The Chinese market is sort of off sync with other parts of the world timing-wise. Can you just talk about the general enthusiasm that you have for automotive as an end market?

Well, let's start with the fact that 2025 was a transition year, right? It was clouded by inventory digestion, primarily coming from Western Tier 1s. And now with that inventory digestion kind of behind us, you can now see that the true story in auto, which is content growth driven by this transition to software-defined vehicles is actually emerging. And this is actually our play. Our play is on that transition to new architectures to higher compute to digitization of the vehicle. And I think this is universal. It's happening in China. It's happening here in the U.S. It's happening in Europe. And this is kind of a better way to actually make a car. And we're all in the core of the NXP strategy is around software-defined vehicle and the processing unit that goes around the different type of architectures. And that's what is making the transition of NXP to a component player to a strategic kind of a strategic vendor for the OEMs since this is kind of a big investment for them in terms of software, in terms of architectures.

So you grew in those accelerated growth drivers, I think, 10% last year. Can you talk about software-defined vehicles specifically and sort of how does that get provisioned? Where are you in kind of the S-curve of adoption with software-defined vehicles?

Great question. So just to get into a little bit on the growth drivers last year, it was essentially flat, maybe slightly declined, but essentially flat year-on-year for auto for NXP. And the growth drivers in the year grew 10%, which is slightly below our model, but they grew in a very tough environment in an environment that was all about kind of inventory digestion. So clearly, kind of basically points to not only the strength of the trend towards software-defined vehicles but also points to the strength of our portfolio target to software-defined vehicles. We're nascent. We're basically in the beginnings of transition to software-defined vehicles and platforms as all OEMs are in different stages in this journey. The way you should think about our strategy from an NXP perspective is software-defined vehicle is the core. Think about a hub and the rest are somewhat attached to the platform, think about like the spokes. The stronger the software-defined vehicle platform is, the stronger your attach story is going to be from a system perspective. And that's the way we approach it. We know there's not a one-size-fits-all for software-defined vehicles. It goes from zone all the way to central compute. But our road map and our technology investment is geared towards higher-performance compute, where all these compute devices are going to basically incorporate what it used to be all this kind of a myriad of ECUs around the nodes. They're going to basically integrate them through high-performance compute and heterogeneous compute. And you need a particular flavor if you're going zonal, you need a particular flavor if you're in central compute. And we made our investments where the industry is going to go towards these high-performance compute platforms.

And where do you see the competition coming from? Is it sort of traditional microcontroller competitors moving in that direction? Is it people doing infotainment sort of centralized CPUs that are trying to incorporate some of those software-defined vehicles into their software like Qualcomm is trying to do? Just how do you see that competitive dynamic? I know your position is fairly unique.

The way we look at automotive from a vehicle architecture perspective, think about it -- and I'm just going to be a bit kind of a primitive in the way I mentioned it just to kind of basically simplify the concept. But think about 3 domains. Think about the infotainment domain, your cockpit. Think about ADAS, which is the big brain that is making decisions, how to -- basically make decisions how the car drives. And then you think about the vehicle core function of the car, right, the vehicle core functions. These are real-time, low latency, high security, high functional safety, high redundancy. The 3 of them, you could make an argument that they're quite distinct from one another, okay? Today, the core function of the vehicle is very specific, and it requires a certain domain expertise and different certain skill sets. And this is where we see today as the legacy players challenge that we have in this market is that this particular function requires higher performance compute, more heterogeneous compute, more cores, more advanced processing nodes. And so I think this is where the advantage of NXP is that we invested way ahead of the competition in 5 nanometers. We're the only company right now with an automotive-grade processor in 5-nanometer. So it's not only one. We have several ones now. And so I think this is where the market is going. And I think we saw that about 6, 7 years ago to actually invest ahead and just be ready for the future demand of the products.

And maybe, Joe, maybe I'll remind everyone, this is our largest accelerated growth driver. In 2021, it was about $500 million, in 2024, it surpassed $1 billion, and we expect it to double in 2027 to $2 billion. The traction we see, as mentioned before, the first half of 2025, it grew in the high single digits -- I'm sorry, in the low -- high single digits, but the second half accelerated to the high double digits. And so that came across very good. We see that trend continuing. That's well on track to grow at those growth rates that we shared during Analyst Day for SDVs.

Very helpful. Can you talk about maybe competing priorities within the automakers between EV power? And I know you're also focused on that as a growth driver as well, doing battery management, but also ADAS and things like that. It seems like those functions have kind of gone back and forth in importance. Are we now prioritizing both? And are people taking -- I think feel like to really do software-defined vehicles, you sort of have to top down, redesign the whole ecosystem of the car. Are those investments happening at your customers now?

For sure. But I think you touched several things. You talk about EVs, and you talk about SDVs. I think the megatrend is SDVs. But just kind of put it, it's just a new way of thinking about the architecture of the vehicle, the way you remove features from hardware, you add them in software, the way you think about a life cycle management of a vehicle, over-the-air updates, how do the OEMs stay connected with the consumer post the initial sale, the whole concept of the car driving the worst the first day you drive it because a year later, it learns how you drive and it just becomes better, this whole paradigm has shifted. So SDV is the main driver, I would say. Now what -- how you choose like the powertrain that you choose is a different aspect of it, whether you choose a hybrid or EV or an ICE. So we continue to see a massive trend of EVs, especially outside the United States and China. And EVs continues to be kind of the main category in some of these high-growing markets in automotive. And we -- again, we play in SDV and in the areas where EV is being taken, we play as part of our kind of basically our BMS portfolio that we have.

Great. And then maybe talk about the role of China a little bit. We're seeing a lot of the innovation in the automotive market is happening in China, incorporating these types of features into lower price point cars and things like that. How is NXP positioned to take advantage of that?

Well, very much -- and by the way, there was a new -- kind of a new regulation that it just kind of passed out in China where the Chinese government are increasing the reliability and the safety standards what it takes to kind of go into market in China. And I'll tell you that, that is a nice tailwind that is going to basically cascade from the high end to the low end of the supply chain. And obviously, the intent that China did there was, listen, there are 3, obviously, multifaceted reasons why they want to do that. They want to basically, I would say, curb the destructive pricing competition they have, what they call involution. They want to, for sure, increase the safety of the vehicles. They want to avoid some of the headlines that the kind of high-profile incidents that happened last year with respect to safety. And the last one they want to do is probably align more to the global standards in safety and security benchmarks to basically enable the Chinese OEM to continue to expand overseas. That is a really nice tailwind for NXP because it's no longer about the product itself, right? It's about the product; it's about the level of service that you provide to the customers in terms of technical support over the life cycle of the product. It's about the safety; it's about the security. And so when -- what's happening in China right now, I think, is customers are assessing or reassessing the semiconductor basic partners in order to really incorporate the kind of these aspects associated with what is it going to take for them to be competitive in China given the new regulations. So I think that is a tailwind for us. In China, it's very clear. The playbook for us doesn't change. We compete by being better and being faster. I mean the level of innovation and competitiveness in the Chinese market is massive.

Yes. And the state of Chinese competition within semiconductors to serve those markets?

Well, for sure. I mean, I think you can argue, arguably, China auto market is probably the most competitive market in the world, and we treat it as that. And for sure, right, it is the customers demand the highest level of innovation, the fastest time to market. And China has an indigenous semiconductor industry. They've been successful in high-performance compute for ADAS. They've been successful in high-performance compute for infotainment. They've been successful in some areas of analog and discrete power. But the area that we compete, which is a core function of the vehicle, I mean, this is an area we still compete with the Western peers. And the reason for that is the products there are not just the spec sheet, right? This is about decades, multi-decades of IP development products that they are automotive grade in terms of reliability, in terms of safety, functionality, longevity. And so this is not just about products, this is about IP and know-how that has been developed over decades. But we still remain very kind of alert to the competition. Again, playbook is you need to be better, you need to be faster in China to be able to compete.

And manufacturing in China for China helps you guys as well?

Well, I think it's mandatory. I think it's just part for the course. I mean it's just kind of it gives us the ability to compete. We have invested heavily in making sure that we remove objections in terms of supply chain. We invested in -- we have several fabs with SMIC, TSMC, and I believe the third one is HS Grace. We have basically a packaging that is NXP-owned in Tianjin close to Beijing. And so we have now a really nice robust supply chain for China for China. And we have also a robust chain for China for the rest of the world as well.

Okay. Great. Other elements of automotive, you've talked about electrification, you've talked about radar as key growth drivers. Can you give us an update on those?

Yes. So we said SDV is the core. And I think you mentioned that we grew in 2025 with SDV in the, I would say, the low teens. Connectivity was a standout in 2025. I think grew in the high 20s above our model. And we have very strong franchises in connectivity as well. And also the car is becoming just a big networking engine, right? When you have so much digitization of the functions, you need high-performance throughput around the car. So connectivity is a standout. It's doing quite well. Radar grew on the low single digits in 2025. I mean that was one of the franchises that got caught up in inventory digestion. But I think now that as normalization happens, we see that going back to a long-term model. BMS was below our model. I think it was also another franchise that got caught into program -- basically inventory digestion, but we also saw a delay in some program ramps that we anticipated. We see 2025 a bit of a deviation from what we wanted to do, but I think we expect this to go back to -- the trend around electrification is very strong and it's going to get back to trend. And overall, I think we expect the investments that we're doing on this transformation of the vehicle are starting to scale right now.

Yes. Okay. So you're seeing return to the automotive market shipping to production, you've got good traction with the good -- accelerated growth drivers and kind of seeing all of them essentially contribute this year.

The way I would frame it is inventory digestion removes the headwind and now the secular drivers are the growth engine for the rest of auto. Okay.

Yes. Just maybe to add, just to give you a feel of the accelerated growth drivers in the auto bucket, in 2024, it represented 39% of the auto composition of the revenue. In 2025, it was 43%. And we feel very comfortable and confident that we'll get this to above 50% by 2027 based on the traction we're seeing. Now what dragged us down the first 2 to 3 quarters was the inventory digestion related to our core business. Our core business inside auto finally turned slightly positive in Q4. And what we expect in Q1, that's back to that low single digit for the core business. And therefore, you also see the total company doing quite well above model at plus 11% in our guide for Q1 as a total company.

But Joe, I think this point that you just made in terms of the composition of the revenue in auto being growth drivers around 50%, basically, that's where we're moving to is very meaningful because that kind of gives you a sense of why we feel so strongly that the auto revenues now become structural. The shift towards kind of the secular trends that make the new cars is tying our products and our revenue to that secular trend is what gives the conviction that auto is becoming more and more and more structural for NXP.

Maybe we shift to industrial. If you could talk about industrial, IoT strong growth year-over-year. Can you talk about where we are there cyclically and then versus the growth drivers in that space?

Yes. I would -- yes, it's a great question, industrial because I think I would -- because 2025 was a tough year for industrial. And the way I would frame industrial is both a broad-based recovery, but also a broad-based structural opportunity, a structural opportunity for us. So we are seeing -- and you can see in Q1 guide was a strong guide, but we're seeing broad-based. We are seeing products emerging in our core industrial and industrial and IoT. We're seeing products in factory automation, energy storage, building automation. It's quite healthy in terms of broad-based. No single vertical is driving the business right now, and that's what it gives us conviction. There are some new areas like energy storage, which is actually coming strong. But I think the way you should think about industrial, it seems like every single segment is getting a transformation. Every real segment is getting reinvention. Now what makes NXP different than other industrial companies is that we are a processor-first company. We lead with processors. And so we kind of really go to this market with the strength of our processing portfolio. And what we're seeing right now based on the engagement that we have our customers is not only new areas they're emerging, but it just seems like there are a lot of use cases that are getting reinvented, whether it's managing kind of deterministic networks for factory automation, whether it's doing new types of controls for building management or secure access or in health care, patient monitoring, different types of instrumentation. These are requiring kind of new products, new type of silicon solutions. And so we see this -- we also see the complexity that these systems are taking. So we're taking an approach that says, listen, product is no longer enough. We need to remove complexity for our customers. So we incorporate connectivity. We're going to incorporate security, the processing, the PMUs, the analog piece. And so we're taking a system approach, very similar to what we do in the auto, we're applying it into industrial. And so we kind of go out there and said, okay, this is how we're going to remove complexity to our customers. So industrial is -- it has this transformation happening. And the true secular trend that is happening in industrial is AI at the edge, physical AI, edge AI at the edge. And it's at the center of every conversation that we have with our industrial customers. You can't avoid walking into a customer and not having a conversation on how to deploy AI to either augment their current offerings or create new ones. We see it. Our customers are developing platforms with room in terms of headroom, in terms of processing. So they don't know what they want to do, but they know that we need to evolve that platform to basically over the air, update these features, make them secure by design, make sure they augment it with other functionality. And this transformation is very visible. They're going from very similar to what happened in automotive to hardware-defined features to more software-defined systems. And NXP is kind of a prime position to help them in that journey.

I guess in autos, it's very clear why that needs to be at the edge because you can't be going to the cloud for stuff when you're driving a car in real time. Is that same thing true in industrial? Why do you need -- why can't I just call in the cloud for that information? Why do I need the actual AI capabilities at the edge?

Well, it depends, right? It depends. I think it's going to be a hybrid model. There are -- I think even the car is still connected because you still need to update the models and actually kind of go back to a client for training. But anything that is low latency, anything that is high level of security, anything that is going to tamper with the actual functionality of the system, you want it to have control. You want it to be at the edge. But I think that's what is just the one thing that is quite wonderful about industrial right now is companies are thinking about what the software-defined systems are going to look like. And these are the type of conversations they're having with us because of the capabilities that we have, not only on processing and on security, on deterministic networking, but the new capabilities that we have with AI. And I tell you the level of interest that they have on our AI platforms, our customers are massive. And this is what gives us the conviction. I think in many ways, we feel internally at NXP that industrial is a sleeping giant for -- in our portfolio for NXP because there's massive interest of AI at the edge.

And maybe, Joe, let me hopefully hear the excitement and the investments we're making in physical AI. Let me for our owners out there and provide some numbers. Remember, our industrial, IoT end market, 60% is industrial, 40% is IoT. In Q4, this was up 24% year over year, and we just guided Q1 up in the low 20s. The accelerated growth drivers, there's a bucket of them focused on processing, analog and security. In 2021, this bucket was about $500 million, 2024 was $600 million, and really, the growth has just taken off. This grew above model near 30% in 2025. And our ambition is to be above $1.2 billion in 2027. During that, the core even declined by about 10% in 2025. So you'll see the strength of our industrial, IoT comes across. And a composition of that end market, think about it in 2024, the growth drivers were 28%. In 2025, they were 36%, and we're well on our path to have the growth drivers be meaningful to plus 40% by 2027.

And just to put, even add another layer kind of data that gives you the strength of our of kind of the ramp of AI in our portfolio. In 2025, the revenue that has some AI component in industrial was only 3%. And we expect this to more than double in 2026, and this doubling also in a year where industrial is going to grow. And so you can see from an actually absolute value perspective, AI is becoming more and more of the portfolio of industrial. And this doesn't even include the Gen AI capabilities that we acquired through Kinara through the neural processing unit that they have that is discrete. It includes nothing of that. This is actually organic growth over our previous AI solution. So again, this is actually a really, really strong tailwind that NXP and it is going to differentiate NXP at industrial.

Yes. I mean I assume you have a lot of long-lived sockets. So when you have that kind of approach, it means a lot of your new activities around those areas.

Yes. Well, I think AI is actually kind of becoming a transformative move. You must have it. And I think the transformation that's happening is challenging incumbencies, right? You can't assume that incumbency the same way you assume incumbency a while back in industrial. AI is going to transform that. Now the good thing about what we are positioned in NXP is incumbency is going to get strengthened by the addition of AI, but it's also an opportunity for us to kind of -- to actually challenge other incumbents in the market with maybe less optimal or less innovative solutions with respect to AI.

Great. So maybe just to round out the product portfolio. Talk about your communications exposure, sort of transitioning from RF to secure cards, just how to think about that.

Well, we've been transparent, right, on the structural change in the revenue of comms and infra. And just to remind you, in 2024 in Analyst Day, we said this segment was going to be essentially flat throughout 2027 with secure cards really kind of mitigating some of the decline on the other 2 segments with digital networking and RF power. So in 2025, secure cards was north of, I would say, north of 50% of the revenue. And the other 2, digital networking and just took the other 2 equally, almost equally. What I want to highlight what is -- and one thing that's probably being masked is this headline of flat over a period of 3 years. I think it's masking a structural shift towards higher quality, more predictable revenue because it goes away from lumpy businesses, our power, digital network that we have -- we stopped investing towards franchises that they are more -- basically, we have more visibility to predict. It's got higher -- kind of higher growth areas. Just to remind you, in secure cards, we have very, very strong franchises like MIFARE QR code, which is our RFID product that's seeing really, really good growth in the market. So that's what I'll leave you is there is a structural change. It comes in infra from revenue from -- away from digital networks and our power towards secure cards, and that gives us more visibility and more predictability.

Okay. Helpful. And then finally, on mobile, you're sort of become more of a niche premium supplier to that market. What are you seeing there? Any turbulence around memory, things like that?

Well, I can't necessarily comment -- we don't purchase memory, so I'm not going to be able to actually have provide -- I mean, memory is a conversation with every customer, okay, no matter what flavor of memory they use. But mobile, we focus on a franchise that has deep technical moats that relies on ecosystem. And this is a mobile franchise. We also do some analog-specific products for some of our customers. Last year, we saw a growth in mobile business close to 6%, is a bit shy of 6%. And it was driven by the strength of our mobile portfolio in the mobile wallet. It was strength by content gains that we have in mobile. The way we see this basically uplift is in terms of longevity, I think this is -- this growth is durable. The franchise that we have in the mobile wallet is strong as deep moats technically in use cases, ecosystem. And the content gains that we have tend to be kind of multiyear design wins. And so we feel strongly that right now, the mobile is in a good trajectory.

Great. So I have one more question, and then I'll turn it to the audience. Maybe if you could just remind us, you talked about a gross margin model of 57% to 63%. Can you talk about the puts and takes there as you approach.

Yes, sure. In the short term, typically, you have utilization, you have mix changes. But more exciting, we use a rule of thumb for every $1 billion, think about 100 basis points on top of our model. Everything is projecting and forecasting to what we expect, and things are just grinding higher as revenue grows at this point in time. The real uplift that will occur probably a step function more material for the company is probably in 2028, 2029 with a 200 basis points increase related to our joint venture with VSMC out of Singapore, where basically we'll just get the economies of scale from going from 200 to 300-millimeter as well as more of a cost-plus type of model so that we can compete with other peers.

See if we have questions from the audience.

Rafael, you mentioned inside the drivers for industrial that you mentioned the energy storage systems. I just wonder if you could size that out for us? And is there any growth outside of data centers for ESS right now?

Well, this is definitely an emerging market for us that we target as a complete total system solution. We don't sell necessarily our BMS product. We actually tend to sell the entire system solution because it needs to be certified with both security and functional safety. So we ended up kind of helping our customers basically go to market with these systems. This is a nascent market. And it actually didn't start with data centers. It started with a lot of systems that were used to actually load balance the grid, and this is where we saw it. Also for some systems that actually create some, I would say, some basically resilience for buildings. We are starting to see now more and more that data centers are going to be driving the need to be able to kind of really load balance the center and the spikes in terms of both power that they need to have is going to be something that is going to be a tailwind for us. I can't tell you right now, the data in terms of how big the market is, it's still evolving. We're going to be looking at. But this is an area we believe is actually an area that is going to be an area of focus for us, given the system solution and the kind of the portfolio that we have is quite good for this market. And it's not just on the energy storage system, but it's all the way to charging and the management of bidirectional charging and bidirectional charging from the buildings to the grid. And this is an area that we -- it's kind of our portfolio is very well suited for it.

Any questions? If not, we'll wrap it up there. Rafael, Bill, thank you so much.

Thank you.

Thank you, Joe.