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

Okay. Thanks, everyone. So this is Morgan Stanley Day 2, 25th Anniversary TMT European Conference. And we're very glad to welcome to the stage, CEO of NXP, Rafael Sotomayor; and the CFO, Bill Betz. So welcome, gentlemen. Welcome to Barcelona. Glad to see you here.

Thank you.

Maybe I'll just kick things off just to level set everyone in the audience. Maybe walk us through again, key points from Q3 and what you talked about for guide, maybe even by end markets, but what was the key points at least anyway?

Thank you. Yes, absolutely. In Q3, well, first of all, I think there are several signals, right, and several reasons why we feel good about the trajectory of the business so far. If you look at -- this Q3 now will be the 2 quarters in a row, we have sequential growth. And let's start with the first one. First one, we delivered $3.17 billion of revenue in Q3, which is slightly above the midpoint of our guidance. And what is encouraging is what's behind it, which is it was broad-based. Every region was up, every end market was up. Our direct customers were up, the channel was up. So when you look at a company with this level of diversity, and you have this, I would say, this synchronized improvement, that is actually a very healthy indication of, I would say, business improvement. When you have this, kind of everything is actually going well. Second piece for us is very, very -- kind of very bullish is the areas where we invest more, the areas where we believe there are growth engines, the differentiator of NXP. Why do you own NXP? Why our customers select NXP? These are the areas that are very NXP specific. And these areas in software-defined vehicle and the transformation of edge, AI and industrial, all these things are actually performing. Performing in an environment that is actually quite challenging, right? So that means they're growing, they're doing even slightly better than what we think they're going to be doing. So that's the second piece. Now when you look at a broad base the KPIs from a cyclical perspective, the things that we watched. Let's start with the most important one that we mentioned in the last earnings call. We see now inventory normalization, right? It seems that inventory digestion for the Tier 1 auto customers that we have is finally coming to an end, which means that we're going to start shipping to end demand, which shipping to end demand is actually quite good in auto, right? For us, we see this as a very nice thing to see finally coming to the end of the inventory normal -- coming to the end of inventory digestion. Now we look at the channel. The backlog of the channel is improving. The sell-through on the channel is improving. And even our direct orders, the EDI signals that we get from our Tier 1s and our customers is also improving. So you can see the overall -- the demand for NXP products is in a different setting. Now visibility continues to be low. That's one thing that seems to be universal. Uncertainty actually kind of makes customers kind of put orders very late, right? And that continues to lead to customer escalations and those escalations have kind of -- every quarter, they have actually increased, the number of escalations, not necessarily the number of dollars, but the escalations have increased. But nonetheless, I think that kind of shows you that demand kind of continues to be improving, the demand conditions. Now that leads me to the final point I will say, which is our Q4 guide. Like we guided to 3.3%, which is 4% sequential growth, which is above seasonality. And that puts us for the first time this year on a year-over-year growth of 6%. And that's at the low end of our guide, but we're quite excited about the fact that we kind of go back to growth year-over-year. And if you look at what also is behind our guide is strength in company-specific drivers in industrial, which is quite encouraging. These are company-specific. These are design wins that we had in energy storage, battery management, some design wins we have on a particular category of wearable devices is doing quite well in smart glasses. And that growth sequentially is 10% just quarter-over-quarter in industrial. And so we feel optimistic on how the year is closing and how we are going to start 2026.

Great. So it seems like most end markets doing well. Channel is quite lean. Industrial strength being seen. Margin and the guide looks like it's at least seasonal. Things are going well, or at least well enough through the bottom of the cycle.

That's how we see it. Listen, I'm not going to go and say this is the lens for the entire industry, and I know that some of my peers have different kind of messages. But the way we see it is things are improving and are going in the right direction, and we like the setup.

Perfect. Okay. Maybe if we drill into some of the end markets. I mean, it did look as though automotive in Q3 was a little flat, but other markets were up. One did stand out, communication infrastructure was down a little bit. Maybe help us understand what's happening there? And then what is the strategy to see that recovery coming back there?

Yes. So if you step back in Comms and Infra, maybe we start with the premise. Comms and Infra today is composed of 3 unrelated businesses. So it's literally kind of a bucket of a cat that is not mobile, not industrial, not auto. Let me put it kind of that way, right? And so there are 3 businesses there. In November Analyst Day, we basically gave a projection that this particular segment, this particular revenue bucket for NXP was going to be flat for the next 3 years. Now let me tell you the dynamics behind that. You have 3 businesses. One is the networking edge business. This is a networking product that we stopped investing a long time ago, and we said that it's going to go to 0 by the end of 2027, okay? So that means it's declined. You have radio power, which is radio for basebands for networking devices and infrastructure, 5G, 4G and 6G. And that's a bit of a choppy business. It's a small business. It is relegated to 2 or 3 big customers, and that's kind of flat over the years. And then you have the Secure Card business, right? So you can imagine, if you have a business that is declining, it means the secure card business is growing. And so that is the dynamic that you see is at the end of the day, at the aggregate level, that bucket is going to stay flat. But in reality, the one portion of the business that is actually performing and keeping it flat is the secure card business. And these are products that go into payment cards, passports, ticketing, MIFARE and RFID.

So Secure Card is really holding the floor together to flat, but somewhat declining...

Declining, yes.

Got you. Bill, maybe if we turn to yourself on margins. I think last print, we talked about return to the long-term financial model. I think 57% to 63% is target. 57.5%, I believe, was margin for the next quarter. So maybe help us understand what are the moving parts in the margin? And how does Q4 set you up potentially for next year as well?

Yes, absolutely. Our model is 57% to 63%. The very simplistic way to think about gross margins is for every $1 billion of revenue, it will drive an incremental 100 basis points to our gross margins. And so think about $15 billion, we should be around 60%. Now what has occurred with gross margin, and there's the structural foundation of gross margin, the discipline inside the company, and that's driven by 2 factors. The first factor is our portfolio. And it happens over many years. When I joined the company and sure when Rafael joined the company, the gross margins were at 45%. And so what that meant with our R&D investment dollars, and you think about the ramps of our products in both industrial and auto, it takes 3, 4 years to that ramp. So the investment dollars and the choices we had 10, 15 years ago, that hurdle rate for the company was 45%. As we expanded and grew, it went to 50%, then it went to 55%. And so think about today, all our new investments and choices that we're making inside our portfolio, that hurdle rate today is at the midpoint of our 57% to 63%, which is 60%. And so those investments that we'll see, which you'll probably see post 2027, be beyond 60%. Now the other foundational change, to get gross margins more resilient. So if you compare us compared to our peers, what has occurred is, in the past, I would say when we went through a cycle, our gross margins would probably have a quite variability. How we managed this last down cycle? We think we navigated it quite well. We kept utilizations at the 70% internally. But more importantly, our hybrid manufacturing strategy allowed us to diversify and become more flexible, more resilient for our supply, but also more cost competitive in our supply. Therefore, I would say, when we joined, Rafael and myself, again, 15 years ago or so, we were 70% fixed. Today, we're only 30% fixed. And so through that transition, your variability on the gross margins become more variable in nature, so you shouldn't see the large swings. We're not taking business to go fill our factories. That's not what we're about. We're about making sure we provide the value to our customers and what they need and at a cost competitive that's helpful for them. So those are the foundational structural changes, I would say. Now what's in the short term to medium term? What's going to drive gross margins? Obviously, the function of revenue, we get back into model with Q4 guide, plus our modeling soft guide gets us back to growth year-over-year. We feel very confident that we will deliver in our financial model for next year and beyond. We're not backing away from the entire financial model. And I would say utilization, as we mentioned, is we're just beginning our 200-millimeter consolidation. And so think about these internal factories that run 90 nanometers and above. And so what's occurring is all our new design wins are on below 90 nanometers. And so there's a shift to make sure we support that. And we will be structurally moving and transferring product to our foundry partners of TSMC or Samsung or GlobalFoundry for that resilience of supply. But more importantly, we also have joint ventures that we're investing in that we own, whether it's 10% of ESMC with the likes of Infineon and Bosch and TSMC for resilience in Europe, or our VSMC, our newest joint venture that will be up and running in 2028 in full production with the likes of Vanguard, our partner, also supported and backed by TSMC. That brings us a very cost competitive and be able to support and supply this ramp. We have this huge design win funnel that we have to go deliver and execute to. And so over the next couple of years, you're going to see us execute and grind gross margin higher through our hybrid manufacturing strategy. So again, short-term utilizations, the 200-millimeter consolidation, when we finally build the end-of-life products, utilization improves. We may sell these factories. Again, that's all -- we'll update that on a quarterly basis. And other things I would say is the mix of our products. Over time, these new product introductions, like I talked about in the portfolio, grind the gross margin higher and also something that we like is the industrial space and our go-to-market through the distribution. The higher the distribution, think about it, it's low volume, higher margin. And so we want to continue to invest in our go-to-market. And so we feel pretty confident to get our gross margins not just at 60%, but think about post 2027, with just the likes of entering and filling up our VSMC joint venture, we will get an incremental 200 basis points on top of the corporate margins post 2028. So we'll update more about that in our next Analyst Day in 2027, but we feel very confident in our gross margin projections with the support of our revenue growth in the markets we play in.

Got you. Sounds very healthy. So sensible hurdle rates, good variability and flexibility in the manufacturing, and a transition as well, where you're managing joint ventures at the same time. So it seems in good shape. And you talked about industrial growth. I think if I stay with you, Bill, actually, there is always pricing pressure. I think low single digits usually is muted. And it's right about this time of the year, you see the price negotiations. So how does that set us up for Q1 and the kind of margins you'd expect? And maybe just if I add on to there, you did talk about migrations, but is there anything else you're doing in cost management or cost reduction programs, restructuring maybe?

Yes, absolutely. So in Investor Day, what we assumed in those growth rates of 6% to 10% is an annual low single-digit price reduction, which majority of it happens in Q1, because with the automotive exposure we have, that's when it takes place. And the way to think about that is you get an impact in that first quarter. But going to your second part of your question is, we then get the cost improvement, the yield improvements, the test time reductions, the cost reductions that we get back. And so the way to think about the low single-digit price, that's just normal being competitive in the business, that all gets offset with the cost improvements that we make throughout the year. So those 2 kind of offset each other really from it. And typically, Q1 is our lowest quarter of the year. So yes, we kind of made sure that we get those gross margins. We still think there'll be a model. It's hard to always hold from Q4 to Q1 with our revenue typically seasonally down and also our price gives, but we feel very confident we'll get and grind those gross margins above current levels of Q4.

Got you. So maybe if we turn back to yourself, Rafael, and I look at autos in particular, 2 main drivers really. We had xEVs obviously pushing for a lot of dollar content increase, but obviously, software-defined vehicles becoming a big play. I know you've got the S32 family. You're well positioned in this space. Help us understand how does the dollar content increase maybe next year and the year after as you see things?

Yes. No, happy to get into that. So I think probably, step back and look at it from the content -- the way to look at NXP is that we don't have a fixed content per vehicle. When you have the exposure that we have in automotive, NXP -- I mean, it's likely you pick a car randomly in the world, there will be an NXP part, right? And so the content per vehicle really varies from as low as $100 all the way to greater than $1,000. And it really varies that much. And it tends to be related with the architecture of the vehicle or the type of vehicle, the type of features, it's an older model or a newer model. Maybe a more probably interesting way to look at it is to look at the subsystems of the car. So you can see kind of the drivers of content growth. And I'll start with what you talked about. You mentioned SDVs, right, the software-defined vehicle, and that is at the core of our processing portfolio. And these are portfolios that go from relatively kind of low power type in the lower-end MCUs, even though for us, they're not super low end. We tend to be in the high-performance part of the spectrum in all our processing. But that will be in, let's say, on the $10. And then you have the products that are driving more the zonal architectures, right, that tend to be in the $15 to the $25 depending on the features and the memory and all that. And then now you get into our 5-nanometer products who are the latest and greatest that we have for the transition that is happening to central compute architectures. And you see the spectrum, right, of choices and architectures, but also spectrum and the choices you select for your microprocessor needs or microcontroller needs. And that tells you -- and where you are in that spectrum is really largely determined by where the OEM is in their SDV journey, okay? And there are customers who are in the zonal architectures, driving a lot of the next-generation zonal architectures with our K5 family, which is our 60-nanometer with MRAM incorporated. And the other ones who are more progressive and going towards more of a central compute, where we are going to see content right there with 4 or 5 of these chips at a $50-plus range. And so that's big content. Big content growth by the selection of the architecture, right? So that's one driver, and that one is undeniable is happening. The move to SDVs is unstoppable. It's just a matter of where the OEM is in that journey. And you have the push for more ADAS, right, the higher level of ADAS, which for us is a driver for sensors. So the way you got to look at it is, the push for higher autonomy in the car drives the need for more radar and NXP is the leader in radar. And so what it means is that not only more cars adopt radar or more cars adopt more nodes of radar, right, as they go from 1 to 3 to even 4 radar nodes. And every radar node is about $20. So there's some content growth there. And then you have, obviously, the electrification piece, right, the BMS piece, which for us will be -- I think it will be for a low-end 400 volt, that will be about $60, and for a higher end, 800 volt will be twice that. And then you have all the other stuff, networking, connectivity, car access. And I'm sure I'm missing some. And all that also is driven by the fact that the car requires more and more data. The car becomes a massive networking machine. And so that is also content growth. And so when you look at 2026 and you put all these things together, obviously, at least you understand the drivers from the content growth. Now if you were to put numbers in it, and I'm going to use the guidance that we gave at Analyst Day, our Auto business is going to grow in a range between 8% and 12%. If you assume a SAR growth of 2% or mid-single digits and low single digits, then you can see, at that level, at that average level, you can see at least a content growth that goes from whether that would be 6% to 10%, right? So that will give you a sense of what the content growth is in the way we're measuring it.

Really interesting, the changes. I mean, I think you alluded to it a little bit there that there are different routes to this EE architecture or SDV play. And it's interesting that you go from microcontroller to microprocessor. What is really pulling that microprocessor story? And why are people coming to NXP in particular, do you think?

Well, first of all, I think the processing needs of the next-generation architectures are just driving higher performance, just driving higher performance. And about literally like almost 7 years ago, we decided to actually basically tilt our focus into not what the business was then or even it is today, but what the business was going to be. And what you will require is that we ended up with investments to develop products that are quite differentiated. I mean some of our products are original. They are original, meaning is there's nothing like them in the market. And now I think the OEMs have caught up the idea that they need to own the software. They need to own the architecture. They need to keep up what happened that was originally initiated by Tesla and some of the Silicon Valley customers, right? It opened a new car entry, but the Chinese quickly follow. So a lot of the legacy OEMs are now coming up to the realization they need to adopt a software-first mindset. And the products that we were using before and the dependencies of the supply chain and the automotive, they just no longer work for them. And so what they now go and say, okay, well, now I need to develop the software. What is the right architecture? Well, the old products don't fit anymore. So you need to look at this high-performance microcontrollers, the high-performance microprocessors, because you need to develop software. And not only that, you need to also have -- you need to have the ability to have extra CPUs, extra cycles to be able to deploy future software, right? So you got to think about a software-defined vehicle, very interesting. The worst driving experience is going to be the first day you buy the car. And the car will get better over time, which is like the reality is backwards, right? How it gets better is it learns how you drive, how you like it. It learns how the vehicle is operating in the environment, it gets updated. And for that, you need a really robust processing foundation and the right software on top of it. And that's what is making NXP different.

So good hardware redundancy for the over-the-air upgrades that you have?

Absolutely.

Got you. Maybe just staying with -- I think you just touched on China and China's adoption of software-defined vehicles. I think China grew 13% Q-on-Q for you guys. I mean help us understand how does that sort of maintain itself going into next year? We do hear about xEVs peaking somewhat in China, 50%, 60% penetration. So it's got to be SDVs, it's got to be the content increase there. But what is driving that?

Yes. So start with the premise that China is a very important market for NXP. And if you look at, the latest we have is 39% of our business was shipped to China. Half of that was for indigenous companies like Chinese headquartered companies. Whether the product stays in China or goes overseas, at the end of the day, it's really hard to tell how much of that is, but it's for Chinese customers, so half of that is. So you can see that China headquartered companies, of course, about 20% of our revenue. So it's a very important part of our business. And I would say that for auto, just kind of step back, for auto, we like China, because they're fast movers. They tend to adopt our latest and greatest innovation the fastest. And so we tend to see what works and what doesn't work first in China. So being there in China allows us to have a peek of what is going to work for the rest of the world a little ahead. So strategically, we tend to consider China very important. If you can keep up with China, you can keep up with anywhere, right? Now I think today, competition in China, especially in auto, it continues to be the same people we compete outside China. It continues to be the Infineon, the ST, the Renesas and the same players. I don't intend to say that the indigenous kind of semiconductor is not going to be there. All I'm saying is that where we play today, we don't see it, right? And when and how we're going to see it, I can't really predict that. But the way we see China is, it's an overcompetitive market among themselves. I'm talking about now car OEMs competing with other car OEMs and competing for the outside world. And I believe that if we continue to play our playbook and execute our playbook, which is be ahead with innovation, be ahead with a product that creates value for the Chinese OEM and the Chinese Tier 1, I think that playbook has worked so far. And I think it will continue to work as long as we're able to keep up with innovation that adds value.

Okay. So innovation is key there in that market. I've got a couple more questions for you, Rafael, and maybe I'll turn back to Bill later. But at this point, I'll ask if there are any questions from the floor before I move on. Not at this point. So Rafael, maybe just changing to Mobile. I think Mobile revenue has been a little volatile, it looks from the outside. So what is your outlook for this segment going forward? And what drives the business for you guys?

Well, maybe I'll rephrase your comment on volatile, especially the volatile part, because I think it's seasonal. I don't consider the Mobile revenue volatile. Actually it's one of the ones that has actually been quite stable. But it's seasonal. I mean the first half tends to be weaker than the second half because both of our biggest customers tend to introduce new products in the second half of the year and then kind of basically be weaker in the first half. But our exposure to Mobile is very limited to a particular niche category, right? Most of our revenue is driven by the mobile wallet, which we have a very big kind of share there, and we have very important customers. We like that. It's a good franchise for us. It's something that we don't talk too much about it because -- but it's mostly an ecosystem play. We got to think about mobile wallet as an ecosystem play. It's a product that develops -- that enables secure transactions. So if you have your pay in your phone, it's very likely to be driven by NXP technology, very likely. And it's a system solution that drives the NFC, near-field communications ecosystem and security. And that evolution of the mobile wallet continues to happen. We continue to add more and more functionality to it. We do mobile payments and then we did mobile transit and now you can use your phone to actually do mobile transit. Even today, here in Spain, you can go to Madrid and use your phone using the mobile wallet from NXP. But the other thing that you can also now do is we are now today working on embedding the SIM. The SIM is another secure product that occupies too much space on the phone, and we're going to incorporate that functionality into the mobile wallet. So the mobile wallet now becomes a very secure enclave for all sorts of secure transactions that they are sensitive from a security perspective. And that business is solid. Again, it's solid. We're a niche player in Mobile. We don't represent more than that, and it will continue to be a good business there.

Got you. So seasonal on product cycles, but a good dependable niche in secure ICs around the payment ecosystem?

You got it.

Perfect. I kind of was also curious that you talked about growth in Industrial. And some of your peers over here in Europe are still somewhat struggling. It looks -- I mean, very much focused on general purpose microcontrollers and are hostages to factory automation and the slow orders we're seeing there. Yes. So maybe help us understand where are you positioned differently? How are you seeing things in Industrial, frankly?

So I did get a lot of questions on Industrial because the performance -- what is considered the overperformance on our guidance for Q4. And again, when you compare us and the comparison to our peers who are representative of Industrial, I think it's not necessarily constructive, let me put it this way, right? Because we -- and I said it in earnings, we don't consider ourselves a bellwether for Industrial. Our exposure to Industrial is different. Our exposure to Industrial is, if you look at our -- our Industrial exposure is mostly on the digital side. So we are not a catalog component. Most of our processing that we -- whether it's microprocessors or microcontrollers, we have both, right, are application-specific. And so I think that if you look at some of the performance that we have in Industrial, I mentioned it, they seem to be driven by very company-specific drivers, right? And so it may not necessarily indicate the health of the industry. It just indicates the health of our business, which I would say we feel good about it. We also have a system solution. Every time we do a processor, likely, we also get the connectivity piece. And energy storage systems, for instance, is -- the way we won in that area is by doing complete system designs, right? And so we deliver system solutions to our customers who are actually prequalified, precertified, because most ESS need certification, by the way. And so there's actually a go-to-market aspect that is very strategic in which we position NXP for Industrial. The other thing not to completely overlook is how we managed the downturn, right? The downturn was managed with us managing the channel. Industrial, 80% of the business goes through the channel. So having a healthy lean position in the channel makes it easy for you to address today's needs, because you don't have to carry the burden of inventory that used to address yesterday's needs. And that is quite different than some of our peers. Today, our inventory in the channel and at the end customer is in a healthy manner. So I'm able to actually stage products that we have high conviction that they're going to sell through, and that's what is going to lead to overperformance in Industrial for NXP.

And just to add related to Industrial & IoT, I think a lot of it also is company-specific, as Rafael said. So there's a subset of products that we shared during Investor Day, which is processing, connectivity, analog and security. And that size was about $600 million in '24. And we are well on pace to double it to $1.2 billion specifically, as now the design wins and the cycles kick in, we feel really good about it. And as we said during earnings, all our accelerated growth drivers are tracking to those plans of that 20% clip. We'll break out a little bit more granularity at the end of the year, but we feel very confident of what we're doing. And those company-specific drivers is what's really probably given us a bit better than our peers, what we can see, along with that lean channel inventory that Rafael mentioned.

Yes. So that was the message I was getting. Good lean channel inventory management, application-specific bias rather than general purpose, and $1.2 billion run rate is target. So it looks healthy. Maybe, Bill, if I turn to you then, last question here for me. So what are the key assumptions? I think your goal for earnings here by 2030 plus. I think you gave something at the Capital Markets Day. It evades me. It's not on my paper here. Maybe help us understand what the long-term game is.

Yes, absolutely. It's quite simple, actually. If you don't remember anything today, remember these 3 or 4 things, right? First off, we're going to grow our revenue. We feel very confident in the high single digits. So we guided 6% to 10%. So I just assume 8% in the model, right? Simple math in excel. Gross margins, we're going to expand them above 60%. Think by 2030, we should be well above that by just applying for every $1 billion, 100 basis points improvement in gross margin. As you all know, we're very disciplined in OpEx, right? We have a model of 23%. We'll probably get some fall through and improve on our SG&A as we adopt AI automation internally and so forth, just like every other company. And we feel very confident, and look what we just did? We basically took on 3 acquisitions that were dilutive to us, because that revenue doesn't show up until 2027 and beyond, and fit it in our portfolio and transformed the company and get it ready. As what Rafael talks about is, we're playing for the future. We're not playing today. We're playing where we believe the future will be 5, 10 years down the road. So if you just put that all in your excel model, it will basically spit out an answer that says that $13 of earnings per share back in 2024 will be easily $26 in 2030. And then on top of it, besides doubling our earnings, we feel very confident about just think about the amount of cash that we return to our owners. That's very important. We have a track record over the last 10 years. We have returned over $20 billion in the form of dividends and buybacks to you all as well as expand and improve the earnings of the company. And I think if I laid out in Investor Day, I think there's another $8 billion we're going to return or $9 billion from now until 2030, as well as delivering to that financial model.

I lied about the last question. I have 2 more, if I could.

We've got 8 minutes left. So...

Yes, go ahead.

Yes. It's interesting, you touched on the acquisitions. And I think you've made 2, I'll call them, bolt-ons, Aviva Links and Kinara. But as far as I understand, I think there's 2 months of impact in the upcoming quarter in OpEx. So how do you manage bolt-ons like this feathering in? I presume there's annual bolt-ons as well. So how do you manage the OpEx integration story there?

Yes, we're quite disciplined. Myself, Rafael or my predecessor, every year, it's about how you deploy that capital internally. We have a quite disciplined process in-house. And it goes back to that foundation of hurdle rates, entitlements and this rigor on execution, speed and focus and innovation, it's all a combination. And when projects are not performing well, we have to shrink them or exit them or, in other areas, we have to place bets and double down and go expand. And so it is a puzzle. But we don't plan on doing 3 acquisitions every year. That just happened from a timing perspective. But what I would tell you is to help fund these acquisitions, we always had in our mind of looking at the portfolio. So we also announced over the last quarter, a divestment of a business called Sensors, and we're actually selling it to one of our peers. And the beauty of that is, it's probably better off in their hands, because guess what, they're manufacturing the front-end wafers for us. And so that business was a good core part of our business, growing 2% to 3%. Margin for us, slightly below the corporate margins, but we saw as a headwind because, guess what, they're manufacturing the wafers for us, they can impact if costs go up, they will pass those costs to us. So in theory, we took that money. We said, "Hey, let's go sell it." We sold that at the right time, get $950 million, take that money and actually move the portfolio and invest in more richer value to our customers, which means higher margin for us in TD Tech, Aviva Links, and Kinara, which is going to transform the company and also help transform our customers and make them successful.

So there's a point that you asked about how do you guys do that. By the way, you mentioned it from an OpEx perspective. But also what I want to say is that there is a big longer-term objective here, which is the reinvention of NXP to position ourselves to be the leader of bringing intelligent systems to the edge in Automotive and Industrial, and that requires different type of capabilities internally. And so when we made these acquisitions, it's also to bring talent that we didn't have, and we shared talent that we probably no longer kind of serves our direction going forward, right? So it's got 2 aspects to it.

Yes. I'll give you -- Rafael just triggered something in my mind is, after these acquisitions, right, more than 50% of our R&D community is software now, because we are focusing on that system aspect, as Rafael said. That's where we want to be. That's where we're going to drive value.

So key value areas, getting the full stack. MEMS sensors, it's not something you can get the hurdle rate to work. So send it to someone else like ST and they'll get the verticalization benefit. It makes sense indeed. I promise this is the last one this time. I did hear you mention energy storage systems. And over here in Europe, we have a couple of players who are getting warmed up to that as a future play, particularly as we look at new power architectures in the data center. So I just wondered if whether or not you saw that as a driver for growth in ESS for you guys or if it's somewhere else. Just help us understand.

Well, I think the electrification of everything, I think, is a big driver for us to actually position ourselves into that secular trend period. I think it drives a lot of innovation. I think it doesn't stop in auto. It doesn't stop with energy storage systems. At the end of the day, you can actually envision a world where autonomous devices, robots are going to be part of our daily life, and all those devices are going to have some form of electrification aspect to it, right? There's going to be some form of cognitive capacity and intelligence, right, and processing needs. You're going to have a lot of engines, a lot of motor control. So the way we position, the way we invest in our portfolio, it tends to be in technologies that are at the core of what the world is going to look like 5 to 10 years from now, whether it's autos today, robots tomorrow, energy storage systems that are static, to energy storage systems that are mobile. And so I think we try to make sure that when we develop IP, we have the ability to catch the secular trends that they're not so obvious today, but then they become inflection points of opportunity in the future.

That's very interesting. I never thought about mobile ESS, but there we are. Yes, there we go. I learned a lot. Gentlemen, I think we'll cut it there, but thanks very much for coming. Enjoy Barcelona.

Thank you.

Thank you so much.