Operator: Good day, and thank you for standing by. Welcome to NXP Second Quarter 2026 Earnings Conference Call. Please be advised that today's conference is being recorded. I would now like to turn the conference over to Jeff Palmer, Senior Vice President of Investor Relations. Please go ahead.
Jeff Palmer: Thank you, Lisa, and good morning, everyone. Welcome to NXP's second quarter earnings call. With me on the call today is Rafael Sotomayor, NXP's President and CEO; Bill Betz, our CFO; and Michael Lucarelli, our incoming Head of Investor Relations. The call today is being recorded and will be available for replay from our corporate website. Today's call will include forward-looking statements that involve risks and uncertainties that could cause NXP's results to differ materially from management's current expectations. These risks and uncertainties include, but are not limited to, statements regarding the macroeconomic impact on the specific end markets in which we operate, the sale of new and existing products and our expectations for the financial results for the third quarter of 2026. NXP undertakes no obligation to revise or update publicly any forward-looking statements. For a full disclosure of forward-looking statements, please refer to our press release. Additionally, we will refer to certain non-GAAP financial measures which are driven primarily by discrete events that management does not consider to be directly related to NXP's underlying core operating performance. Pursuant to Regulation G, NXP has provided reconciliations of the non-GAAP financial measures to the most directly comparable GAAP measures in our second quarter 2026 earnings press release, which will be furnished to the SEC on Form 8-K and is available on NXP's website in the Investor Relations section at nxp.com. Now I'll turn the call over to Rafael.
Rafael Sotomayor: Thank you, Jeff, and good afternoon. Our second quarter performance exceeded expectations once again as the strong momentum we saw in the first quarter continued into Q2, setting the stage for a strong second half. Demand improved across all end markets highlighted by strength in both our company-specific growth drivers and core business. This combination of secular and cyclical growth is enabling a durable revenue stream that expands margins and drives strong earnings growth. Second quarter revenue was $3.5 billion, up 19% year-over-year, non-GAAP operating margin of 35% and non-GAAP EPS of $3.61, all exceeding the midpoint of our guidance. All end markets and regions grew versus the prior year. Our company-specific growth drivers grew in the mid-20% range year-over-year, and represented roughly 1/3 of second quarter revenue. In addition, our core businesses increased in the high teens range year-over-year -- that broad-based momentum is also contributing to our growth. Now turning to end market performance. In automotive, revenue was $1.94 billion, up 12% year-over-year and slightly above expectations. Adjusted for the sale of the MEMS sensor business earlier this year, automotive growth was up 17% year-over-year. The company-specific growth drivers grew in the low 20% range year-on-year and represented 47% of the auto business. Growth was driven primarily by software-defined vehicle, electrification and connectivity. SDV processor design wins continue to accelerate, including S32 and S32K series platforms. Additionally, we secured new design awards for our next-generation multi-gigabit Ethernet switches, purpose-built for SDV in-vehicle network architectures. These are multiyear platform commitments that expand NXP's content per vehicle. In industrial and IoT, revenue was $755 million, up 38% year-over-year and in line with our guidance. The company-specific growth drivers, which include our newest processing portfolio of i.MX and RT and MCX grew at 40% year-on-year and represented 36% of the industrial and IoT business. Communication Infrastructure revenue was $452 million, up 41% year-on-year at the high end of guidance. Growth was driven by digital networking exposure to data center and continued ramps of our UCODE RFID products. And lastly, mobile revenue was $351 million, up 6% year-over-year and in line with guidance, reflecting normal midyear seasonal trends in our secure mobile transactions franchise. Now turning to our data center exposure. 90 days ago, we quantified this exposure for the first time. To recap, 2025, revenue was approximately $200 million, and we expect to exceed $500 million in 2026. Our position is squarely in the control plane of AI infrastructure, the same domain where NXP has built deep expertise across vehicles and factories for decades, now operating at hyperscale infrastructure specifically in 2 franchises. First, top-of-rack switching to SmartNIC control anchored by our Layerscape family, which is ramping across leading hyperscalers. With every new data plane switch generation as speed increases, the control plane performance must also increase as there is simply more to manage, monitor and secure, hence we are accelerating our Layerscape road map to deliver the control plane performance each new generation demands. Customer engagement gives us confidence that these programs materially broaden our addressable content and extend the franchise well into the future. Second, the processes that control, monitor, cool and secure every component within a rack. Data center infrastructure is converging towards industrial-grade principles where reliability, real-time monitoring control and zero tolerance for downtime is critical. NXP is uniquely positioned as these functions thrive on key industrial processing attributes, where our portfolio is differentiated. Now I want to address something fundamental. AI is moving from the cloud to the physical world into vehicles, factories and robots. It is moving directly into the markets where NXP already has leadership positions. Intelligence deployed at the edge demands real-time performance, ultra-low power and design in safety and security. True physical intelligence also requires distributing AI workloads across multiple layers of the system, an architecture we call the Neural Axis, which is the foundation for deterministic and safe operation of physical AI. These are capabilities, again, NXP has spent decades building. Our differentiated position rests in 3 areas: First, NXP offers the industry's broadest and most differentiated edge AI compute platform. Our portfolio places the right intelligence at the right layer from high-performance recently in coordination in our i.MX and S32N processors to real time reflex and control in our S32K and i.MX RT families, all unified under our EiQ software environment. This has already translated into measurable growth. We estimate AI-enabled processors will represent approximately 15% of industrial and IoT processor revenue in 2026, more than doubling from last year. Second, physical AI is a system problem, not just a compute problem. Intelligent machines must sense, connect and act in real time. NXP is the only company that delivers all of this in one integrated trusted platform something no compute-only competitor can replicate. Third, winning in physical AI requires reaching a highly fragmented market at scale. Unlike Cloud AI, the edge spans thousands of applications and customers across automotive, industrial and IoT markets. NXP's ecosystem of distribution partners, reference designs and field support gives us unmatched reach into this market. Taken together, compute, system and reach, physical AI is already showing up in our revenue, and we expect it to accelerate. Now turning to the third quarter. The operational metrics we track to assess business health continued to strengthen, and our outlook is better than we anticipated 90 days ago. We are guiding third quarter revenue to $3.75 billion, up 21% year-over-year adjusted for the MEMS sensor sale and up 7% sequentially. We expect all regions and all end markets to be up sequentially, a reflection of expanded customer adoption of our differentiated portfolio. At the midpoint, we expect the following trends in our business during Q3. Automotive is expected to be up in the low double-digit percent range year-over-year and up in the mid-single-digit range sequentially. Adjusted for the sale of the MEMS sensor business, our guidance implies a high-teens percentage growth year-over-year. Industrial and IoT is expected to be up in the high 30% range year-over-year and up in the mid-single-digit range sequentially, continuing the strength we saw in Q2. Mobile is expected to be down in the mid-single-digit percent range year-over-year and up in the mid-teens digit range on a sequential basis. And finally, Communications infrastructure and other is expected to be about -- to be up about 50% year-over-year and up in the high single-digit range versus Q2 2026. What you saw this quarter, double-digit growth driven by company-specific growth drivers, and a 35% operating margin is the compounding result of staying disciplined on the right priorities. And now I would like to pass the call to Bill for a review of our financial performance.
Bill Betz: Thank you, Rafael, and good afternoon to everyone on today's call. Q2 was a strong quarter with record revenue of $3.5 billion, up 19% year-on-year and 10% sequentially. All end markets performed above the midpoint of guidance, led by our company-specific growth drivers in software-defined vehicles, electrification, industrial edge processing, connectivity and increasing data center content. Non-GAAP gross profit was $2.03 billion, an increase of $376 million or 23% year-on-year. Non-GAAP gross margin was 58%, in line with guidance and expanding approximately 150 basis points year-on-year and 90 basis points sequentially. Our gross margin performance reflects better product mix, improved factory utilization and solid fall-through on higher revenue. Non-GAAP operating expenses were $794 million or 22.7% of revenue within our long-term operating model. Non-GAAP operating profit was $1.23 billion, up 31% year-on-year. Non-GAAP operating margin was 35.1%, expanding 310 basis points year-on-year and 40 basis points above the midpoint of guidance. Taken together, our second quarter results demonstrate that the margin expansion we are delivering is structural, driven by product mix, factory utilization discipline and operational leverage across our fixed cost base. Below the line, non-GAAP interest expense was $87 million. Taxes were $205 million, noncontrolling interest was $15 million and results from equity accounted investees were a $3 million loss, collectively in line with guidance. This resulted in non-GAAP earnings per share of $3.61, $0.11 above guidance. Turning to changes in cash, debt and capital returns. Our balance sheet remains strong and provides flexibility to invest in our strategic priorities and hybrid manufacturing plans. We ended Q2 with $10.98 billion in total debt and $3.2 billion in cash. Net debt was $7.7 billion or 1.5x adjusted EBITDA, and our adjusted EBITDA interest coverage ratio was 15x. In Q2, we returned $360 million to our owners made up of $256 million in dividends and $104 million in share repurchases. We remain committed to our long-term capital allocation strategy balancing returns to shareholders with disciplined investments in the business to support long-term profitable growth. Turning to working capital. Our cash conversion cycle improved to 129 days from 140 days in Q1. Days of inventory improved to 156 from 165 days inclusive of approximately 9 days of prebuilds for our planned front-end factory consolidations. Receivables were 33 days, and payables were 60 days slightly better than last quarter. During the quarter, we generated $860 million in operating cash flow, which helped fund the $750 million debt retirement, $360 million of capital returns, $174 million into VSMC, $12 million into ESMC and $69 million of net CapEx. Taken together, we generated non-GAAP free cash flow of $791 million or approximately 23% of revenue. On a trailing 12-month basis, free cash flow was approximately $2.8 billion or 21% of revenue. Now turning to our expectations for Q3. We expect revenue of $3.75 billion plus or minus $100 million, which is up 18% year-on-year and 7% sequentially. We expect non-GAAP gross margin of 58.5%, plus or minus 50 basis points which is up 150 basis points year-on-year and up 50 basis points sequentially, driven by the higher revenue and our manufacturing utilization. We expect operating expenses of $810 million, plus or minus $10 million. At the midpoint, this results in a non-GAAP operating margin of 36.9%. Below the line, we expect non-GAAP financial expenses to be approximately $85 million and our non-GAAP tax rate to be 18%. We expect noncontrolling interest to be $15 million including $5 million losses in our equity accounted investees for VSMC and ESMC. This implies Q3 non-GAAP earnings per share of $4.11 at the midpoint. Turning to Q3 uses of cash. We expect capital expenditures to be approximately 3% of revenue with VSMC capacity access fee of $70 million and equity investment of $80 million. For ESMC, we expect equity investments of $30 million. This brings our cumulative investment in VSMC and ESMC to approximately $2.4 billion or about 70% of the total planned commitment across the 2 joint ventures. At the midpoint of our Q3 guidance, the year-to-date revenue was $10.4 billion, up 17% versus the same period in 2025 and consistent with a double-digit growth trajectory. Our growth continues to be driven by the compounding effect of our company-specific growth drivers and the return to expansion of our core franchises. We remain confident that we will deliver on our financial commitments. I would like to now turn the call back to the operator for your questions.
Operator: And our first question is coming from the line of Joe Moore of Morgan Stanley.
Joseph Moore: I was quite interested in the conversation about physical AI and how it's maybe changing your thinking and the products that you have aligned to that. Can you just talk to how that intercepts customer interest areas like software-defined vehicles, are they thinking along those lines? And just how quickly do you think that this transforms into something that's more of a physical AI driven market?
Rafael Sotomayor: Yes. Thanks, Joe, for the question. The whole notion of cognification of the edge, whether it's industrial and auto, conversation is now. I don't think we will get a design win without having a big part of a very, very strong value proposition with respect to AI. So it's happening now. But everything starts before you put AI in the system, everything starts with creating a software-defined system whether it's a vehicle or whether it's industrial. And so I think that's the beginning, right? The beginning is how do you create a software-defined system, what is the platform? And then what is the overlay that you do with AI? It's already happening. I think one of the prepared remarks we said that AI-enabled products in industrial and IoT already composed, I think, 15% of our revenue was around basically AI-enabled products. I think this is going to become even more material and bigger as we go into 2027, but there is simply no actually activity with our customer today rarely without having a really material conversation around how AI is going to get deployed.
Joseph Moore: Okay. And then just on the broader -- on the automotive market. Can you talk about some of the dynamics there. Some of your peers have talked about maybe the beginnings of restocking from very low levels? I know you've talked about Tier 1 inventories being quite low. Are you seeing anything that's different along those lines?
Rafael Sotomayor: Yes. Great question on the auto side since auto, I think this is where our secular story shows up very clearly. You saw in Q2, we're growing into the high teens. This is removing sensor. And I would say that for us, we actually see no restocking, right? What we see and what's driving our revenue, if you look at it, our accelerated growth drivers grew 22% this year, as we stated. They are becoming almost close to half of the revenue. And so the drive there is around content. It's around content driven by an architecture transformation that SDV is doing with the vehicles, and we have aligned a road map to lead this architecture shift towards SDV, which by the way, is still in early, early phases of adoption. So I don't think necessarily auto is about restocking and is it about cycle. I think this is compounding on content and I think we're very well positioned with our SDV road map.
Operator: And our next question is coming from the line of Matthew Prisco of Cantor.
Matthew Prisco: Starting on the industrial side, can you maybe break down the trends you're seeing between the core industrial and the IoT side of the business? And also anything to highlight in this segment from a geographic demand perspective?
Rafael Sotomayor: The industrial -- you saw the industrial growth is quite strong and is growing. And I think in the -- what is it, the high 30s into Q2. The industrial and IoT, the accelerated growth drivers were in the 40% range year-on-year for Q2, Matthew. But also what's happening in the core business is coming back. And you see that our core business also grew in the high 30s. And so very strong growth in industrial and IoT for a total growth for the quarter into the high 30s. And by the way, this is now -- I think that now is the second year and the year that were in the high 30s. This is actually a market that's performing quite well for us.
Matthew Prisco: And then maybe the pricing side, how are you seeing that as a benefit today? And maybe how much of that is impacting the 3Q guide? And how are you thinking about pricing dynamics through the year?
Rafael Sotomayor: Yes. I think your question highlights something that has been on the press quite a lot, which is this issue that we -- that all of us are facing with respect to inflationary pressure and I'll start the answer by saying our first move is always to mitigate the price pressure through operational efficiency. We did make some price adjustments to selected products. These adjustments are not across the board. Price for Q2 was essentially neutral. In Q3, our guide already incorporates an estimate, but we won't know the exact impact until much later.
Operator: And our next question is coming from the line of Francois Bouvignies of UBS.
Francois-Xavier Bouvignies: My first question is on SDVs. I mean, is it possible to get an update on where you are in terms of revenues. I believe you said that it represented $1 billion revenues in '24, and you expect it to double at $2 billion in '27 to reach your targets? And you said that SDV doing very well with high growth. Can you maybe help us quantify it for this year or the growth rate? Anything on that would be great? That's my first question.
Rafael Sotomayor: No. Thank you, Francois. Just to maybe recap what we said, right, the accelerated growth drivers in automotive, they grew 22% in Q2 year-on-year, and they become close to 50% of the total revenue of the company. SDV is the highest growing part of the accelerated growth drivers. Remember, we have radar, electrification, connectivity and SDV. And SDV is the one is performing the best. And obviously, that is the driver for the architecture transformation that has happened in automotive. And we're very well positioned with respect to our road map. What is even more encouraging to think about this is that the current outperformance in auto, and I think our auto business is doing quite well is happening without our latest products, S32N 5-nanometer products, the S32K5, which is our flagship zonal products and 16-nanometer, they haven't even begun ramping yet. And so they're in the design win phase. And so I think we're quite encouraged about the performance of our SDV portfolio because they're more to come with the next-generation products.
Francois-Xavier Bouvignies: And my follow-up would be, again, on the automotive side. I mean, you guided Q3 mid-single digit quarter-on-quarter, if I'm not mistaken, which seems to be like roughly in line with what we have seen before. I mean if I look at your 9-year seasonality, it was even a bit higher than 5%. So it doesn't translate so much the automotive recovery when we look at the quarter-on-quarter pattern if that makes sense? So year-on-year, yes, because maybe you destocked last year, but we don't see a strong recovery than maybe one would expect when we hear TI, STM. Is there any drag we should be aware of that is limiting your growth? Or it just to come or it could come later?
Rafael Sotomayor: Well, Francois, I mean the way we look at our business given the fact that more and more of the revenue in automotive is driven by the accelerated growth drivers as year-on-year because year-on-year, this is where true content growth shows up. You will always have sequentially blurred product ramps. And so I think the better way to look at it is to look at our growth year-on-year. And if you look at what -- where we're guiding into Q3 is a high teens -- is a mid-teens growth year-on-year, which ex-sensor, that is quite strong, and it is above our model.
Operator: Our next question is coming from the line of Vivek Arya of Bank of America Securities.
Vivek Arya: Rafael, historically, NXP visibility has kind of been quarter-on-quarter plus or so. And I'm curious how far does your visibility extend right now? Which areas would you say you have better visibility? And where do you see lead times stretching out?
Rafael Sotomayor: Well, lead times are stretching out and visibility has improved across all end segments. We have better visibility into Q4, we have visibility into Q1. And we've seen that really happen -- basically visibility has improved throughout the year, Vivek.
Bill Betz: Yes. Maybe I'd just add to what Rafael said versus 90 days ago, some of those other health signals that we measure internally. Rafael is absolutely correct. Our backlog continues to grow quarter 1, quarter plus 2, quarter plus 3. So we have a signal of 18 months out. Our distribution backlogs when we look into their books, they continue to follow similar patterns as ours. Clearly, our book-to-bill ratio is above 1, and it's above last quarter again. As lead times will continue to extend and we start to extend them, if you remember, maybe 3 or 4 quarters ago. Customer escalations, which we track in the quarter have doubled since last quarter. So everything, all the signals continue to show even the turn parts that we get, the late orders in the quarter continue to climb as well. So everything much better or continues to improve versus 90 days ago, in fact.
Vivek Arya: Got it. And for my follow-up, I was hoping you could remind us of how you view your typical seasonal pattern in Q4. And given there's visibility, if you could care to give us some directional indication of how it might shape up. And if I zoom out, I think on the last call, Rafael, you mentioned you're still comfortable with the '27 outlook, right, which at a midpoint of $16 billion or so. But at that time, when you had given that outlook, data center was not expected to be a big driver, and now data center is a bigger driver. So is there a different way we should think about the NXP prospects for '27, including the upside from data center.
Bill Betz: All right. I think, Vivek, you're essentially asking me to guide Q4 in 2027. So I'm going to pick your Q4 number real quick. Listen, I think what you're asking me with Q4 is one way to ask me to guide. As you know, we guide one quarter at a time. But I'll leave it with the following. I think we feel very good what's happening right now with our business, right? The signals that we track, they're all pointing in the right way. And I think Bill mentioned some of these signals, right? There are quarters continue to strengthen. Backlog continues to build. And very importantly, for 2027, I think design win ramps are going to climb. So we like the momentum. And I think that we're not -- we're not going to give you a Q4 number, but I'll give you that the momentum that we have of -- into Q3 continues into Q4. And then with respect to '27, I think that the strength of the business and the strength that we have into 2026, it only translates to a strong 2027. And things have improved, continue to improve. And I think our confidence in our long-term growth rate has only increased in the portfolio behind this even better.
Operator: And our next question is coming from the line of Jim Schneider of Goldman Sachs.
James Schneider: I was wondering if you could maybe following up with the prior question, just sort of reiterate the double-digit growth outlook you expressed last quarter on the call in terms of 2027. And specifically, the 60% kind of gross margin level that you expressed. Anything that kind of like changes your confidence there? Or is your confidence, in fact, increased on those targets for next year?
Bill Betz: Jim, this is Bill. There's no change of what we previously said. As you hear, we continue to see things improve on the revenue side. So that's all intact. Related to gross margins, again, we feel very confident of hitting what we shared using our rule of thumb. You can see that play out in our results from a year-over-year perspective, actually do a bit better. So everything intact linked to the higher revenues we plan for and scales very nicely into our model.
Rafael Sotomayor: Jim, let me -- with respect to '27, I think that the right way to think about 2027, right, is the way we think about it, and we're excited about the strength that we have in 2026 is the runway that we have ahead, the opportunity that we have ahead and our ability to compound into it. I think I mentioned it before, one of the perspectives that we take is we have the strength and the newest products in automotive have not even launched. And physical AI is in early stages right now basically the signings and the heavy deployments have not even started. And so I think 2027, right now, we're looking at a very constructive manner, and it really underpins our long-term growth rates that we have established.
Bill Betz: Yes. Maybe I just build on that because what Rafael said about physical AI, I know you shared about the 15% enablement more than doubling. There's another metric we do track, as you mentioned, it's early innings. As you all know, we acquired an asset called Kinara and the design win funnel that we shared with you of the engagement, the excitement, the amount of input we're getting from our customers. That grew last quarter to over $1 billion. I would say this quarter, it's sitting over $1.5 billion in the funnel. And obviously, we'll have to convert those into design wins. But that's a leading indicator. Again, it's an early inning stage for physical AI, and we're excited about it.
Jeff Palmer: Just to add to that, Bill. We have the $1.5 billion pipeline represents over 200 unique and distinct customers. So it's very broad-based, Jim.
James Schneider: And then just a quick follow-up. Can you maybe comment on what channel inventories did in the quarter? What you're expecting for next quarter in terms of weeks, et cetera?
Bill Betz: Yes. No. Like we said in the past, we want to run in that target. It was 11 weeks last quarter. So we feel good about it and make sure we service our customers and get our fair share of market share.
Operator: The next question is coming from the line of Joshua Buchalter of TD Cowen.
Joshua Buchalter: Maybe following up on Francois' question from earlier. You've had a couple of your peers very clearly call out restocking in the auto market. You guys during this past cycle, I think, for good reason, were conservative with inventory on your books and in the channel. I guess is there any reason -- anything about your portfolio that makes it -- a reason why you would see restocking later? Or is there any conservatism on your part that's driving the comment about not seeing restocking?
Bill Betz: Yes, Joshua, this is Bill. I think one of the unique things about NXP is our company-specific growth drivers, where Rafael shared in his prepared remarks, that 47% is coming from this content that is typically, I would say, 3x larger than our core. And so we are in a different area. We play in a different area of auto. And we expect that 47% to growth towards 50% next year. Related to restocking, again, we have a very good handle on distribution. We know exactly what's going in and what's going out, that's serving more broader customers. And specifically in auto, a majority of our Asia customers go through the channel. But in the Western Tier 1s, we track on that and we triangulate it. And as you all know, the Tier 1s, the working capital needs are quite tight. We still see late orders coming in and hand to mouth and margins for them are quite not that healthy. And so they know we have some inventory. You see it on our balance sheet. And they are still providing late orders to us. And so we monitor this very carefully, but we have not seen the restocking effect specifically with our Western Tier 1s.
Joshua Buchalter: Okay. I appreciate that. And then I was just hoping to ask about gross margins in the second half. I think on the previous call, you talked about utilization rates going from the low 80s to the mid-80s. Is that sort of still the right metric to think about as we continue in this up cycle. And as we think about the third quarter gross margin guide, last quarter, you called out some higher, I think wafer access fees that could potentially impact you? Did those play any role in the quarter of the guide?
Bill Betz: Sure. Let me first address. So obviously, our gross margins are doing quite well. From Q1 year-over-year, they're up 150 basis points. Last year, they're going up another 150 basis points in [indiscernible] plan to bring them into the mid-80s so I'd like to confirm that, and that does help the second half of their gross margins to continue to improve. And related to increased cost from foundries and access fees, we haven't seen that play out yet. That's something more, I would think that comes into us maybe in Q4 but more in 2027 when we enter into new agreements with our foundry partners.
Rafael Sotomayor: Sure. And Josh, what we said last quarter was we did see inflationary input costs on back-end type of things. So piece parts, substrates, precious metals, things like that. But in terms of wafers on the front end, we operate within kind of a boundary condition and an envelope. As long as we operate within that envelope agreed to with our partners, we don't see tactical price increases. If we go outside of that envelope, yes, we would see price increases and we would then have to pass that along to our customers.
Operator: Our next question is coming from the line of Tom O'Malley of Barclays.
Thomas O'Malley: I wanted to dive back into the Kinara commentary. So the funnel is expanding going to $1.5 billion. You've seen in the industry a lot of acquisitions taking place, both Synaptics and Hailo. So do you think that these acquisitions are going after that same area that you guys have already kind of targeted. And then when you look at the TAM and how big that can be, maybe could you try to size what that market looks like in a couple of years just because the funnel would indicate it's a pretty large opportunity.
Rafael Sotomayor: Indeed, I think that -- I think you pointed out to basically -- the acquisitions that I think you mentioned is just a confirmation of our strategy, the cognification of the edge is happening. And I think you are not able to play without having a strong AI platform and a strong AI road map. And so yes, I think that's exactly the case. I think we do believe that our platform that we have and the asset that we acquired now has been incorporated into NXP roadmap is best-in-class. We have incorporated now that IP of AI into monolithic integration into our i.MX processors and our S32N platform in automotive. We are discussing how to actually kind of engage with customers on a discrete NPU, which attaches to other platforms, even non-NXP. And not only that, we're developing a very complex AI software framework, which includes now agentic AI, which is going to be the way that the edge becomes completely autonomous. So it's now while other companies are acquiring assets and trying to integrate them, we are trying to evolve into what is going to be the next phase of AI, which we strongly believe is going to be the deployment of AI and agentic AI.
Thomas O'Malley: And then on the auto side, I know that you're saying that you're not seeing Tier 1 bringing back up inventory. But I know that kind of around the last quarter, there was a great variation among your customers where some were well below kind of the standard channel number of weeks and some are well above. Have you at least seen some normalization there where there's been some standardization around that 11-week mark, both an end customer and maybe at disty, you can't comment on the metrics as you will. But -- or do you still see this big disparity where some people really aren't getting it? I just want to see if there's still some normalization yet to come is the genesis of the question.
Bill Betz: Yes. Tom, it's similar to what we've seen in the previous quarters. There is a dislocation between low and high, there's a mix for whatever reason, how they want to control their own working capital and so forth. So no change there. But we track this very carefully. We're just basically -- finally, I think in Q4, Q1, we were finally shipping to real end demand, inventory digestion is behind us, but we have not seen any pull forwards or restocking efforts specifically with their Tier 1s in the Western world.
Operator: The next question is coming from the line of Tore Svanberg of Stifel.
Tore Svanberg: And congratulations on the record revenue. Rafael, I wanted to go back to the physical AI pipeline, the $1.5 billion. How broad-based is that? Is this a few verticals? Is it many different applications? And which sort of application should we assume you're going to ramp the earliest?
Rafael Sotomayor: Well, the excitement is that there is actually broad-based interest on deployment of physical AI or edge AI. And it's happening, whether is think about HMI applications basically where a human controls the device via voice commands and replaces whether mechanical buttons or display buttons, there is -- and the support is -- there's all sorts of use cases around that, there's plenty of use cases around visual and vision. There's plenty of cases of predictive maintenance. The deployment of physical AI is really broad-based. I think we see a very strong interest in Industrial. We saw a strong interest also in automotive or in-cabin type of applications that we're getting pulled into. And so I think the -- really it's going to be such a massive, massive opportunity, massive, I would say, transformation that is happening at the edge. And it's going to be physical AI. We have determined is the driver for content growth in industrial, and it's also an additional content driver for automotive moving forward.
Tore Svanberg: And then as my follow-up, if we think about the current environment, it's a bit strange because you're seeing obviously, great bookings momentum, a lot of demand. And on the other hand, we have these, obviously, capacity constraints with other components so on and so forth. So I'm just wondering, are you seeing any of that potentially impacting some of your customers' demand, meaning they want to do more, but they can't because there's shortages of other components.
Rafael Sotomayor: Well, I think that -- let me answer the question with a specific. I mean you can see our Q3 guide has mobile already down year-on-year, even though it grew sequentially, and our position with respect to market share has not changed, it already seen an impact, right, year-over-year. Mobile is down. And so I would say that you're starting to -- you see it. You saw it in mobile. Now the concern about memory, especially in memory and the constraints that right now the market is, and literally everybody is talking about it. Everybody is trying to actually design around it whether it's different products, different packages, different types of DDR, and I think we are literally helping our customers to actually go and make sure that we help them with the constraints they have. So in pockets, we see.
Operator: The next question is coming from the line of William Stein of Truist Securities.
William Stein: I wanted to ask about the interplay between the backlog that keeps growing at your lead times and your customer intentions, if your lead times were shorter, would you have been shipping more? In other words, are customers wanting more than what you can deliver now and you're sort of -- you're facing some constraints have grown, and that's why the backlog is growing? Or is it that customers feel emboldened given the demand signals they see in the market? And so they're just lining up to place orders with longer duration. If you could linger on that for a moment, I think it would help us understand what's going on there.
Bill Betz: Will, this is Bill. Yes. I mean, I think what you're saying is what we see, right, we do see escalations. You see our inventory coming down. late orders have been coming in over the last couple of quarters. And now they're realizing they have to place orders, specifically in areas where lead times are longer. I mean, it's not our whole entire portfolio, but I would say that it's greater than 16 weeks, a big chunk of our portfolio has extended versus last quarter. And so there's -- you got to place your orders in line. And so we're seeing that play out, but we're still seeing these late orders come in. So it's a combination of both, I would say, but it's going back to more of a normal type of way of how you should place orders with us in the appropriate lead times. But there are folks that are still trying to place late orders and struggle with it.
William Stein: Great. And then one follow-up, if I can. There are some things going on in the world that are normally disruptive to the electronic supply chain, the wars in particular. Have you seen any change in the impact on your supply chain from the renewed activity in Iran or anything else. Any other geopolitical developments? Or are you seeing any effect of that that's different from what we've seen over the last few months?
Bill Betz: No, I would say it's similar. Obviously, we are getting a direct impact on higher input costs. And like as Rafael said, we try to offset those operationally first. And if we can't, we want to protect our gross margins. And unfortunately, we have to pass those on to our customers. So that has been the mode we start to work in, and we started to see this in Q2. We're in the mode of Q3 related to it. Indirectly, I mean, of course, there's probably things that will impact us indirectly, which is more macro. But if you look at the macro indications, both PMIs are doing quite well. GDP has ticked up slightly. Next year's GDP, as you know, is sitting at 3.2% versus today's, I believe, 2.5% or 2.6%. So I mean, it's a balance, I would say. And so we see the macro signals, we see our own internal signals, and we just want to make sure that we're there to support our customers and provide the value for them.
Jeff Palmer: And Lisa, we'll take our last question here today.
Operator: And that last question will be coming from the line of Chris Caso of Wolfe Research.
Christopher Caso: Just a follow-up question with regard to what you had said on pricing. And when pricing was starting to move higher during the last cycle, you guys were very specific about it being neutral to gross margins. Is that the case today? And in terms of what you're seeing with regard to pricing, could you level set us with regard to the magnitude of that? I know you said you're going to have to wait a while to see the magnitude, but you've obviously made some assumption in terms of your third quarter guidance.
Rafael Sotomayor: Yes, Chris, let me tackle the -- I'll let Bill tackle the gross margin piece, but let me tackle the way we account for pricing because I want to be clear, right, pricing is not necessarily what drives our model. The driver is content growth, this architecture-led content growth per system. And so that's the main driver for our revenue. In pricing, with respect to the way right now we're framing it, pricing seems to be a little bit more holistic in the way you're framing the question with respect to just a tactical move to overcome higher input costs. But in reality, pricing is very dynamic, and it's a strategic lever for us, right? We use pricing sometimes to capture value. Sometimes we use it to increase market share and in some cases like the one we discussed right now to offset input costs, but we do price adjustments every quarter. And every quarter, we provide a next quarter guide that has an estimate of that. And so I don't think this time is any different. So Bill, do you want to...
Bill Betz: Yes, I would just add to what Rafael said is on pricing, right, typically, we give an update once a year. In the beginning of this year, we said we'd be down in the low single digits. At the end of this year, we'll update that we'll probably be a bit better, I would say, because again, with the selective pricing we're doing because of these higher input costs. And the related to gross margins, obviously, we want to make sure the value we capture, we pass that on to the owners. And obviously, if there's one that we can offset and it's a higher input cost like inflation, we have to make sure that -- Unfortunately, we have to pass that to our customers and you see that play out throughout the entire supply chain on it. But I think we've been very disciplined here. It is a bit different than COVID, which was more broad-based, and it was a supply issue. This is more of an inflationary issue, I would say, at this time of where we are.
Christopher Caso: Got it. And as a follow-up, if you go a little more detail, you talked a bit about the AI-enabled processors. And how does that value come to NXP? Is it a form of higher content, higher ASPs for the products that drive unit growth? Is it market share, perhaps a combination of all those?
Rafael Sotomayor: I think you kind of answered the question, but I'm going to -- let me -- I'll just rephrase what you said. It starts with the products themselves they have more content, right? So physical AI will drive content from a product perspective. They tend to be higher performance processors, more AI inference content, more software content and more enablement then the system itself becomes also more complex. You have more connectivity because you have AI, you have more security. And in the case of physical AI and robotics, you must have functional safety. And then from -- and then you go in the evolution of what's going to happen right now with respect to agentic AI, now you have a little bit more of a software framework associated with that. So in reality, physical AI for us is a very important driver of content growth and we intend to actually position our road map to be in this market.
Operator: Thank you. And that concludes the Q&A session for today. I would like to turn the call back over to Rafael, CEO, for closing remarks. Please go ahead.
Rafael Sotomayor: Thank you, everyone, for joining us and for your thoughtful questions. I want to leave you with 3 thoughts: First, our growth is structural, driven by software-defined vehicles physical AI at the industrial edge and a nascent data center franchise. Second, we're entering a decade-long adoption of physical AI, which is transforming industries. Through relentless innovation and customer intimacy, NXP is best positioned to meet this transformation. Third, our financial model is scaling exactly as designed. Margin expansion is structural, capital allocation is disciplined, and we are positioned to deliver expanding profitability and growing returns for years to come. The long-term opportunity for NXP has never been clearer. Thank you.
Operator: Thank you so much for joining. You may now disconnect.
