GLOBALFOUNDRIES Inc. (GFS) Earnings Call Transcript & Summary

All right. Welcome back, everybody. I'm Joe Moore from the Morgan Stanley semiconductor team. Very happy to have with us today the executive team of GlobalFoundries, Tim Breen, CEO; Sam Franklin, the CFO. Thank you, guys, for being here.

I guess maybe we could just start. You described the company being at an interesting inflection point on the last earnings call. What did you mean by that? What has you excited about the direction that the business is headed?

Yes. Thank you, Joe. Thanks for being here. No, it's a super interesting moment to be GlobalFoundries. And I think it's because we see 3 trends that perhaps we've tracked for some time, but now moving to different pace of momentum. And the 3 that I see, 2 are demand and 1 is on the supply side. On the demand side, clearly, the rollout of the data center, we're seeing only acceleration in the requirements, the scale, the deployments globally, but also some of the bottlenecks those data center deployments create, whether that's power or networking areas that we've been investing in our portfolio for some time and now seeing that pull-through in the business and the pull-through for the road maps that we have for the future. Even more excited about the longer-term trends of AI entering the physical world, right? When AI is around us in the cars we ride in, no longer driving because they'll drive themselves, the home environment, the workplace, the factory, that physical layer transition, we're starting to see that have real roots and pull-through into engagements we have with customers. But the third piece of it is how we were born is doing all of that in a globally secured footprint. And that was hard for a lot of our journey because it's not easy to run fabs in different places. But now that's not a nice to have. it's a hard requirement for a lot of these applications. And so when you put demand and supply trends like that at the same time, it really positions GF in a way that we've never been positioned in our history.

Yes, that makes a lot of sense. Some of the recent M&A that you guys have done and talked about, you've talked about being a full spectrum RISC-V process provider, things like that. How do you think about that M&A? And where is the balance between developing IP that your customers do and things like that, how do you think about those dynamics?

We've tried to ground our M&A strategy very much in our customers' requirements. And so we've said, where can we buy businesses, companies, teams that add differentiation to our technology footprint that deepen those engagements with those key customers and actually enhance our global manufacturing base, 1 or more of those 3 pillars. The recent M&A have done exactly that. And you picked on the topic of kind of IP, custom silicon and software. As you know, we acquired MIPS in the middle of last year. We announced earlier this year the acquisition of Synopsys' ARC business. And this is all centered around providing more to our customers than manufacturing alone. We provide processor IP, a range of cores for different applications. We give them an alternative in the ecosystem so they have choices. But we do it in a way that they can trust because we're a neutral party, right? We're not a competitor to them. We're not a supplier to their competitor only, right? We're across the broad spectrum. It allows us to engage with them earlier on their architectures, earlier on how they think about design and it leads to enhanced foundry business as well as IP, software and customer engagements as well.

Great. And you -- before the IPO, you had a decent sized custom silicon business, which later on became Avera acquired by Marvell. Can you contrast that with what's happening? It doesn't seem like you're doing that. It seems like you're more design services element, but how do you think about that dynamic?

No. And I think that's a great question. I mean Avera goes back now '20 sort of '19 time frame. That was a business that was doing 5-nanometer designs in 2019, right? So that road map is highly oriented towards applications that GF couldn't add value to that point, and we said we are not the natural owner of that business, but the thesis that says being a differentiated IP provider, especially for process applications in that edge AI, physical AI kind of domain plus being able to build custom designs in process nodes that take advantage of those technologies like our FD-SOI platform, FDX and so on, you've got a great synergy there. And so MIPS is a better fit for that relative to where we were in the past.

And is that designed to grow the foundry business? Or are there other monetizations that you can do around that?

So you can think of it in really 3 buckets and then a sort of knock-on effect, right? So bucket one, direct IP sales, this business -- both businesses exist. They have direct IP customers today, and that will continue to grow. Both businesses, again, I'm talking about Synopsys as well, have software tools that are used today. Synopsys has over 300 customers in that business that we're buying, using tools like ASIP designer in the case of Synopsys. We have a tool called Atlas Explorer in the case of MIPS. This allows designers to build structures in silicon in virtual silicon, well before they're actually putting things into the real world. And so that software business is also there and growing. And then we have more and more interest in customer engagement. By the way, with existing GlobalFoundries customers, but also new customers who say, look, I have a particular use case for my medical application. I don't have my in-house design team. I don't have in-house IP. Can you help me put that together? And we can do that working with them, but also bringing third parties to some of the design activity as well. So all of that leads to a different set of monetizations. Of course, a lot of it leads to additional wafer volumes as well down the road.

I mean there was a time when people use GlobalFoundries because they were cheap, right? And then really, even by the time of the IPO, you had started to establish much more of a partnership mentality and people -- the geopolitics, I'm sure helps. But it seems like this is another aspect that kind of brings you closer to the customer, it makes you more central, not like a secondary supplier of foundry services?

I think that's fair. And I think it's about do you bring differentiation? And do you help earlier in that stage of development so that when they're trying to crack their own product competitiveness, you play a role in that. But by the way, it also helps you exploit your process technology differentiation even more. What we learned in that journey is that it's one thing having a different process technology that may be actually superior, but not necessarily the biggest in the industry, while actually having the design skills means you can figure out how to use it. So our FD-SOI platform today is dominating the field of high-performance radar. Why? Because we've been working with these customers now for many years to build that capability to how to use that technology the right way. So I think this fits with this idea of being highly differentiated.

Yes. Okay. Great. Geopolitics has obviously been front and center for what you guys have been doing for a decade now, really the desire to diversify away from Taiwan specifically, but also the desire for different regions to localize, and you guys have been a U.S. company, but with fabs everywhere. Can you talk about the importance of geopolitics, the importance of government subsidies? And how that helps you build those relationships out?

No. I mean it's obviously a trend that we read about every single day. I mean we live in an increasingly fractured world in many ways. The old order of globalization with no guardrails, I think, is an order that I think no one would say is coming back anytime soon. But it means that customers in the industry have to navigate risk and find ways to mitigate those risks in the business. And they actually want optionality, right? They don't just want to supply security, but they want choices because who knows what happens in the future. Maybe their business grows in one region more than another, and perhaps they want to be able to localize content. And so we said it's not just enough to build fabs in different locations. You need to qualify technologies across those fabs, so you can choose, I tape out my FDX chip, but I can make it in 3 fabs or 4 fabs around the world. So you get them that optionality of that for their business security. And of course, going alongside that is you build deep partnerships with governments. And we've had that, of course, in places like Singapore for many years. Germany is bigger supporter of us in the EU more broadly. And then, of course, the U.S. has now accelerated also its support for our industry. So that levels the CapEx playing field, meaning when we invest a dollar in the U.S., we're bringing more than $0.50 back in terms of government support. That allows us to get scale where previously it was hard to do so.

Okay. Great. Can you talk about the current economics of the business? We just had the TI CEO downstairs talking about, okay, the China foundries are starting to be full. We're seeing hotspots. We think that this could become a very tight environment for foundries, which would obviously benefit them, but obviously would also benefit GlobalFoundries. Do you see that? And I guess part of that is TI moving some of the older nodes to support substrates for CoWoS or just to support more advanced nodes, that seems like there's actually supply being taken out of your business. Is that fair?

Yes. Look, I think if you take from a year ago to today, every end market is incrementally tighter, some more than others. For sure, anything that's touching the data center is in high demand and visibility is very strong. Purchase orders are talking about '27 in many cases, not '26. And so you have different levels of visibility. But every part of the industry has basically seen a tightening. And again, some of them are on the supply side, like you mentioned, TSMC strategy changing and some of it, of course, is demand. And then, of course, there's idiosyncratic factors, where, for example, let me give a real example for GF. We have a great business in high-performance silicon germanium. It's been around for some time. But now that, that's getting pulled through on data center applications like within a pluggable optical transceiver for a TIA or a driver, suddenly, we see excessive demand in that we're able to invest with confidence to then supply that demand.

Can you talk about silicon photonics in more detail in optical? We've -- the Lumentum and Coherent were probably the 2 hardest rooms to get in this week at the conference because of the deal they struck, but very clear that this is a key bottleneck to data center, and you guys are very well positioned to help that.

Yes. It's interesting because a few years ago, it would be hard to talk about photonics with anyone to be interested. But in a way, this transition has always been underway because we'll be moving more and more data for more and more complex workloads. Obviously, AI accelerates those workloads dramatically. You add to it that in the data center, the density of the data center has changed dramatically, right? You look at the number of GPUs per rack, racks per pod, pods per data center. This is now a very dense environment, where a lot of data and power, by the way, has moved around on a daily basis. And you start to hit physical constraints that need to be solved in different ways. How are you seeing that in the photonics demand in the market today? You're seeing a really strong business in pluggable optical transceivers. If you walk around a data center, it will be full of pluggables in every different application scale across and scale out in particular. But what you're also starting to hear about is the growth of co-packaged optics, and I think a lot of this week's announcements were about that because at some point, at that rack scale, there's not enough beachfront to put the optical transceivers in place in a pluggable format. You need to move to something called package inside that CoWoS on the substrates or on the interposer. We've been working on that for a while. Again, sometimes I think we're early, but we're right and it's better to be probably early and right than late or later wrong. But those applications in co-packaged optics have taken time to develop. You need to integrate more components. You need to bring silicon process technology, which we have but also advanced packaging so that you can bond an electrical on the photonic IC. You need to understand micro optics, so partnerships like the one we have with Corning about delivering a detachable fiber connected to that board. You have to solve a lot of different challenges, all of which we've been working on for some time. And so now we see the proof points in that in, okay, pluggable demand today, but also strong demand for the road map in the future of moving to co-packaged optics. We've said consistently that '27 would probably be the inflection point of CPO. I think we still believe that's true. I think what people are maybe getting a bit more incrementally bullish on is just how fast CPO will penetrate, particularly for those scale-up applications. And for avoidance of doubt, pluggables is not going anywhere. Pluggables remain very important, but the scale of application is just becoming huge in terms of demand.

I mean you guys were talking about doubling this year, right, before we even get into that.

Exactly right. I mean you look at the revenue from 2025, that was roughly a doubling from 2024 to a little over $200 million. The expectation for this year is that we've got good line of sight to double again. And to Tim's point, a very large percentage of that, in fact, the vast majority of that is from pluggables. But the early stage ramp and the early opportunities and tape-outs that we're doing with customers is to then support the growth from a CPO perspective.

And how would you handicap your guys' capabilities versus other foundry competitors? I know you've done the acquisitions of AMF and InfiniLink [indiscernible] talked about this as well, other -- certainly, this is going to be an obvious opportunity for everyone. Where do you stand?

So I think we think about leadership here in 3 main categories, right? First of all, how strong is your technology today and your road map for the future. And because this has moved beyond just silicon innovation, you have to think about packaging, you have to think about how you put the whole system together. A lot of the innovations we've been investing in are actually allowing us to have a road map that can serve the 200 gig per lane kind of bandwidth today, which is what people talk about generally, moving to 400 gig per lane probably as fast as anybody, if not faster, from an industry sort of development point of view, that enables those 1.6T transistors, the 3.2T transceivers and so on going forward. So that enables the next generation of performance, both in pluggable and co-packaged. The second piece is building an ecosystem and the enablement structure. So we have an extremely robust PDK, we've been developing for some time. What does it mean? It means customers can design quickly into your process with standard product, standard sales, standard IP that they can use. That's something that others haven't had time to do and others haven't put in place. Some of the partnerships on test, assembly, detachable fiber connect we talked about as well. That ecosystem is super, super important. The last piece of it is manufacturing footprint. So back to where we -- our roots in a way. We're doing this in 300-millimeter manufacturing today in the U.S. And so when it comes to security and scalability, that's hard to beat, right? So it means we can meet the demand of the industry as it grows with CapEx efficient and short time to market.

And are you guys -- when you talk about that 100% growth this year, are you guys going to be supply limited with that? I know there -- it seems like there are other bottlenecks that impact the growth on you.

Our internal job is to try and match supply and demand as close as we can. I'd say right now, they're pretty closely matched. But we're definitely seeing incrementally stronger demand signals. And as a result, we are also increasing in some cases, our investment. What I like about the investment side, again, in the short term, this is adding tools to existing fabs. So you're able to get time-to-market advantages. It's very capital efficient. From a long-term point of view, we're also getting the support back from the government side on that. So I'm more confident to increase the investment as the demand signals come in. What I also like about the demand is it's not 1 or 2 customers with 1 or 2 ramps, right? It's broader and broader. And in fact, the acquisition of AMF brought a whole suite of customers to GF that we didn't have before or didn't have significant scale out before, and that's giving us a much more diversified ramp as well. So I feel like we'll be able to match supply and demand quite well. Of course, that's what we think about every day of the week.

And you -- it seems like investors are starting to get the message on the importance of this. You said a lot of your questions and meetings are around this opportunity.

Yes. I think like everything, it's a question of the technology story was always quite clear, but you need to see the revenue numbers. And obviously, we've talked about that more, and that I think gives people confidence that now we can set a goal of being a $1 billion by 2030. In our last earnings call, we talked about that being run rate by the end of 2028. And so I think we have incrementally significantly more confidence on that revenue trajectory.

When you start to see these scale-up opportunities just emerge from nowhere as tens of billions of dollars, pretty impressive. Okay. Great. On the other parts of your business, consumer-centric, you have still a pretty significant mobile business. There's angst about memory issues in that part of the market. But the biggest customer seems to be okay from what we see so far. Second biggest customer has a memory business. How do you feel about mobile? Do you anticipate there being headwinds over the course of the year?

So look, bigger picture, mobile is still a significant part of our business, but we've been growing substantially in the other parts. And our strategy has been to develop a much more broad base of end demand and definitely data centers played a role in that. The auto story we can come back to if you want. The auto story has also been a really good multiyear growth story and double-digit growth rates, which has made mobile smaller than it's ever been as a percentage of our business. I think overall, we'll track the overall mobile market. Our portfolio, if you kind of say where are we most geared, it's more geared to the high-end phone. That's kind of obvious in a way because high-end phones have more content upgraded components but think of it in RF front-end module for 5G, think of it display, haptics, audio, power management for the battery life, all of these components. And so we haven't seen tactically any significant customer changes of behaviors and actually we came off a relatively solid kind of end of year launch cycle and so on for some of those phones and those customers that you alluded to. So I'd say overall so far, kind of tentatively good. I think, look, it's worth watching closely what happens on the lower end and do you start to see incrementally more nervousness on there. So look, we'll track the market overall. Look, longer term in mobile, what's been working well for us is really focusing on those differentiated applications that are multiyear cycles, not the tactical annual cycles where our content is particularly distinguished. And then, of course, getting involved in new form factors. And we're hearing incremental bullishness on smart glasses and application that requires some similar ICs but also some different things like display and so on that, again, we have good technology for.

Yes. I mean the only place we've actually seen softness, there's a lot of concerns about memory and smartphone is really China Android and -- which should be pretty [indiscernible].

Relatively small.

Yes. You mentioned auto, which has been a great story really for years now. I mean you came public with low single-digit percentage and you've grown it every year a lot. You recently just extended your partnership with Renesas. Can you just give us a sense of what's happening in auto long term? And it does seem like that's one of the markets that's most focused on sourcing in the region.

Auto is, in a way, that perfect combination of the 3 things we always talk about, right? The technologies we bring to bear are particularly suited to auto, and it's not just microcontrollers, it's also battery management, in-cabin sensing, radar, LiDAR, kind of a broad spectrum of semiconductors, more than 90% is in areas that we can service. The single-digit nanometer part is very, very small. So very, very good technology fit. We've been able to work deeply with those customers now for multiyears. By the way, those design wins take time, they ramp slowly, but then they ramp and they stay for a very long time. And actually, the supply crisis of COVID allowed us to also partner with the Tier 1s and the OEMs. So we spend time with the whole value chain. And they are instrumental in influencing the sourcing decisions by the way, partly because of trust and technology, but of course, the geopolitical factors play a big role. And so look, auto has been for us a secular growth story. We've grown -- we've outgrown the market every year as a public company, and we've actually had our best year of automotive design wins last year. Despite everything in terms of growth, we're seeing more momentum come into the market. So it tees up the discussion about Renesas. Renesas, a great company, broad portfolio, 3 areas I think are particularly interesting in their portfolio. Of course, automotive is strong. They're ramping in data center power, which we talked a little bit briefly about. And then they have a broader sort of edge AI microcontroller type applications for different nonautomotive use cases. All of those, again, feed very well into our technology mix and as a supplier with many customers around the world, they said, I need a partner who I can work with globally. I want U.S., I want Europe, I want Asia. I want optionality in China with our manufacturing partnership we have. And actually, even would like maybe some optionality in Japan, and we've agreed to do that with them. And so you've got a situation where it would be very hard for anyone else to deliver that kind of global sourcing in the technologies that business is focused on.

Yes. I mean, I was really struck a few years ago by the automotive OEMs striking deals with you guys where I don't think they knew who you were 2 years before that. How persistent has that been? I mean have you -- obviously, the shortages are not as much in main memory as they were. But we actually heard from [indiscernible] yesterday that the OEMs are asking the Tier 1s to build inventory, Tier 1s say, no, so the OEMs are starting to stockpile inventory so that there's a buffer. Are you seeing that OEM relationship still be important to your business?

I think on the supply chain side, it's maintained extremely high level of importance, right? We have very regular dialogue on planning. Again, they have a little bit more visibility from their side versus the Tier 1s. By the way, we spent a lot of time with the Tier 1s, too. We've been very public about AUMOVIO, formerly Continental, Bosch, Hyundai and others -- Mobis and others that have worked with us. And so I think our breadth in the Tier 1s is also quite strong. I think, look, what they're all thinking about is what do they want to own from an architecture point of view and design over the long haul. Of course, they look to the cases of companies like Tesla who've had a much more kind of silicon native structure and many have been working on that. And by the way, that brings us back to topics like processor IP. And so for example, having access to IP for processor designs, that's again another vector of engagement with the Tier 1s and the OEMs that was perhaps harder for us to do in the past when it was all about manufacturing. And so look, I think this is going to deepen. And those end up becoming multi-company partnerships because people play a different role in bringing the systems together.

That's helpful. Moving to the financials. Your gross margins have continued to go up. You have a long-term target of 40%. Can you kind of tell us what has to happen to get to that 40%?

Sure. And look, I think the important metric that we're focused on now is the relative progression from a gross margin perspective. And if you look at the fourth quarter and do the comparison to same period the year prior, we delivered revenue that was roughly flat. We delivered margin that was almost 4 points higher. So why is that the case? A couple of fundamental drivers that Tim has alluded to already. One of those is the mix within our business. And there's 2 layers to the mix within our business. There's the mix as we think about it from the end markets that we're servicing and the secular drivers within those end markets and there's the evolving mix as it relates to how the acquisitions we've made over the last year are starting to evolve from both wafer revenue services as well as technology revenue services as well. MIPS is a good example of that. The Synopsys IP acquisition will be another good example of that as well. So really, it's that mix evolution that has been a strong tailwind. And by the way, we've shown that gross margin improvement, notwithstanding the fact that on a year-over-year basis, the legacy of underutilization payments had actually fallen out of that revenue stack as well. So continued mix shift is key. If you look at fourth quarter combination across comms infrastructure and data center and our automotive business, that's 35% of our revenue, okay? You take that and you do the comparison to smart mobile in the fourth quarter, that was 36% of our revenue. If you rewound the clock to prior to when we went public as you rightly said earlier, Joe, automotive was in the low single digits. Smart mobile was greater than 50%. The mix shift doesn't happen overnight. It's been a multiyear journey, but it's a multiyear journey with very disciplined investment that's driving incremental margin fall-through as well. The second piece of the story is really how we think about the scale opportunities across our business. And scale matters in foundry, right, particularly given that we're at the highest fixed cost end of the entire semiconductor ecosystem. And so we have, again, focused our investments over the course of the last few years to really kind of drive that scale across our Dresden footprint, across our Singapore footprint. The opportunity for us as it relates to scale is now principally in the U.S. It's scaling and remixing our Burlington, Vermont facility and it's scaling and remixing our multi-facility in Upstate New York. And again, we've been really driving towards that for the best part of 1.5 years now. Malta was traditionally a FinFET facility. Well, now we've been transferring technologies in. We have FinFET. We have silicon photonics. We have our FDX technology. We have a 40-nanometer automotive qualified corridor as well. So we are driving scale across our footprint, which again is a tailwind as it relates to the broader margin story as well. And then the final piece of it is really around productivity. At our core, we're a manufacturing business. And so continuing to focus on productivity is critical. Our cash cost per mask layer matters, all the components that go into our manufacturing processes and where we look to drive incremental margin growth as well. So what I will say is that all of the strategic decisions you've seen us take over the last year as well are not because we view 40% gross margin as a final destination. We view that as next stop in the journey as we continue to focus on this evolution of our business as a holistic technology services provider. We think that creates more opportunity over the long term.

Okay. That's helpful. You talked about a little bit on the earnings, a meaningful increase in capital intensity this year. How should we think about that?

Yes. Again, Tim alluded to a little bit of this earlier. We've made a really concerted effort, particularly as it relates to CapEx intensity over the last couple of years. And more importantly, how we think about it for 2026. The first core principle is debottlenecking, making sure to your earlier question, that we have the capacity that's needed to meet the demand that our customers are telling us that they need. A couple of great examples of that are everything related to silicon photonics in the pluggable space, our high-performance silicon germanium, which is up in Burlington, Vermont, but also our FDX technologies as well. Our FDX technology is one of the most versatile that we have across the business as it relates to all of the end markets that we service. So debottlenecking that demand key. And that's really where we see a significant envelope of growth from a CapEx point of view this year, but it's because the demand is there and it's accretive demand to our long-term margin targets. The second piece is how do we do that capacity growth and deploy that CapEx in an efficient manner? And from a GlobalFoundries point of view, we've got a footprint that allows us to grow into the existing 4 walls that we have across most of our sites. I mentioned the remixing going on with Malta, opportunity to continue growing within the 4 walls. The same is true as we remix Burlington. In fact, in Dresden, we have our legacy Bump-Test-Facility, we're going to convert that into incremental manufacturing capacity as well. And as we ramp in AMF, the silicon photonics business we acquired last year, again, using the existing 4 walls. So that efficiency really matters. And then the final piece, Tim kind of mentioned earlier is capital efficiency and partnerships with governments across all of the geographies that we operate in as well as partnerships with customers as well. There's an increasing focus from customers that demand certainty matters in some of these critical technologies that we're putting on. So if I take the government as an example, Tim mentioned earlier, in the U.S. we can expect upwards of $0.55 on the dollar. I think you can assume a rough rule of thumb that anywhere between $0.30 and $0.55 on the dollar is what we target across all of the geographies that we operate in. So I'd say they are really the 3 core components that we think about from a capital intensity point of view. Now the last 2 years, we've been well below our target model. Our target model is net CapEx sort of 20% of revenue. We've been operating in the roughly 8% to 10%. So actually, it's natural that you see a bit of a float up this year and I indicated on the call that I expect that to be more like 15% to 20% this year. But for all the reasons I outlined, we feel quite good about making those decisions.

I mean one of the nice things about the businesses you're in is the relatively low capital intensity compared to the guys competing gate-all-around and the cash flow that the business throws off. Can you talk about the priorities for free cash flow moving forward?

Absolutely. I mean, look, we've been on a relatively early journey in terms of becoming quite a high generative free cash flow business last year, close to $1.2 billion a year before, over $1 billion that we want to make sure that we're maintaining this flywheel of free cash generation. But we also have a very clear hierarchy about how we think about what to do with that cash generation. We want to make sure, first and foremost, that we're investing in the business in high-growth, high-margin opportunities. That's investment from both an organic perspective and an inorganic perspective. I mentioned the technology corridors where we're really focused. Inorganically, Tim alluded to earlier, the acquisitions that we've made across the last year as it relates to MIPS and AMF. The broader opportunity that comes through with the Synopsys IP business as well. Again, really focused on investing in the business for growth in high-margin areas. Secondly, we want to make sure that we have a very robust financial position. Balance sheet is key. We want to continue making sure that, that's strong. We ended the year with roughly $4 billion of cash on the balance sheet. We have another $1 billion of unutilized credit facility as well. We feel quite good about the strength of the balance sheet at this point. And then finally, it's how do we look to thoughtfully look to redistribute some of that capital back to shareholders as well. Look, we're going to do that in a very orderly manner. As you heard from our last earnings call, we've had authorization from our Board to start with a $500 million buyback. That buyback is format neutral, whether it's privately or through public market buybacks as well. But I think it's a good endorsement as it relates to the long-term opportunity and the resilience and the discipline that we have within the business.

Very helpful. Let me pause there and see if we have questions from the audience. If not, maybe just we have 3 minutes left. Any final remarks. Can you close that. What do you want investors to focus on the most when it comes to GlobalFoundries?

Yes, I think the way I think about this is preparation meeting opportunity, right? And it doesn't happen overnight that you position a global manufacturing footprint with flexible technology. It doesn't happen overnight that you develop technologies that just have breakthrough performance for connectivity in the data center, for power management in the data center. It doesn't happen overnight that you build relationships in automotive to be able to grow double digits year-on-year for multiple years. And so we've been working on this for a while. I think what the inflection point is kind of where we started the discussion is you now see this driver across everything in the industry called AI. And of course, we internalize that mostly as this virtual cloud-based interface, but you're hearing at this conference a lot more about agentic, you're hearing about physical. And so we're at the beginning of the secular shift of how technology is driven and that's going to pull through a huge amount of content in our industry. And GF has been preparing for that for some time in terms of positioning the portfolio as best we can. Of course, things move faster in some areas and slower in others. But we're at a really incredible point where we can now deliver on that model. And as Sam said, we've got the financial firepower to do it. We've got the accretive model where every dollar of revenue comes with significant fall-through and higher margins. And so I feel like we're at the beginning of a really good sort of virtuous cycle for our company with not just one demand driver. I think the last piece of this is there are lots of ways to place semi, but betting on one thing. I think what's great about our model is it's betting across that whole kind of technology stack along the bottom, the benefits.

Great. Well, thank you very much. Appreciate it.

Thank you very much.