Hello, everybody. Welcome back to the Deutsche Bank Technology Conference. We will kick off our final afternoon session today with Jon Kemp, CEO of Qnity. I'm Melissa Weathers; I'm one of the lead semis analysts here at the bank. Jon, thank you so much for being here. This is your first time at our Deutsche Bank Technology Conference. It is. Thank you. We're happy to have you here. Wow, some applause. I think this is a technology conference. Obviously, Qnity has had a very interesting last nine-12 months. You've been very busy. I think, with the tech investors in the room, it would be helpful to start out with, talk about what Qnity is, talk about the reasons for the spin from DuPont, and just kind of level set us on what a tech investor can look forward to in owning shares of Qnity. Yeah, that's a lot. Thanks, Melissa. Look, I'm really excited to be here. It's great to see all of you. We've had a great day talking to a number of investors all day today. Look, we're about nine months old as a company now, and we've had a terrific nine-month start. We sort of picked a perfect launch timing. If I had to pick any time in the last 30 years to have launched a semiconductor company, it would've been the timing that we just had. We've had an exciting nine months. When I think about Qnity, what makes us unique is really we're the largest pure-play end-to-end solutions provider across the semiconductor ecosystem, with materials that run the full gamut of the semiconductor stack, from front-end semi-fab materials to the middle of the line in the advanced packaging materials to back-end, high-value assembly and AI PCB materials. That affords us the opportunity to capture and benefit from leading-edge technology, working with some of the world's largest and most respected technology companies, where over decades, we've earned a seat at the design table, working with them to enable their technology roadmaps. We're really well-positioned to capitalize on the AI-led transformation that we're seeing, whether that's in investments in the data center or increasingly investments in other parts of the AI-led transformation, across both high-performance computing and advanced connectivities. What I like to say is AI is changing the unit of innovation and how things get built, and materials innovation is really the hidden hero of the AI era in the sense that it's enabling kind of how those designs and how compute can be delivered in new and innovative ways to customers around the world. We support those customers with a local-for-local operating model that has allowed us to co-locate our innovation and manufacturing capabilities in the major centers of geography for this industry, whether that's the U.S., Taiwan, Korea, Japan, or China. Within DuPont, this business thrived and did really well for a number of years, but it was sort of hidden in the corner of a company that was going through a lot of portfolio changes over the last decade. So it was relatively undiscovered, and I'd say it's still relatively undiscovered. When I think about the things that are fundamentally different, certainly the advantage of the strategic focus of being a pure-play company is first and foremost. But really when I think about what this industry requires, it's really a different sense of the pace of play. When we were part of a larger, somewhat more bureaucratic organization, both the risk appetite and the pace of play were challenging. Our ability to operate, to kind of step-change the clock speed with how the company operates in lockstep with our customers, has been refreshing to us and to them in terms of what we've been able to do already in terms of affecting a culture change within the organization. The different portfolio moves that were happening inside of DuPont also became a too internally focused, where we wanted to reorient the customer with a much more customer-centric mindset. As we came out, customer and innovation and speed became three of our core values. We've really taken that to heart and really tried to drive that, and I think that culture is underpinning some of the success that we've seen over the last nine months as we've been able to capitalize on the growth momentum in the industry. That's great. I think that's a great setup. I guess if as we think about governance and capital allocation, now that you're out of. I feel bad; I sit right next to our chemicals team, and I think there's a different tone between the semi side and the chem team. But, talk about from a dollars perspective, being an independent publicly traded company, what kind of decisions, I know we've talked about CapEx in the past, R&D investment, those kinds of decisions, what kind of value can you guys unlock as a standalone company, just from a governance side? Yeah, it's a great question. Maybe I'll use a couple of examples. Certainly, it gives us more strategic flexibility to act quickly. One example that I'll use: we announced in the first quarter of this year two fairly significant capacity investments, one in Delaware and another one of a new building that we acquired in Taiwan. One of the biggest constraints in the industry has been clean room capacity. This facility in Taiwan, it was less than 100 yards from our existing Taiwan manufacturing facility, and it was a five-story building with three floors of already built-out clean room capacity. Now, it was a $60 million price tag. That's what we ended up paying for it. We didn't know what the price tag was going to be, so we needed some flexibility to negotiate that. I think previously, that would've been a six-month process in order to get authorization all the way to the board level to go through that. We did it in 10 days. We had the purchase contract signed in 10 days. So much more flexibility to time the strategic investments and operate with the speed that this industry requires, because there was a building that had come up before that, and literally it was gone in a week. Because that's how fast this industry moves, particularly in Asia. The other thing I would say is we were fortunate enough to do a couple of acquisitions even while we were part of DuPont, but doing acquisitions was always a little bit of a challenge because there was so much other portfolio activity that there was a pull on the resources. The same resources that would do the acquisition were doing some of the other portfolio moves, and so there was this constant tension of, is now the right time? Not to mention, when you're trading at an industrial multiple or a chemistry industry multiple and trying to do acquisitions in the semiconductor industry, the multiple divergence makes the economics a bit more challenging. So now that we're a pure play, we've seen some revaluation that has already happened. I think there's more opportunity to go further. But certainly that unlocks a little bit more strategic flexibility for the economics that would make sense in our industry. Our portfolio has been built through a series of really smart acquisitions over time, and we continue to be interested in leading technology positions that would expand the breadth and depth of our portfolio, especially in fast-growing areas like advanced packaging and thermal management. One more kind of high-level question, just on some news flow that happened last week and a couple of weeks ago. You have a new partner; you have just announced a new CFO and then also a new head of the, I think, the semis business. Correct. Any public comments about your excitement on those hires? Yeah. Really thrilled. Kate Dei Cas is the new President of our semiconductor segment, and Ken Rizvi was announced as the new Chief Financial Officer. Both of them bring more than two decades of semiconductor ecosystem experience. Kate, kind of when it came up in the Air Products, Versum, Merck environment, or EMD Electronics, as sometimes they are known in the U.S., came up from the shop floor as a process engineer, is very familiar with all of our customers, their process technology, and the way the semiconductor ecosystem works. She is an operator at heart. What is common about both of them is they are a great culture fit for our organization. Like when you do acquisitions, most acquisitions fail, not because they got the economics or the systems right, but the cultures do not fit. I would say that is the same thing with executive hires as well. It can be great on paper, but if the culture does not fit and there is not strong alignment in how we operate and how we want the company to run, it is not going to fit. What I love about both Kate and Ken is it is a great cultural fit on top of having all the—Ken has more than 25 years of semiconductor industry experience from large companies like Micron. He has been up and down the semiconductor ecosystem, worked for an OSAT company in Asia and Singapore. He worked for a back-end company in printed circuit board materials. He has worked in both 10 years as a public company CFO, a lot of transformation experience during his private equity days, a lot of acquisition experience and portfolio management in some of his other stops. He is currently the CFO at Synaptics. Obviously, they are in the middle of a transaction with onsemi. They should close in the middle of next year. Kate has already joined and has been with us for about three weeks now, and Ken will join us officially on October 1st. Very exciting. Hopefully, you will all get a chance to meet Ken soon. Yep. He should have joined earlier so you could have joined. It would have been great. What I would say is we had two interim leaders, Sam Ponzo and Michael Goss, who are stepping into their prior roles. Sam was our Chief Commercial and Strategy Officer. Mike was our Controller and Chief Accounting Officer. Both executives are staying with us. This is definitely addition to the bench strength of our executive team because both of those executives will continue to play a prominent part and role in the company going forward. They did a terrific role during the interim period. Onto the fundamentals of the business. One of the framings that I like that you guys have is the shrink-to-stack transition. Can you walk us through what you mean when you say the world is moving from shrink to stack? It's my favorite topic. Thanks, Melissa. Look, I've been around this industry a long time, right? We all know that the pace of progression in this industry has been powered by a single idea, and that's more performance in less space and usually at lower cost as well, right? With every node migration of Moore's Law, we got more performance from smaller space. I call that the power of shrink, right? The power of shrink still matters, and it will continue to matter for a long time. But the economics of shrink and the complexity of shrink have increased to the point that on its own, it can't keep up with the pace that AI demands, with the computing demands of AI. That's why I think we've come to this alternative architecture, and it's really not an alternative because it's both at once. This idea of not just shrink, but shrink and stack. What do I mean by the and stack? And stack happens within the process of node progression and node complexity as you get more layers, whether that's a NAND architecture or a logic architecture or even an HBM or a DRAM architecture. You get a layer multiplying effect and vertical scaling happens. But more important than that, stack really refers to the ability to combine chips, GPUs, and CPUs with memory. In the future, it may be with other components as well, analog and discrete components, to be able to drive the next era of computing and the next frontier of computing, so through this really smart and powerful combination of both shrink and stack. That is where the power of the Qnity portfolio becomes really important because we are really unique and the only materials company that does the complete end of front to back-end materials, and advanced packaging is where both of those roadmaps converge. We are bringing front-end materials like CMP materials in to do advanced packaging, while also supporting it through metallization chemistries, dielectrics, IC substrates, and thermal materials. Advanced packaging has been the fastest growing part of our business for the last two years, and I expect that growth momentum to continue as customers continue to invest more and more CapEx to include it in their roadmap architectures, as well as build out the capacity to support that growth. I want to ask on some drama that has been circulating today. I am always good for some drama. On some of these next-gen AI processors, I think the back-end complexity is certainly going up. The extent that back-end complexity does go up, I think is kind of in question. There are some worries about HBM maybe not coming in as high capacity as what people were thinking before. I guess, can you talk about the evolving landscape? Like how fast are these designs moving? What's your visibility into where these architectures are going and- You're talking particularly on the back end? On the back end. Yeah. Look, on the front end of the technology roadmap, there's been a lot of convergence on what the node migrations look like. It's back to the Wild Wild West in terms of advanced packaging and some of the back-end processes, and that's a lot of fun. It creates a lot of opportunity. But sometimes the choice is, how do you know which one of those opportunities are really commercially viable and will end up Because you could invest in an enormous array of opportunities that don't end up being commercialized. The way we think about it when we prioritize it, we make sure that both, as we're putting skin in the game, that our customers are also putting skin in the game as well to make sure that we're investing in the right pathways. What we see, though, are a couple of fundamental trends. Fundamentally, everybody is trying to get as much capacity as possible from advanced packaging. As you do that, there is a bias to get towards larger package sizes, larger package format sizes, whether that's a CoWoS or panel-level packaging or even CoWoS-L versus CoWoS-R. All of that fundamentally is to get larger format sizes. That's great for a materials company because that ends up creating a lot more content opportunities for every unit. I guess the other trend that we're seeing on the back end is the convergence of the roadmap. You're trying to get finer lines and spaces because there are sort of two ways to get, and this is taking it beyond advanced packaging to where does that package intersect at the board level, and that's where we see this combination of high-layer-count circuit boards as well as high-density interconnects. If you're just using more layer counts, then that's a simple scaling effect, and you'll get some benefit from doing that. But if you want to really drive compute to the next level, in addition to doing more layers, you'll also move to a high-density interconnect, which is going to be smaller geometries, tighter lines, and spaces so that you can put more circuit layers on every level. Right? Fundamentally, when you do that, you have to upgrade the materials' content to be able to do that. Ironically enough, it requires techniques that we would've considered to be really, really old technologies on the semi roadmap, are now being reapplied to the next generation circuit board environments. That's another way where our portfolio has been benefited because we're taking our old semiconductor technology and repurposing it for cutting-edge circuit board technology and adapting it to fit the processing techniques that they're using. For example, something that might be in a very old lithography, we convert to a dry film, and it's now cutting-edge circuit board technology. I want to stick on the back-end theme, and I'll kind of poke some fun. At your investor day a year ago, you said that your, I think it was advanced packaging would grow at a high single-digit CAGR. Last quarter, your interconnect solutions business in total grew 28% organically. I think I just left off a zero. As we think about these changing architectures, clearly the back end is becoming a lot more important to these systems. What is the right CAGR that we should be thinking about over the next couple of years? Yeah. Clearly, as much as six or nine months ago, the world has evolved much more quickly than we anticipated. We thought that there were lots of exciting opportunities, and we were very excited about advanced packaging. We didn't anticipate the explosive growth, the pace with which the downstream customers and the OEMs would adopt the technology in everything that they're doing. And obviously, most prominently, that's happened in data centers, but we're seeing it more broadly than just data centers: into premium consumer electronics, into some automotive applications, and people looking to take it into other industrial applications as well. I think that the pace with which the industry scaled up the advanced packaging capacity and adoption of that technology was faster than anything we've ever seen and faster than what we expected. Fortunately, we were prepared with plenty of capacity and supply to be able to support that leap. As that technology continues to evolve, I don't think now we're at the point where the bottleneck isn't how fast can it evolve, it's the bottleneck is how fast can we add incrementally more capacity. That's now what we're working with customers on, is how fast can they add capacity. They're now extending beyond just their own boundaries to work with the OSATs and partner with them. We've had great relationships with the OSATs for decades, so we're thrilled to bring the OSATs into the loop to help them to figure out advanced packaging as well. They've been great customers of ours for a long time. In many cases, that same technology specification that are working at the foundries, whether that's HBM or CoWoS, will extend to the OSATs as well as they scale up capacity there. I don't know that I'm quite ready to put a new CAGR out there. Clearly, my last CAGR didn't go so well. We expect it to continue to be. What I would say is we think there's an opportunity for that business to double over the next few years. Great. The other business that within your interconnect solutions business, which is half your business, so it's a big segment. I want to talk about the thermal segment. This is a piece that some of these investors aren't quite as familiar with the trends in this business, but it is one of the fastest-growing parts of your business. So can you talk about the trends that you're seeing and what kind of growth should we expect out of the thermal side? Yeah. So thermal, the only thing that's really outpacing advanced packaging right now is thermal management and the growth that we're seeing in the thermal management. Thermal management is a great example of a really smart acquisition that we did in 2021. When I started running this business a little more than 10 years ago, the number one capability I felt like we didn't have in our portfolio was thermal management. At the time, I looked for who were the leading companies that had thermal management portfolios, and a British company called Laird Technologies had the leading thermal portfolio in the industry. They were struggling a little bit. They had an antenna business and some other businesses that weren't going very well. We approached them; we tried to buy that thermal business. They wanted somebody to buy the whole company. We waited for a little bit. They got sold to a private equity company. The private equity company broke them up. We ended up getting the thermal business in 2021, and it's been just a terrific acquisition for us with an industry-leading portfolio of thermal solutions at the chip level, at the package level, and at the device level. What we've seen there is that fundamentally as OEMs are trying to get as much compute as possible into smaller format sizes, they're putting all of this compute into tighter spaces. They're trying to manage power density. All of these systems fundamentally have three system-level challenges that are bottlenecks to performance. The first one is signal reliability and signal integrity. How do you get that effective chip-to-chip or chip-to-board communication? The second one is power density and power distribution, and how do you have steady, consistent power flowing? But if you get the signal flow right and you get the power flow right, both of those generate heat. You've got to have something that's going to take that heat away at volumes we've never even seen before. What our team has done is we have rapidly launched probably more than a dozen new products over the last two years at different levels of thermal efficiency so that no matter what design our OEM customers are doing it, and this part of the business is the one that's got the most direct OEM interface. It's really, at this point, to the extent that it's the hyperscalers and the data center operators that are specifying the thermal materials. Because if something goes wrong in the rack or in the server set, they're the ones that are on the line for that performance failure. We're working closely with them on how do you optimize the thermal performance at the chip level, at the package level, and at the ultimate design level. We're doing system-level testing for them to identify the isolated hotspots and then recommending solutions for how to mitigate that all the way through the architecture from the front end. Then we'll work with their manufacturing partner. Sometimes it may be all the way back to the foundry level. More often than not, though, it's at the OSAT or the packaging house, and then certainly as they do the final device assembly. Whether it's a liquid, a phase change material, a gel, a putty, a grease, or a pad, no matter what format they want that thermal material to be in, we've got the ability to customize it and develop it for them. Maybe to round out your interconnect solutions business, then we can get to the semi side of things. On the PCB side, it seems like there's two kind of diverging trends in that business. You've got high-end AI PCBs going to servers, whether it's CPU, GPU. But then you've also got some non-AI PCB business, maybe some smartphones and PCs. Those businesses aren't doing fantastic. Can you summarize how you're thinking about the PCB market and what trends you're seeing there? Yeah. If I started and I kind of combine the last two questions. When we think about it, we talk about kind of our big three leading-edge technology positions in the interconnect space. Thermal management is the largest of the three, then advanced packaging, and then the AI PCBs would be number three. Together, those three comprise about 30% of our interconnect portfolio. In the first half of this year, those three combined were all growing more than 50% year-over-year. So really strong growth from all three. Specifically, what we're seeing on the circuit board side is back to some of the dynamics that I talked about around the high layer count boards and the high-density interconnects, where you're trying to get smaller and smaller geometries so that you can maximize the circuit space on every board. When you do that, fundamentally, that upgrades the material set that you have to use in order to get the electrical performance and the reliability that you need at the smaller geometry size. So we've seen some nice share gain opportunities there on top of the content gains as our customers more broadly adopt those technologies in there. But we've also seen what I would call a little bit of—I call it a "data center halo effect"—in that as customers put in the equipment to do these higher, more capable circuit-level boards, they're not just limiting the sale of those boards to their data center customers. They're making them available to their consumer device customers. They're making them available to their automotive customers. What we're seeing is that several of those customers, knowing that eventually they'd like to see AI move from the cloud and the data center to the physical AI world of vehicles, devices, and machines. In preparation of that, they're adopting some of those architectures even early so they can get a sense of how that's going to fit in within their overall system dynamics. We're seeing them already adopt some of those architectures and some of those higher, more sophisticated boards into their application. So we're getting that effect, the growth, not just in data center, but we're seeing really solid content growth and opportunities in automotive, in premium consumer electronics, in some of the broader industrial applications, even in environments where unit volumes may be a little bit flattish. Let's switch over to the semi-side. I think a good starting point would be as we think about the big segments of semiconductors, so you've got foundry logic, leading edge; you've got mainstream nodes; and then you've got memory. Can you summarize what you're seeing from a fab utilization perspective and where volumes are in each of those big buckets? Yeah. Certainly the largest, most significant growth is coming from advanced logic and HBM, DRAM. We see utilization in advanced logic. Kind of the end of the second quarter, we talked about it kind of being in the mid-80s, kind of probably going into the high 80s, maybe even into the low 90s as we moved into the second half of the year. DRAM and HBM is already there. They're already operating in the high 80s, maybe low 90s. I'm not sure how much more we can squeeze out of there. That's pretty much as sold out as you can get. Although if there's a way to squeeze more out, they'll figure out how to do it. On mainstream logic or legacy logic, both mainstream logic and NAND are kind of both sitting in the low 80s right now. I think we're having very constructive conversations with the mainstream logic side of the house. It's a little bit uneven. Those who maybe have more consumer exposure, they're struggling a bit more, but many of them have figured out how to find really exciting off-growth opportunities, whether that's in data center or some of these other industrial applications. And net, we think that's positive for overall utilization. We expect kind of continued cadence of slow and steady improvement in legacy logic, and NAND continues to inflect upward as well, a little bit more gradual. I know there's a lot of work going on to NAND to say, "Can you utilize anything in NAND technology to relieve some of the bottlenecks in DRAM or HBM?" We're excited by those opportunities. We're working with customers on some of those, but ultimately, we're also trying to help them advance to the next node of scaling within NAND as well, whether that's getting from 300 layer counts up to 400 layer count NAND architectures. Then on the mainstream side, higher memory pricing—are you seeing a slowdown from higher memory pricing from impacting that mainstream business? Yeah. It's something that we're watching across our mainstream customers. I think the ones that have more consumer exposure—I think it'd be naive to. The customers have been pretty transparent in terms of they've made public comments around, it's restraining their growth a little bit. What I've been encouraged about is how nimble and adaptive and agile many of them have been to be resilient, even in a downturn with some of their consumer electronics, to find other ways to find meaningful growth opportunities. You see that in some of the power and sensing modules that they're commercializing within data centers. But many of them are doing quite well in automotive and, really, in the premium consumer electronics. When I look at our business, even in a year where volumes are a little bit restrained in consumer electronics and in automotive, we're seeing high single digits to double-digit growth in both of those segments because of the content uplift that we're seeing, and some of that is coming through mainstream logic; other is coming from kind of the more back-end technologies. On the leading-edge side of things, we're at a very exciting point. We're starting to move into 2-nanometer and sub-2-nanometer nodes coming into really high-volume manufacturing. Can you remind us how does your content per wafer scales as you move from a FinFET node to gate-all-around and maybe as you go from 2 nanometers to the sub-2 nodes? Yes, it's a great question. I think this is the—I get really excited about these technology migrations and the opportunities that they create. I think when we think about going from FinFET to gate-all-around architecture, fundamentally, there's a layer scaling effect that benefits the material intensity. If you go from, say, a 14 nanometer construction to a 2 nanometer construction, you'll roughly double the number of layers that you're doing. There's a nice, particularly, and where that benefits us the most, where we have the most exposure to these content uplifts, is really in the chemical mechanical planarization or polishing part of the semiconductor process. That's a little bit more than 50% of our semiconductor segment. Anything that increases the CMP process intensity is going to have a disproportionate benefit. We're the market leader for both polishing pads as well as cleans. We've got a strong slurry offering as well that's been growing very nicely for us. What I'm really excited about is not only do you see a doubling of the layer count that you're using on the number of layers within those architectures, but several of those layers, especially on the advanced logic side, the process intensity is such that you're doing multiple polishing steps in several of those layers. It's almost a double multiplier effect. If I take back, we don't disclose the specifics by customer because it gets into some proprietary information, but there's some pretty good third-party research out there that talks about the material intensity increasing for each node migration of, say, somewhere between 20%-30%. For us, if you think about a migration that is a little bit more CMP-intensive, you'd be on the upper end of that range. Something that's maybe a little less CMP intensive; it's a little bit on the lower end. As we move forward and we look forward to the Angstrom era nodes, what I'm most excited about is the growth and the adoption of backside power. We're starting to see some of the early adoption of that, and even with some of the 18A technology, but getting to backside power in some of the 16 and 14 architectures because that's a real step change in CMP process intensity. Our teams have been actively working on 16, 14, and even 10 technology for more than a year now, and the number of wins that we've seen, a significant number of wins in both the 16 and the 14 node migrations, particularly as they adopt backside power architectures. What is more important in CMP: pad, slurry, or clean? Look, there's probably others that have different opinions. I think my answer would always be pad. Right? I think that's where the engineering and the architecture of the process of the CMP process really starts with the pad. The pad controls the interface with the wafer. The pad also controls the interface and the mechanics of the flow for the slurry. The efficiency of that process combined ultimately determines how much of the cleaning solution is actually necessary. Because if you had a perfectly efficient polishing process, cleans wouldn't even be necessary. Now, fortunately for us, that's never the case. You always need a cleaning solution. But ultimately, we think that, and for us, the pads business is our largest business within CMP, and we have a sizable cleans and slurry business as well. As we get to more and more advanced Angstrom era nodes, the degrees of flexibility in terms of how the technology profile in the window operates, becomes smaller and smaller and more challenging, so there's more and more opportunities to leverage technology across those domains to help optimize yield and performance for the customer. Because ultimately, it's about yield in the fab, right? All the customers care about is if you can help them with even a 1% point in an advanced node fab; it's worth $200 million. If you can get even a tenth of a point of yield through something that we're doing on the CMP step, it's worth more than the price of anything that we have, because the bill of material on everything in that CMP process step is very small relative to the value of a tenth of a point of yield. What is your capacity situation in CMP? Because it is growing very strongly. Are you capacity constrained anywhere? We've been steadily adding capacity in CMP, really since the peak in 2022. In 2022, our cleans and slurry business, for example, we're only manufacturing in the U.S., and consistent with our local-for-local model, we knew we needed to add capacity primarily in Asia. Since that time, we've added cleans and slurry capacity in Taiwan, Korea, and China. As a result, we've seen that business continue to grow significantly over the last several years. In addition to that, we've added capacity in the U.S. and in Taiwan and in Korea for pads. We announced two large projects, capital expansions in the first quarter, one in Delaware in the U.S., the other in Taiwan. The first line, both of those are oriented, or the first thing in both of those facilities will be expansions to our pad capacity. The first line is up and running in Delaware. We're doing customer qualifications now, and we should have commercial sales maybe by the end of the year, knock on wood, but certainly will be commercial volumes next year. The next phase of investment is already going into that facility. We're doing investments in the Taiwan facility. We'll have the equipment installs done by the end of the year, and we'll be in qualification process early next year, and depending on how that goes with customer qualifications, we'll have volumes out of that facility sometime next year. Great. I want to talk about pricing. You guys are firing on a lot of cylinders. The cycle's turning up pretty strongly. Can you talk about how you guys are approaching pricing from a materials side? Is it you're passing through cost increases? Are you getting opportunistic? How are you guys approaching the pricing side? Yeah, it's a little bit of both. I would say, going back for several years now, whenever we see market dislocations on raw material costs, sometimes that could have been tariff-induced; sometimes we saw some of that during COVID. We certainly have seen it with some of the Middle East disruptions. We pass those costs directly onto the customer. There might be a quarter-to-quarter, even maybe a one-quarter, lag on the timing of that. But we talked about earlier this year, in the first quarter, we talked about just what's going on in the Middle East, having about a $20 million impact, really not on our raw material buy; it's mostly on freight and logistics. We've already taken actions to pass those costs through to customers so that it's neutral from a margin perspective for us for the year. By far, the more impactful way that we think about pricing, though, is on new product launches. Every year we're launching a steady stream of new products. I think there may have been even some announcements in the news flow today on some new products that were launched in our interconnect business. But typically, the way we think about pricing is to price for value on the launch of the new products. Just mechanically, the way that shows up in the numbers is because there's no prior reference point; there's no reference point in the prior year for those new products, it shows up as part of the volume number, right? One of the reasons why the volume outperformance has been so strong for this year is, yes, we're getting content gains, but we're also getting pricing power from the new products that we're launching. Then on the legacy products, there's always a little bit of a technology price fade on legacy products. That's kind of flattish this year. Historically, it's averaged minus 1% a year. That's a little bit less of an issue this year, but where we're seeing significant benefit is on the launch of new products. Interesting. Maybe in the last couple of minutes that we have, I want to talk on the margin line. Can you talk about, with all this revenue growth, how do we balance this against your EBITDA expansion? You've got some nice EBITDA expansion. Some of your revenues do carry lower relative EBITDA margins, if I'm not mistaken. Can you talk about how you're viewing EBITDA margin expansion from here? Yeah. I think we are really happy with kind of an industry-leading margin performance, kind of where we sit today. We are excited with opportunities to continue to see margin expansion and margin growth, I would say fundamentally driven by kind of three core drivers that will contribute to opportunities for margin expansion over the next few years. The first one, and maybe the most obvious one, is just the volume growth that we are seeing and the natural operating leverage that you get from volume growth through our existing facilities. That creates a nice tailwind. The second win is really all the fastest-growing products that we have talked about, whether that is the advanced nodes within the semiconductor realm or the advanced packaging and interconnect and the thermal management on the interconnect side. All of those are the fastest-growing parts of our portfolio, are also the highest-value parts of our portfolio. So you are getting a nice mix effect while we get the volume growth. The mix effect is also adding to that. Then the third component is really; we announced earlier this year a transformation program that was really consistent with the spin. When we spun, the way I would say it is we got pieces of a lot of things, whether we wanted it or not. So there was an opportunity to clean up a little bit through a transformation program so that we were optimized for a semiconductor pure-play company, and we sized that at roughly $100 million of EBITDA benefit over the next couple of years. We are making great progress with that. Just to bring that to life a little bit, about 50% of that benefit is going to come through productivity efforts within our existing facilities. A lot of that may be things like putting in automation. We see lots of opportunities to do more with AI and automation and machine learning in our existing facilities. That will improve quality, it will also improve capacity, and it will be at lower cost. So the combination of more capacity with better quality at lower cost is a very high return initiative, so we have kind of doubled down on that as part of this transformation initiative. The other 50% is really through a combination of commercial and innovation excellence, where the goal is how do we accelerate the product, the innovation engine, so that we can go after more opportunities. We are seeing a significant increase in the number of opportunities to engage with customers and OEMs. We want to increase our ability to respond more effectively without having to add incrementally a lot more cost. How do we make the scientists and engineers that we have as efficient and effective as possible? Some of that may be through molecular modeling, simulation, and machine learning technology applied in the R&D construct. Some of that may be engaging earlier with customers in the design process so that the earlier we engage with them, the more efficient the R&D development can be. Then the last area is really on the local-for-local model and optimizing that. The example that we used is when we spun, we have a very low percentage of our business that is in Europe, but as part of the separation, we got way too many distribution points in Europe than we know what to do with. Part of that is maybe simply consolidating some of that footprint within the distribution network to something that's more fit for purpose, more intentional. Again, customer service will go up, costs will go down, and it'll be just a much more efficient operating model. When you take that and you work that through the entirety of the system, we expect we'll see some of that benefit even in the back half of this year. It's baked into our guidance for 2026 that we talked about at the last earnings call. But the bulk of that will really start to come in in the back half of 2027 as we get those investments up and qualified, and they start to flow through. Is Ken on board with all those projects? Absolutely. He's excited. He's actively led these kind of large-scale transformation initiatives, and I'm pleased that Mike, who's been kind of in the interim seat, is also heading up our transformation office as well. It'll be a natural flow and transition for him to be able to continue leading those to make sure that we don't lose any momentum with that transformation opportunity. So those three combined, the transformation, the mix benefit, and the volume benefit all combine to give us kind of a pathway towards nice margin expansion. At the gross margin level, I would say, puts us into; we're kind of sitting in the high 40s today. It gives us an opportunity to maybe get at the company level, kind of up to or above into the low 50s, 50%, or into the low 50s, and nice margin progression over time on the EBITDA line as well. Great. Last topic I want to touch on is capital allocation. Specifically, I want to talk about, sorry to put you on the spot, the inorganic side of things. You've got a really nice portfolio, a lot of really great assets within it. You've talked about being open to doing maybe some bolt-on or tuck-in M&A. I guess where's your head, where's Ken's head at in inorganic opportunities that you guys are looking at? Yeah, I'm excited to allow Ken the opportunity to talk about this. I know he's passionate on this subject. He and I have had the opportunity to have many conversations about this over the last couple of months. Really, the way we think about M&A, our first capital allocation priority is really around organic reinvestment in CapEx, in capacity, and in R&D to continue to drive the growth and the success of the company. But after that, this portfolio has taken shape over time through a combination of really smart acquisitions. Most of them have been the bolt-on, tuck-in variety versus the transformational. We want to continue to be very open towards building out, and we've got a nice pipeline and a disciplined process around kind of smaller bolt-ins or tuck-ins to continue to expand our portfolio of market-leading technologies. The areas we've talked about of being of most interest are the ones that are also the fastest-growing: advanced packaging, thermal management. We've got a nice kind of semi-consumable business. It's more of an equipment consumable. We think there's lots of other areas within the equipment landscape that are kind of pseudo-consumables that we think would be an interesting addition to the portfolio. Ultimately, those things will help us to kind of fill out the portfolio and continue to build and accelerate the growth trajectory and the momentum that we've already seen over the last six to nine months. Great. Maybe last quick one on buybacks and dividends. How are you thinking about cash returns? We've got an active buyback program in place, a $500 million authorization. It's fairly open-ended. The way that we're thinking about that right now is primarily to offset normal annual dilution. And obviously, we're really happy with the cash flow potential of this business. It generates a lot of cash, kind of normally would be in that mid-teens percent of sales kind of cash flow potential. The balance sheet has plenty of flexibility. We finished the second quarter at kind of two times net leverage, so some balance sheet flexibility of firepower to do deals that we wanted. We have a modest dividend in place. We've got kind of the share buyback program. So shareholder remuneration is out there. We've got flexibility there. But I think our priority and our focus is going to be how do we continue to accelerate growth. Perfect. Well, we're excited to watch that continue to accelerate. Thank you so much, Jon, for joining us, and- Excellent. Thank you, Melissa, and thank you all for being here today. I know it's been a full day for everybody. All right, thanks everybody.