Merck KGaA (MRK) Earnings Call Transcript & Summary

Dear ladies and gentlemen, welcome to the Semiconductor Solutions deep dive session as part of Merck's First Virtual Capital Markets Day. [Operator Instructions] May I now hand over to llja Doering, Director of Investor Relationships, who will lead you through this session. Please go ahead, sir.

Thank you very much, Brian. Dear ladies and gentlemen, my name is llja Doering, Director Investor Relations and Principal Lead for the Performance Materials sector at Merck. Thank you very much for joining our Semiconductor Solutions deep dive. As you have heard from Kai Beckmann, CEO of Performance Materials this morning, Semiconductor Solutions accounted for close to 60% of Performance Materials in the first half of 2020. Later in the presentation, we will refer to Semiconductor Materials and Delivery Systems and Services. For your information, the split between Semiconductor Materials and Delivery Systems and Services within Semiconductor Solutions is something we do not disclose in our reports. However, having followed resume prior to the acquisition, you will know that delivery systems and services should be somewhere in the range of 20% of the overall Semiconductor Solutions revenues. With me today, I have the 2 business leaders of Semiconductor Solutions, Anand Nambiar, Head of Semiconductor Materials; and Jeff White, Head of Delivery Systems and Services. Joining them is also John Langan, Chief Technology Officer of Performance Materials. Now as to the frame colors of the pictures you can see on this slide, they refer to the presentation. On many of the slides, you will see that materials are highlighted in magenta while DS&S is highlighted in yellow. As to the agenda, first Anand will speak to semiconductor industry trends and explain how Semiconductor Solutions is enabling the chipmakers. Jeff will then give you an overview of Delivery Systems and Services, and John will share with you a Chief Technology Officer perspective on how Semiconductor Materials are driving the technology road maps in the semiconductor industry before Anand then concludes again with the executive summary. Dear Anand, over to you.

Thank you, llja. Good afternoon, everyone. My name is Anand Nambiar. I head the Semiconductor Materials business. Thank you for the opportunity to speak here in CMD 2020 in our first virtual format. The key takeaway for -- if you would take away something from this presentation is that the electronics ecosystem and us as consumers have really developed an unending appetite for data. More pictures, more applications, faster Internet, all of which requires more and more chips from our customers. And this, of course, means more demand for our products from market. And we are quite uniquely positioned in this very strong market, which is growing at roughly 4% to 6%, with a large customer base of over 100 with a diversified and comprehensive portfolio and an unparalleled supply chain to serve our customers. In other words, more data means more chips, it means more products for us. What we see here is sort of the data ecosystem. The data created worldwide is projected to grow at roughly 30% annually, and this view has been accelerated post COVID, and I'll talk about that in a second. Today, we're looking at 1 zettabyte of data created, which is an astronomical 10^21 bytes. And this growth is exponential, and it doesn't have a trough because if you think about it, if your kids want to look up their childhood photos in 3 decades, you need to be able to retrieve it. And all aspects of this picture when it was created, transferred, processed, stored somewhere and then pulled up 3 decades later and seen on your phone, all of this touches Merck. If you just think about what has happened, all aspects of this picture have some impact from our products in some form across the data wheel with our semi and display businesses. And if we extrapolate this notion of a photograph across various technologies and applications such as 5G, AI, mobiles, data centers, the need for more and more data is pretty clear, and therefore, more opportunities for us to capitalize and grow. Before I get into the COVID-19 situation, it's important to understand the key drivers of the industry. The semi market is primarily driven by mobile phones, PCs and servers, essentially data centers. And this is over 50% of the semi market value. And as Kai mentioned prior, the semiconductor market indicators and external models have really not been helpful during these uncertain times, especially in the first half. So coming out of a recessionary year in 2019, particularly, we were -- the semi industry was already expected to return to growth, particularly from data center investments by the hyperscalers, such as Google, Amazon, Facebook and Microsoft. And the expected launch from a 5G in 2020 and the consequential mobile call uplift. And with COVID, the dynamics change in the market grew more rapidly than expected amidst all these lockdowns. And of course, no one has the crystal ball, and therefore, we are developing our own intelligence and indicators. And based on that, the external end-use indicators such as online traffic and bandwidth utilization, we can see an acceleration of data consumption post COVID, primarily coming from a work-from-home economy. Driving more PCs and Notebook sales, higher internet traffic, and therefore, more and more servers needed to support this growth. And according to McKinsey, COVID has possibly accelerated the digital transformation by 5 years. And you can see how the data growth curve just shown in the prior slide, is probably being bumped up by 20%. And all of this is just music to our years. Let's take a look at our business and how -- what drives our own business. The semi materials market is -- we play in a EUR 2.2 trillion electronics market, of which EUR 368 billion comes from the semiconductor devices. And of the EUR 368 billion, we play in a EUR 47 billion materials market and a EUR 54 billion equipment market. In the materials space, we participate in all major attractive segments, all of which are growing stably, but we do not play in silicon wafers and bulk gases. In the material space, you can think of our playing field is proportional to the wafers put into manufacturing. And this is about 12 billion square inches of wafer area. And you can think of it like as 1,100 soccer fields where we need to apply our materials. Think of it like painting 1,100 soccer fields with really high-quality and high-purity materials. And now the industry is moving into another dimension, a third dimension, and John will cover this in a bit, you can think of applying multiple layers of paint or our materials on the soccer field. So 1,100 soccer fields, which is going into third dimension, adding more and more layers on top. Now let's look at what does this all mean when it comes to an end device. So let's look at a life of a phone through this ecosystem. So when Apple designs a new iPhone, they work with chipmakers who design and manufacture their chips and their chips include CPUs, memories, communication devices, storage, et cetera. We work with these chipmakers to make this design a reality and help the chip makers with materials and equipment to produce these chips with the desired features and functionalities, basically, faster performance, lower energy consumption, lower cost, et cetera. And finally, an assembly house like Foxconn will assemble all these components and put it into an iPhone. And Jeff will also talk about some of the details behind the equipment business. So where does materials and where does DS&S play? Now semi materials and DS&S serve the same customer but different stakeholders within the customer. And more importantly, they have different buying cycles, interfaces and the nature of business itself is very different, as you can see from the market growth charts. Materials are consumables. The decision-makers are fab engineers who are responsible for manufacturing. Some of our materials are really dangerous, and therefore, the materials are stored and delivered from the subfab, you can think of it like a basement. The key materials growth driver is the number of wafers, as I mentioned before, that's put into production, and that's our playing field, and the number of wafers are proportional to the demand of chips, so more chips, more wafers, more materials. DS&S systems are sold to facility engineers in the sub fab to support the safe handling and delivery of our materials to the fab above. The major growth driver here is the construction of new fabs and the expansion of new fabs. In the semiconductor industry, our customers value agility, innovation prowess, application knowhow and supply chain resilience. And we have a strong track record of growth, both through organic and inorganic activities, helping us create one of the most comprehensive and integrated portfolios in the industry. Each of these acquisitions, and I'll talk about what each of these acquisitions brought to us, each of these acquisitions strengthened our offering to our customers and realizing an overall journey from 2014 to 2019 of carefully crafted acquisitions, giving us a growth of roughly 3.3 pacs over the last few years. Let's take a little deeper look into our portfolio. What you see here is the overall value chain from chip design to end use. Our materials are consumed in wafer processing and packaging segments, the one that you see in magenta and the DS&S projects take on customer projects you see in the yellow on the bottom, which is taken off to set up new fabs. Our first acquisition with AZ, we entered into the patterning planarization, spin on dielectrics, cleans and packaging segments. With Sigma-Aldrich, we acquired a small portfolio to enter into the deposition segment. Ormet brought us into the conductive pace in packaging. Intermolecular brings us a unique capability to prototype and test materials electrically, which we expect to help us get more shots on goals. And with Versum, we further strengthened our deposition, planarization, clean, doping etching portfolios to become a leading material supplier to our customers. In other words, we have many eggs in many baskets. Our customer base includes all major semi manufacturers in all regions. 75% of the worldwide capacity is split between Japan, Korea, Taiwan and China, and the remaining 25% is split between U.S., EU and the rest of the world. A few of the industry bellwethers are listed here, but the capacity alone is not very representative of the business from these customers. And I'll explain why. The value in the industry is driven by leading-edge technology players, driving new innovation in the industry where quality and performance is of top priority. So let's take a look at some our customers from a technological perspective. On the x-axis, from left to right, indicates a technological node advancement with the left being legacy technology and the right being leading edge players. The bars indicate the approximate number of companies playing in each of these technology nodes. Think of it like your first smartphone, maybe a decade ago, used the CPU probably produced using a 65-nanometer or 45 nanometer process, which was then the leading edge technology. Your latest smartphone, if you acquired it in the last few months, will probably have a 7-nanometer chip or a 10-nanometer chip, while lower range smartphones, which have less performance features will still use older nodes such as 45 or 32 nanometers. Now CPU and memory are the most critical semiconductor components using leading edge technology. However, smartphones have number of other chips such as Bluetooth and WiFi and sensors that are -- that could all be produced in legacy technologies like 130 nanometers and 90 nanometers and so on. So leading and legacy technologies have unique drivers and have unique applications -- end applications. Leading edge is all about innovation, higher performance and a very few companies driving it, such as TSMC, Intel and Samsung for logic and Samsung, Micron, SK hynix and KIOXIA for memory. Legacy technology is all about productivity, yields, costs and there are several companies worldwide playing in that field. And you can see on the bottom where some of the concentration of geographies that play in leading edge versus legacy technologies. So these dynamics at play, we are set to outgrow this really highly attractive semi market. We are very uniquely positioned in this market, which is expected to grow at roughly 4% to 6% of by wafer starts, with a very large diversified customer portfolio, inclusive of all top chipmakers with a very diverse and comprehensive portfolio and an unparalleled supply chain to serve our customers. Therefore, more data, more chips, more products from us. Let me quickly turn over to Jeff, my colleague, who's going to talk about Delivery Systems and Services.

Thank you, Anand. I appreciate the introduction. Would like to take just a minute or 2 to introduce you to the Delivery Systems and Services business. As Anand alluded to, it's unique in its customer touch, the footprint, the market access, the semi cycle participation and its focus. Yet it's integral to this group we're calling Semiconductor Solutions. We'll kind of walk through that, starting with this first slide. Delivery Systems and Services was an independent company, we would market it as a safety company. We surround the molecules with safety equipment, safety processes and safety personnel that guarantee the safe delivery, quality and execution of the molecules that we've produced. The molecules and materials that we invent have very specific features that create very specific benefits to our customers. Those benefits and features must be tightly controlled from manufacturing through their use points, so that they remain benefits and don't become simply hazards. So delivery systems strives to have a piece of equipment available for each and every one of the materials that we invent. Over the last 3 decades, going on for 4 decades, we've installed tens of thousands of these units. They're interoperable, they're upgradable, they're reliable. They simply do not fail. We've created a subset of brands and products each focused on one of the nodes that Anand talked about, one of the manufacturing nodes. Each new molecule we invent, low vapor pressure, high viscosity, pyrophoric, corrosive, toxic, fixes needs a system to control and maintain their quality and their safety. Let's go into the details of the business a little bit more on the next slide. We become a trusted global player in the safe and reliable delivery of hazardous materials to this industry. Anand talked about the value drivers, the difference between leading edge and more legacy nodes. He talked about how leading edge requires new factories, a new equipment. That's where Delivery Systems and Services plays most poignantly. We have access and provide Merck access to the capital equipment investment cycle of our customers. We start with large fab projects, partnering with our customers in developing and building an entirely new factory. This gives us early, visible, intimate, timely insight into what's coming, what materials and what volumes and when. Really important insight. Anand talked about through this pandemic, how we've had difficulty getting good data from some of the industry players. We have the unique position and luxury of having some insight into our customers, both in the fab where Anand and his group play and in the sub fab, where the delivery systems groups play. And I'd like to give you just a couple of short anecdotes of why this matters. As we entered this pandemic and the materials business was holding up strongly, but a question among ourselves, is this just an inventory build cycle, we're worried about that? Is that all this is. And our Delivery Systems and Services business confirm that, no, our customers continue to invest in new plant and equipment. Our customers continue to see a future beyond this pandemic, and we can have confidence in our business. Another really great insight. A few years ago, the industry was talking about the start of a new particular fab and the great promise of the materials that was going to consume. What was interesting, though, is the delivery systems business of Merck could show that those systems weren't yet connected. That, that fab was not ready yet. And so we knew that those promises and that hope for the future was about another quarter away, and we could plan accordingly. But it gave us great insight. We create these gas and chemical delivery systems. You can think of them as something about the size of a phone booth, if you're old enough to remember that, about the cost of a high-end luxury car, it's about what they cost, at margins that are appropriate. We have equipment and services and processes that are important to our customers, enable our materials, but we made a business out of it. A business that creates benefit for our materials group, creates benefit for our customers and creates financial benefit for Merck. Finally, we have a MEGASYS in-fab services and field services group. Here is where we maintain our own equipment, kind of like the Maytag repairman. Here is where we look after and guarantee the quality and reliability and safe handling of the materials that we've invented and created. So we have now in Semiconductor Solutions, sort of a surround the molecule approach, sort of a way to participate in every step of the process. As I close on the next slide. Just the key strengths, the things that the delivery systems and services group brings. We bring uptime. We make it so that the very, very large investments of our customers and their factories pay off by always having on the materials that Merck has created available to them. We have a flawless safety record. It's actually benchmark in the industry, really quite good. We're very proud of that. We sell to that. We benchmark ourselves to that. We challenge other groups within our company to that safer performance. We have a global footprint, able to flex our factory production to the needs of these great, big customers that would like to start their factories up as soon as possible. We're able to flex with them. And then the equipment, the knowledge of the interoperability of the materials, the services piece that handles and manages those materials creates an inherent additional credibility for Merck across this ecosystem that we call semiconductor solutions. I appreciate you listening today. Thanks for your time. With that, I'm going to hand this off to John, our Chief Technology Officer, and he's going to talk about some of those really cool materials that we invent, molecules that we invent. I've always been impressed with what his group does. Thanks.

Well, thank you, Jeff, and let me confirm it really is an exciting time to be an electronics materials solution provider to the semiconductor industry. What I'd like to show you is some of the trends driving this, some of the macro trends and what it means for the technology used to support that. On the left-hand side of this slide, you can see some of the societal changes that have come to us in terms of interconnectivity. The introduction and use of artificial intelligence, trends like big data, augmented reality, not only do they mean more data, as Anand talked about originally, but they also mean different kinds of data. And we're interacting with our data differently. We would like our computers to hear us and understand us, to recognize us, we would perhaps like our cars to drive up to us in a parking lot. That means we not only have to continue to compute the way that we have over the last 40 to 50 years of Moore's Law. But we also have to complete differently, compute differently so that we enable some of these advanced applications. And so what that means on the right-hand side is that we continue to see improvements required in the standard computing technology, but also, we need new computing architectures so that we enable some of those new applications, things like quantum computing or neuromorphic computing, where we enable those different applications. In addition, the devices have to change as well to become denser, faster, more cost effective and support those increasing new architectures and changes. And finally, we have to communicate more efficiently between the systems on the chip, between different systems and between different locations. We need to continue to improve and enhance the interconnection of the data. Well, how do we do that? To do that, we need new materials and different materials that are more capable. And what you can see on the next slide is that over the last 40 to 50 years of the industry, we've really gone from a handful of elements that were the basis of our semiconductor devices to doubling or tripling the number of elements that we'll see in advanced chips because those materials will work better in the more demanding application. So as a result, on the right-hand side, you can see that we're constantly being challenged to introduce new materials and yet do it in a way that enables high quality manufacturing, consistent supply, so that we can make these more advanced devices. This is a great opportunity for us because we're bringing many new enabling solutions to the industry. But it's also a challenge because each of the solutions can become more specialized, more demanding and more diverse. So how will we respond to this? How will we take advantage of this opportunity and address the challenge? And you can see that here in the next slide. What this shows, on the left-hand side are the steps required to introduce a new material into the semiconductor industry. You first have to invent the material, then you have to test it in a unit process. And then you actually have to make functioning electrical devices to see if the new material and process is actually delivering the benefit that you intend. And then finally, you have to introduce that new design and process into high-volume manufacturing and support the ramp. Historically, companies like Merck and Versum Materials have really focused on the first 1 or 2 steps of that process. And even Merck and Versum had specialty areas that they focused on or maybe 2 or 3 of the key steps that Anand talked about previously. But as a result of our acquisitions over the past, as a result of Performance Materials revised strategy, you can see on the bottom left that Merck now has the ability to introduce new materials and processes across the entire semiconductor manufacturing range and actually test the functioning electrical devices internally. So that we can accelerate and derisk the introduction of these new materials, which is critical because so many new materials are required for the advanced devices and architectures coming to us from these global macro trends. On the next slide, I'll give you just one example of a challenge and an opportunity for Merck that represents the kind of processes we'll follow and opportunities we'll pursue. As semiconductor devices as the chips, as the transistors have continued to shrink. Some of the metal lines that connect the signals from the outside world to the transistor have shrunk into dimensions that are below 20 nanometers. As a result, the standard metal that we use for interconnecting those signals, which is copper, has run into some real challenges because as copper lines shrink below 20 nanometers, their resistance to passing current begins to increase. And as a result, it starts to degrade the performance of the chip. But you can see on the left-hand side are a number of different metals that are options to perhaps replace or improve the copper. However, not -- even though some of those metals may have better materials properties in terms of small feature resistance or much more stability in aggressive operating conditions, not all of those metals can be turned into products, which the semiconductor industry can use to build those advanced devices. There might be challenges in terms of purity or the ability to deliver the material or cost. What Merck's role is, is to invent molecules and products like you see on the right-hand side, which actually enable those new materials, those new metals in this application to be introduced into the semiconductor manufacturing process in a way that allows our customers to manufacture advanced chips in high yield. And now because we understand the steps that have happened before this process and after, we can make sure that these new solutions that we introduce are much more appropriate for that application. This is just one example, but there are hundreds of steps required to manufacture advanced transistors. And there are material challenges and opportunities in all of those processes. And with that, I'd like to hand it back to Anand for just a quick [indiscernible].

Thank you, John. So in conclusion, you heard from John and Jeff about the technology and the delivery service and systems -- Delivery Systems and Services and materials business. So in conclusion, the global trends are continuing to drive this exponential data growth of roughly 30% or more and more given the COVID acceleration. And this is generating more demand for chips to manage all this data, which means more demand placed on our customers to produce more and more chips. And as we are seeing all the top chipmakers as our customers, we are approaching our customers with an integrated and integrated solution in a comprehensive portfolio, we expect to see a robust growth through very strong customer engagement, through our R&D capabilities and supply chain excellence. Thank you.

[Operator Instructions] First question is from Richard Vosser.

A couple of questions here. Maybe one short-term one, just thinking about the potential economic downturn from COVID beyond this bolus of working from home. Once the furlough period get rolled back. Do you think in the short term, you might see a little bit of a slowdown next year as people -- unemployment goes up. Just thoughts there and would that slow down, maybe the adoption of 5G, et cetera. So just a more short term question. Second question, just thinking about pricing and revenue split. So you talked about the leading edge and the old chips, if you like. So just an idea of your revenue split between old and leading-edge and how that develops over time. And the pricing of your materials needed in each of those manufacturing processes so we can get an idea of the switch over time of your business? And then third question. Just we've seen in many organizations or areas, China developing processes or technologies to sort of or copy areas of production. Is there any risk from China in this area in the next sort of 2, 3, 5 years?

Thank you, Richard. Let me first tackle the economic impact beyond work-from-home economy and beyond COVID. So what we experienced in the first half was essentially that was a downturn of consumer electronics, automotive, including and mobile phones, part of the consumer electronics. And going back to what I stated earlier, the main drivers of the industry are still mobile phones, PCs and servers. So we did experience some drop in mobile phone shipments early in this year, which had an impact, however, the growth in the work-from-home economy driving PCs and Notebooks and the acceleration in bandwidth consumption, deploying more servers has overcompensated for any of the declines in consumer electronics. And as I mentioned, consumer electronics is still less than 10% of the overall semiconductor market. So really has a minor impact in the overall business. In the sort of the longer term, we're sort of in a watch and monitor phase. We do expect the digital revolution to continue and what the new normal will be, will be -- we'll see. However, I don't -- we do believe the digital acceleration is here to stay. And therefore, even if there is a impact to consumer electronics in the future, I expect the market to be robust long-term. The second part of your question when I came to leading edge versus legacy technologies. It's a great question. The leading edge customers, their primary motivation is performance. It's all about quality and performance. So price points are high because these are innovative materials, very difficult to produce, very cutting edge, very high requirements in terms of quality and standards. And of course, therefore, the price points are relatively high. And then from thereon, as volumes grow in high volume, we gain efficiencies through volume. So therefore, our profitability actually grows as our volumes grow and prices adjust when it comes to high volume. The legacy customers, their motivation is all about productivity, yields and production improvement. So how can we squeeze out a little bit of capacity, how can we squeeze out a little bit more yields. And for that, the requirements are quite different. The material sets are different. And obviously, the material sets are legacy. Those customers are more focused on application support, tech support to help them -- for us to tweak, they will gauge us and help them squeeze a little bit more yields out. And the third question with regards to China, of course, needless to say, as a group, we support free trade. The impact of the China-U.S. trade war is still being monitored. We are carefully watching it. As I mentioned earlier, China is a very important part of the global ecosystem for semiconductors with roughly over 15% of worldwide capacity. However, China is still trailing in technological nodes with the rest of the world, at least by 2 nodes. And that is still the case, and we are watching that carefully as things progress.

We will now take our next question from Michael Leuchten from UBS.

Two questions, please. One just thinking about your business model and the margin pull-through. It was made very clear this morning that we shouldn't really assume for Performance Materials overall that there would be a significant mix effect as semi is growing quicker. I'm just trying to square that with your presentation right now, which seems to suggest that you get phenomenal pull-through from the consumables on top of the equipment. But I guess that doesn't mean margin improvement. I'm trying to understand why that is, what's capital intensity versus the investment intensity that holds it back? Any color would be helpful. And then the second question is about assembly. Is that something you'd be considering going into? You talked about processes. You talked about the equipment, you talked about the inputs, but is there another step further down the value chain on assembly? Or is that something we should not be thinking about.

With regards to consumables, yes, materials is a consumable business, and it is directly proportional to the number of wafers that are put into production. And as I mentioned before, the number of wafers is what we track is MSI and million square inches of wafers produced. And currently, there's about 1,200 billion square inches of wafers being produced. As the wafer starts growing, when we talked about the 4.4% to 6% growth is wafer starts, a number of wafers being put into production. As these grow, the material consumption is proportional to that. And as we grow on the third dimension on these wafers, it grows even faster. And that's why we see the additional kicker besides just the market growth, there is a market share growth as well as the layer growth on top of it, which is adding to our buffer or the additional buffer. I'll talk to -- I'll leave it with Jeff to talk a little bit about the CapEx drivers for DS&S. With regards to assembly, we believe that it's more of a commodity business, where it is not part of our core competence. And that's something that companies -- big companies like Foxconn that they take on.

So your question on CapEx drivers. Leading edge production requires new facility, requires new equipment. The more leading edge it is, the more greenfield it is. So that's the investment that drives the delivery systems business is partnering with those new investments. So most of our business is focused in sort of the leading edge. The more legacy modes are about maintenance, or about upgrades, about some of those kinds of things. The margins in the legacy piece can be attractive, but the volumes are smaller. The business is focused on the leading edge portion.

Moving on to our next question. We'll take it from Matthew Weston from Crédit Suisse.

Three questions, please. The first is about manufacturing technology and how your portfolio is exposed. So as the nodes get smaller, manufacturing technologies change, and presumably, that's the case as new materials could be added in the future. So how agnostic is your portfolio to manufacturing technology. I'm thinking of historic processes versus EUV. Can you give us some reassurance there? Secondly, as technology becomes more of a political footfall, if I can call it that. Can you walk us through your geographic supply chain security and how exposed you could be to people drawing geographic lines around tech moving forward? Are you comfortable that you have sufficient security there? And then finally it was very interesting the comments that the delivery systems business gives you visibility on an individual fab level and how customers are doing. But I presume that's a relatively short-term visibility. How much midterm visibility do you have? I mean, you've set out this great long-term picture, you've set out that you have great short-term focus. But how much midterm visibility do you have into '21 and maybe early '22? Or are investors as we get more and more focused on the semis business within PM going to have to get more used to a degree of quarterly and intra-year volatility going forward.

So Anand, I think maybe all 3 of those are me, or at least I'll start, is that all right?

Yes. Go ahead. Go forward.

Okay. So manufacturing technology exposure, particularly as it pertains to equipment. As those nodes go down, certainly, the preponderance of the wafer fab equipment market would be OEMS, so think about some of the big OEM makers around the world, applied materials, Lam Research, some of those sort of folks. They drive that, and that equipment changes substantially. The equipment that we sell changes less so, but the specific components and materials of construction and whatnot change, as the molecules we deliver change. So as long as the molecules are consistent, and in many cases, they are from node to node, then to use your words, the equipment is agnostic to that technology. As the molecules change then the advantage we have is intimacy with our own materials development to make sure that our equipment changes with those molecules and is available when that comes up. As far as the geographic supply chain and equipment, really, what we try to do is align in region, in culture, in language, in time zone, in technology with our customers. And so we have a footprint that aligns with our customers in assembly, in engineering, in sourcing, in all those kinds of things. So we sort of have the benefit of both a global supply chain and the local supply chain, trying to insulate ourselves from that as much as we can. And then your third question, I think, was short-term versus midterm visibility. It depends on the level of granularity that we're looking for, right? Midterm, you can certainly see when the ground breaks on a new factory. And we participate at that point or before that point. Long term, we're intimate with our customers as to what their plans are on both sides of the business, the materials and the equipment. And then mirror in, we're active participants with our customers as they build those factories. So as much as anyone has visibility into what's happening with specific customers, I think we do.