We're ready to go? Yeah, we're ready to go. Okay, good morning, everyone. Thank you so much for coming. I'm Toshiya Hari. I cover the semiconductor space and the capital equipment space at Goldman Sachs. Very pleased, very excited to host Lam Research with us this morning. We have Doug Bettinger- Thanks for having me. EVP and CFO. Before I go into questions, I'll hand it over to Doug. Great speaker. Listen, I hope everybody got all bulled up listening to Jensen this morning, 'cause if he's excited, I'm excited. 'Cause you don't get GPUs for free, you gotta build them using equipment, so anyway, I always like to keep my lawyers happy and flash up the safe harbor. Let me read this real quick, and then we can jump into it. Today's discussion may include forward-looking statements subject to risks and uncertainties, and actual results may differ materially. Additional information concerning factors that could cause actual results to differ materially from those forward-looking statements can be found in the risk factors disclosed in our public filings with the SEC and our 10-K and 10-Q. Okay, we can take that down now, and we can jump into it. But all kidding aside, listen, everybody is excited about AI, and certainly I am as well. At the end of the day, these great big accelerator chips consume a lot of wafer capacity, leading-edge wafer capacity, which drives a lot of demand for equipment. Again, all kidding aside, it is one of the things I think at the end of the day we're all really excited about in this industry. So I just wanted to jump on the back of Jensen for a minute, and tee that up. Okay. So with that, we can jump into what you want to talk about here, great. Thanks, Doug. So kicking off with your WFE market outlook, just starting off high level. Yep. On the earnings call, you reiterated your 2024 outlook, in the mid-$90 billion range. Yep. Just remind us what your assumptions are by device type. I know it's early. I don't expect a 2025 number from you today. I'm not gonna- If you're willing, please do. But maybe I'll give you a low call. Yeah, how are you thinking about 2025 preliminary expectations? Where do you have the best visibility? Where do you have the least visibility into next year? Yeah, first, maybe, let me unpack 2024 a little bit in terms of what we're seeing. You're right, our outlook is mid-nineties. It did tick up a little bit as we went through the year, but that's how we see things today. If I unpack that by end device segment, leading-edge foundry and logic is actually pretty strong this year. I think it goes without saying everybody understands what's going on there. It is a lot of the accelerator chips that I was talking about when I first sat down here. So that's driving a lot of demand for leading-edge foundry and logic. Seems pretty strong this year. I think it's probably as strong or a little bit stronger next year. If you think about kinda twenty-four to twenty-five, I don't think that is gonna go away anytime soon. DRAM pretty strong this year, partially because of the move from DDR4 to DDR5. Obviously, everybody's excited about AI. You gotta be excited about high-bandwidth memory. That is certainly a bright spot in terms of what's driving DRAM spending this year. I can't see that abating next year, right? I think DRAM is gonna continue to be pretty strong. There was some level of DRAM spending in China. That was a little bit first half-weighted this year. And so that's softening a little bit as we go through the second half here, but I think DRAM continues to be pretty strong next year. NAND is pretty weak this year. I don't think anybody is surprised to hear me say that. I've been saying it, coming up on two years now. I have to believe that as we look into next year, Toshiya, it's gonna be somewhat stronger next year for NAND, right? We're moving our way through the normal inventory cycle there. I think you'll see next year largely driven by a lot of conversions, technology conversions, which by the way, when that happens, we do really, really well 'cause we are the constraint tools in a NAND fab for the most part, and so when you go through your conversion spending, our share of wallet is actually quite high, so I'm, you know, excited to finally see NAND spending a little bit more. I'm not ready to give you numbers quite yet, but that's kinda how I think about where we're at today and likely what we're gonna see next year. Okay. And then quick follow-up, Doug, anything on the trailing edge? You talked about leading edge. Yeah, no. What about trailing edge? Yep, that's a great question. Trailing edge is kind of a tale of geographies. If you look geographically, the trailing edge in China is quite strong this year, right? You've got a whole bunch of customers investing strategically with an eye towards building a business a year or two into the future. If you look around, what's going on there, it's analog, industrial, automotive, microcontrollers, some RF stuff, some image sensors. So the mature node stuff in China is actually pretty strong this year. Outside of China, it's relatively soft. I think everybody in the room, if you follow semiconductors, knows we're in the middle of an inventory cycle there, in some of the industrial and automotive stuff, and so there's not a lot of investment occurring until the inventory clears. That's kinda what we see going on in the mature node spending. Okay, great. Thank you. Wanted to transition to some of the big technology inflections, but before we do, again, high level, earlier on in the year, I think you talked about or you guided us to model R&D up nicely or a healthy amount. You were less sure about the revenue. We're spending money this year. You are spending money this year, which is good, right? I mean, that's how you maintain your competitive advantage. Curious, when you sit down with Tim and your broader team, what are some of the inflections that you guys are excited about and- Yep investing in today? No, it's a great question. Yeah, I mean, if you follow Lam, as we came into the year, we guided, "Hey, we're gonna grow spending this year." And I was unsure exactly what revenue was gonna do, but we knew we were gonna grow spending because we saw some of these really quite large technology inflections ahead of us that when we looked at it, and we continue to look at it, we see big opportunities that played to the strength of what we're good at doing, and so we felt it important to make sure we're positioning ourself well to drive revenue growth several years into the future. So that's really what the thought process was. And when you look at the spending at Lam, roughly 70% goes towards R&D. So a lot of the money we spend at this company goes into developing new products, driving innovation, trying to move us forward in some of these areas where we see large opportunities. What are those opportunities? It's all across the spectrum. Everybody knows Gate-All-Around is coming. It's a big opportunity that plays to the strength of our selective etch capability and our ALD capability. It is a true 3D structure in foundry and logic that we're gonna do well with. We saw backside power beginning to show up, right? That plays to the strength of our metallization capability and deposition. It's our SABRE tool. We've been investing in dry photoresist, which is a disruptive capability that we believe is gonna be quite a large business three, four, five years into the future, that we're striving to disrupt some of the track business. There's new metallization schemes beginning to show up in the industry, right? Tungsten is moving to molybdenum or moly. Everybody sees that happening. This may be the biggest metallization change since the industry went to copper. So when we look at all these things. And And by the way, there's a technology change in DRAM that, while still pretty far away, 3D DRAM is being worked on today. So when you wrap a bow around all of these things that I just kinda alluded to, it plays to the strength of things that we're good at doing: depositing films and etching high aspect ratio structures. Architectures are moving in the third dimension. I didn't even advanced packaging, which is another one, which is a 3D type structure. These are things we're good at doing, that require innovation, new capabilities, that we felt like we just needed to make sure we were spending the right amount of money to be well-positioned. Now, I don't want anybody to run away and think, "Hey, you know, Lam, you used to deliver a lot of leverage in the financial model. You're not doing that anymore." We will continue to do that. It's just this was a one-year pause, if you will, in terms of driving the financial model a little bit because we saw the need to spend a little bit more on R&D this year. Got it. Okay, that makes sense. Wanna hit those sort of one by one. Gate-All-Around and Backside Power and leading-edge logic, you talked about selective etch, you talked about ALD. Maybe sort of talk through where your customers are in that evolution, your competitive footing, in Gate-All-Around, vis-a-vis FinFET, the opportunities that going forward. Sure. Let me unpack a couple of these, and if I forget some of them, come back and Yeah ... and ask me about it. First, Gate-All-Around. Yes, it is the first really true 3D structure in foundry and logic, at least that's the way I think about it. And when you look at developing the sheets, you need to deposit films using an ALD-type process, and you need to etch them selectively, right? You're kinda going around the corner of things. That's our selective etch suite of products. When we look at this opportunity, for every 100,000 wafer starts, our capacity the industry puts in place, we see an incremental opportunity for our equipment of $1 billion. Now, we won't win all of that opportunity, but we will do well, because it plays, like I said, to the strength of things we're good at doing. So we're excited about this. It's beginning to show up now and will grow into next year. We talked about the spending on this node across all of the industry this year is delivering to us $1 billion in revenue for the entire node, not just the creation of the Gate-All-Around structure. The whole node this year will be $1 billion in revenue for us. So we're excited about that, obviously, and into the future, this will get bigger. So that's one. Second is backside power. Now, I would tell you, each customer is at a little different cadence with these things, and I'm not gonna unpack which customer is where, but I see a lot of familiar faces in the room. You probably know who's where. But likewise, for backside power, it's about the same amount, right? It is nearly $1 billion in incremental opportunity for every 100,000 wafer starts. And so that is beginning to show up this year, will grow into the future as the whole industry goes to backside advanced packaging is another one of these billion-dollar numbers that we talk advanced packaging, think about CoWoS, think about chiplets, think about the heterogeneous integration that everybody talks about in foundry and logic, and think about high-bandwidth memory. For advanced packaging this year, and we do it through-silicon via process steps, which is the most enabling, in my opinion, steps in advanced packaging solution set. This would be $1 billion in revenue for us this year, growing into next year. So there's a whole bunch of these things that, frankly, is broad in foundry and logic. It's also showing up in memory. Yeah. Really excited about the opportunities we see in front of us. Got it. DRAM and HBM, again, you talked about this a little bit earlier. You know, I realize you're coming off a low base, but your DRAM business has grown, you know, more than two X on a year-over-year basis, the past few quarters. Yeah, investment this year in DRAM is pretty strong. Right. Like I said, it's DDR4. We're going to DDR5. You need DDR5 with some of these new server CPUs that are out there. And then you got the high-bandwidth memory construct, which is a much larger die with some yield penalties as you construct that eight-die stack. And so that's driven a lot of the spending in DRAM this year, Toshiya. Right. Right. And earlier, Doug, you talked about DRAM staying strong. I don't wanna put words in your mouth, but I think you sort of said something along those lines into 2025. Is that mostly HBM, or do you see sort of conventional DRAM spending coming back as well? It's, it's- A combination of both? It's both. It's both. Yeah, it's for sure, it's both. I mean, everybody's excited about HBM, but there's also a product cycle in there. Like I said, DDR4 going to DDR5, you need DRAM. You need low latency DRAM to feed the AI compute engine, so. Got it. Okay. You mentioned 3D DRAM. I believe that's a sort of twenty twenty-nine, twenty thirty sort of- You know, yeah, that's the way I would have everybody thinking about it. Late decade- Mm-hmm ...... I think, is what I would have people thinking about. But the industry's working on it today, right? R&D is being spent today to try to figure out how to create the structures. So, but yeah, it feels like it's a late decade phenomenon. Right. The 4F squared, the vertical transistor, sort of half step or baby step, if you will, before the industry gets to 3D DRAM. Yes. Is that an opportunity for, for etch and dep as well, or should we not focus on it? It is. No, it's an incremental opportunity. It's not all the way to 3D DRAM. It's a couple nodes away still today. But yeah, the industry's working on an interim step, 6F squared, going to 4F squared, that will have some very high aspect ratio structures, conductor etch structures, and is one of the reasons we talked about this new pulsing technology that we're bringing out, or Tim talked about that on our last earnings call. That will be quite enabling for 4F squared. Got it. You mentioned NAND is very weak this year, and we all see that. Which was pretty weak last year, too. Pretty... right. Two consecutive years. Two years now. I think you are beginning to hear my customers talk about utilization ticking up. We saw that a little bit in the last quarter, shows up in our spares and service business. So we're beginning to see some indications that, you know, maybe there's some green shoots, but there really is almost no equipment spending occurring this year. Got it. Okay. Longer term, again, sticking to NAND, some of your customers seem to think or believe that capital intensity is lower, permanently, structurally, cycle to cycle. What are your thoughts specifically on dep and etch? And I think you did mention you guys essentially enable these upgrades, so perhaps you're a little bit more agnostic to greenfield versus upgrades, but how should we think about that? Yeah, I mean, the interesting thing when you look at our business is, quite honestly, for every% bit growth, whether it's enabled through a conversion or new wafer capacity, our revenue opportunity is the same. Because in a conversion environment, the stuff that gets converted is the equipment that deposits that stack. That is all of our equipment. It is the etch down through the structure. That's pretty much all of our equipment. Certainly, that high aspect ratio channel hole is all our equipment, and then the metallization is all our equipment. That's the stuff that needs to get upgraded in a conversion year, which is likely what you're gonna see happening next year. And so, frankly, whether new wafer capacity gets added or it's a conversion year, we're gonna do well. And like I said, in a conversion year, our share of the spending is as high as it could be because of what I just described. Got it. Okay. You talked a little bit about dry resist earlier on in the session. I think it was the 2020 Analyst Day when we at least- Yep, when we first started talking about it. When you started- We've been working on it that long. Of course. Yeah. Yeah. Yeah. So it's been, you know, four and a half years since we heard about it. What's been sort of the general feedback from customers, to the extent you're willing to say? How many customers have already qualified your tool? Are any in HBM? You know, how should we think about the opportunity going forward? Yeah, nobody in HBM yet, but what I would tell you is everybody that uses EUV has our hardware in the lab evaluating what it is capable of doing. And honestly, that is a pretty strong endorsement of the fact that we've got something of value here, right? The floor space in R&D lab of our customers is pretty valuable space, and they won't allocate that space unless they see some compelling reason to do so. So everybody that uses EUV is evaluating the hardware, is the first thing I would say. There have been three tool DTOR decisions where the tool has been selected. Now, a DTOR selection doesn't mean you're in production yet. But we're pretty excited about this. We're pretty encouraged by what we're gonna be able to do. The collaboration with ASML is very close. We're working very closely together with them. They see the value in this. We're collaborating with them in Imec. It perhaps has not been selected quite as quickly as we thought back in 2020, but it's absolutely moving forward in a pretty compelling way. I had my CTO at a lunch at SEMICON, and his statement was: "This is a question of when, not if. Got it. And he knows this technology better than anybody at the company, so excited about this. I'm still describing it as a $1.5 billion cumulative opportunity over a five-year horizon. But if you think about that, by the time you get to year five and six, this should be, again, delivering nearly $1 billion in revenue for the company. Okay. I like talking about billion-dollar stuff. Right. Mm-hmm. Nice, big, round numbers, so this is, this eventually gets there. Got it. So a slight push in timing, if you will, but the cumulative pie or the opportunity set unchanged? Yep. Okay. Absolutely. Great. Maybe talk a little bit advanced packaging, 2.5D, 3D across, you know, high-performance computing and HBM. Lots of exciting things going on there. There is. Where is your current revenue run rate, if you could remind us, and how should we think about that going forward? Yeah, we sell a lot of things advanced packaging, but the real strong positions the company has is in the through-silicon via applications, if you will. That's where you create the space to deposit the conductive material and then depositing the conductive material. I call it kind of tongue in cheek, the drill and fill. We drill the hole, and we deposit the material. We own pretty much all of that across all of the industry. It's a tool we call Syndion. It's our deep silicon etch, and a tool we call SABRE 3D, which is an electroplating tool that deposits the tungsten. You have this across the board in the CoWoS process, in the chiplet stuff, no matter who is doing it, and it is extremely enabling for high-bandwidth memory as well. So when you look at the totality of this business, it's more than $1 billion in revenue for us this year, and that has grown quite a lot this year. Driven by both the chiplet foundry logic solution set, as well as high-bandwidth memory. For us, when we look at high-bandwidth memory this year, what we sell into that is tripling this year, so that's a big part of what is driving growth for us this year. Got it. Shifting gears a little bit, I feel like it's hard to have a conversation with investors on Semicap or Lam Research, specifically without talking about China. It's a topic that comes up in almost every conversation. China's been strong this year. It has been. I think in the most recent quarter, China, as a percentage of total, came down to 39%. What are your near-term and medium-term expectations for the region? I think investors are increasingly worried that a big portion of, not just your business, but the industry's business, has been pulled in. To what extent do you have a gauge on how utilized or not utilized the tools are in China? Yeah, let me describe China a little bit, in case you haven't heard us talk about it. It's a broad set of customers in China, a lot of which, not all, but a lot of which I would describe as new customers, so they're spending with an eye towards growing a business into the future. And when you're a new customer, you're learning how to do things for the first time, if you will. So not efficiently deploying capital necessarily, 'cause they're early in the ramp of some of the technologies they're trying to ramp. This is largely 28 nanometer and above, so it's not the most advanced process nodes, but also, if you haven't done it before, it's not easy to do. But it's a broad set of customers, is the first thing I would describe to you. It's not one or another. There's some customers that have been around for decades, but a lot of newer customers doing things, like I said, in industrial, automotive, analog, microcontrollers, RF, CMOS image sensors. It's the broad set of semiconductor devices that don't necessarily need the most advanced process technology. So that's the way to think about it. China is growing this year in terms of spending on WFE. Last year, China was a little bit of a second half-weighted story in terms of the profile of spending. This year, it's a little bit first half-weighted, so as we go through the year, that percentage, you're right, it was 39% last quarter. It'll reduce as the rest of the world begins to invest a little bit more, and the spending in China being a little bit first half-weighted. Now, what I would describe to you is, I think people get concerned, "Hey, this spending is gonna go away." It is not. I do not believe, and I say that because, you know, when we talk to this set of customers, they all have roadmaps that are many years in duration. So the intention is clearly, I believe, for this set of customers in China to supply a lot of the demand in China with supply in China. I think that's what they're striving to do. Okay. Okay, great. I guess, given everything that we've sort of talked about, I'm curious how you and the broader team think about long-term WFE intensity. I think on a past 10-year look back, it's sort of been in the 10%-11% range. It's been more 13%, 14%, 15% on a three-to-five-year look back. As you think about the industry, you know, hopefully progressing to that $1 trillion mark- Yep ... that we all talk about- Yep ... how do you all think about industry capital intensity? Yeah, you know, listen, what I know for sure is capital intensity per wafer put in place is growing at every subsequent process node. As you go forward, it's getting more expensive to create these structures. What do I mean? Planar went to FinFET, is now going to Gate-All-Around. Backside Power is showing up. It costs more to manufacture these wafers, and I think the good thing is, when I step back and look at the industry, the whole industry is getting paid for this increase in complexity. Capital intensity, I believe, will continue to go up. The metric, I think, people need to look at more than, you know, WFE per revenue dollar is WFE per profit dollar. Actually, Toshi, if you look at that, and I know you probably have, but I'd encourage everybody in the room to go look at that. That's not going up, because profitability in the industry in total has gotten better than it used to be. The industry is getting paid for the innovation and the value that it's delivering, and so if you think about from an affordability standpoint, is this investment amount level affordable? It absolutely is. Okay, great. Maybe talk a little bit about competition. Over many years, you guys have generally gained share across etch and dep. I think the Novellus acquisition obviously was a home run. Timing was perfect. Perfect. Yeah, um- Your CEO's from Novellus. That's right. In fact, a lot of leadership team, Right ... came from Novellus. At this point, it's hard to know who came from where, frankly. Right ... 'cause it's all, it's all Lam at this point. Long time ago. time ago. I'm kind of aging myself, but yeah, so looking ahead, anything that you're excited about from a competitiveness standpoint? And I guess part B is a lot of talk around cryo-etch. You know, one competitor in particular is talking quite a big game. You guys sort of responded, I thought, on the recent earnings call. Yeah, we tried to put out some of the technical merits of, frankly, what we're doing. In cryo-etch, there's only one company in the industry that has anything in production, and it's Lam Research. We've got a thousand chambers of cryo-etch in production today. We own all of it, so that's as much as I can point to and say we're good at doing what we do. Okay. It doesn't mean others don't aspire to kinda try to get some of this business. They're trying, but we keep moving the bar. So anyway, that's as much as I have to say about that one. Okay. All right, great. There's no Fireside Chat with Doug without getting into CSBG. Love CSBG. It's a great business. 40% of total revenue, it's grown really nicely through cycle. Apart from your install base growing pretty much every year- Yeah ... any sort of other drivers, revenue opportunity per chamber, any initiatives at the company- Yep - that could drive CSBG further? First, let me unpack CSBG a little bit for those of you that don't listen to us talk every single day. CSBG at Lam is the Customer Support Business Group. It's composed of four lines of business, if you will: spare parts, service, equipment upgrades, and the Reliant product line, which is the mature equipment that we've been selling for years, actually, at this point. You're right, it's in the most recent quarter, it's about 40% of our revenue. It is a quite profitable business because it doesn't require a ton of research and development spending. All the R&D is deployed upfront when the equipment is first designed, and then this is just the ongoing revenue generation of the tools in the installed base. Some of the things we're trying to do and what I think we've done quite effectively is our strategic intention with this business is to grow revenue dollars faster than just the chamber count in the field. So we give you a chamber count every year. The important thing to understand with our equipment is it really never goes away. We honestly have equipment that's been in the installed base for decades, and the opportunity over time is when you're running the equipment, it needs spares. It can be upgraded to a next-generation capability. You need to service the tool, and then often. Well, we used to anyway buy equipment, refurbish it, and resell it. Really, nobody is letting used equipment come back. And so what we're doing is selling new, older equipment there. and we've been pretty effective in growing dollars faster than just the raw chamber count through a combination of spare parts as well as advanced service offerings, using AI-type capabilities to do more predictive kind of maintenance on the tools, as compared to historically, service in this industry was a lot about show up and do a task. We're bringing to bear a lot of predictive-type things, which frankly is quite value-added for the customers. Within CSBG, your Reliant business has been soft recently. It has. How should we think about that business, going forward? Yeah, it's what you asked me about earlier, which is what's going on with mature node investment. That's substantially where the Reliant product line shows up. So when you think about industrial, automotive, analog, that's kind of... It's CMOS image sensors. It's kind of where those products are sold into, and obviously, outside of China, that business is a little bit soft this year because of the inventory cycle that the industry's working its way through. Got it. Okay, I've got more, but I'm gonna pause here and see if we have any questions from the audience. If you do, please raise your hand. We've got mics running around. It looks like you're doing such a phenomenal job asking the question. Actually, there's a hand. Oh, there's one. There's a- We got a mic over here. Just about you, what you were saying. That's right. I was patting you on the back a little bit. Yeah, here we go. We got a mic coming up. Morning. Please, can I just ask a, I guess, a slightly tangential question? So on China, you're effectively saying there's a lot of new fabs coming online, taking equipment. It's not leading-edge kit, and when I listen to the analog players, their answer is that they will remain very competitive in China, they're not gonna back out of China, and the secret sauce in their existing equipment keeps them ahead of the game. But is that really the case if there is a level playing field in terms of the equipment that's been brought in? And we've seen this, if you just look back historically, with what happened with Huawei and Nokia and Ericsson. Nokia and Ericsson told me that there would never be a real competition from Huawei, and now the kit's as good as theirs. So in analog semis, how long before the risk that that Chinese kit is as good as? That's a good question that maybe I'm not 100% qualified to answer. Listen, I think that our customer base in China is gonna continue to get better as they ramp these technologies and learn how to do what they're doing. It's not easy, though. I would definitely say that, and I'm sure what the mature analog companies are telling you, there's truth in that. I don't think they're sitting still, but certainly, I think the Chinese, over time, are gonna get better and better as they ramp technology and try to learn how to do what they're learning how to do. Question the same as, Oops, I don't think the mic is on. There you go. Okay. Now it is. My question's similar on competition in China, but just in semi cap equipment. Can you just map for us where you see Chinese equipment makers closing the gap? Yeah, I mean, Yeah, no, it's a good question. There's a handful of smaller Chinese equipment companies. They're pretty far behind where the global leaders in the industry are at, certainly where we're at today. But there's a set of customers also in China that we are restricted from selling to, right? The U.S. government has restricted certain customers from certain technology nodes. And so if you're one of those customers in China, what choice do you have but to buy what you can buy and try to make it work as well as you can? And I think largely, when you look at some of the equipment companies in China and the growth you've seen from them, it's because of that. We're pretty far ahead technologically, and we're definitely not sitting still. You know, in my thirty-odd years in the industry, the best technology almost always wins, and we intend that to be more often than not coming from Lam Research. Doug, I'm curious, do you guys use AI at all internally, whether it be, you know, designing your tools? Is AI embedded in your tools, you know? Yeah, we do. Listen, some of that advanced service offering is using AI-type capabilities. Our simulation capability in what we call Semiverse Solutions is enabled by a lot of AI. We use it for process development. Actually, my CTO, my former CTO wrote a pretty cool paper that actually showed up in the magazine Nature, about how we're using AI to do some of the R&D work that we do to develop process recipes. That's a huge area where we spend money in. We're using it in. There's a lot of software in our equipment. We're using it to help with code generation, like I think everybody in the world is. So it's showing up all across the board, Toshiya. Okay, great. Maybe in the last two minutes, you know, I meant to ask about gross margin and capital allocation. I kinda skipped those because I think I know what you're gonna say, so, Same thing I've been saying forever. You know, you're a well-covered company, well-covered stock, but when you talk to people like us, and more importantly, investors, as a collective unit, anything that we missed or underappreciate about you guys and your story? Listen, I think, and thank you for asking about CSBG. Often, I'll sit through, like, I'm gonna sit through, I don't know, a dozen meetings today, and I know we'll get through 80% of the meeting, and no one will have asked me about the installed base business, which honestly, thanks for mentioning. It's 40% of the revenue of the company right now, and a big part of the profit generation machine. I think people underappreciate the quality of that component of the business. I don't know, my perception anyway. Okay, great. Thank you so much for attending, and, Thanks for having me. Congrats on everything you're doing. Awesome. Thank you. Thanks, Doug.