Applied Materials, Inc. (AMAT) Earnings Call Transcript & Summary

First I'll go ahead and get started here. I'm Joe Quatrochi, the semi-cap equipment analyst for Wells Fargo. Excited to have Brice Hill, CFO of Applied Materials here to join us.

Good morning, Joe. We love -- everybody loves this facility and the granola bars.

I'll pass my compliments to the chef.

Maybe to start kind of high level a little bit, I guess when we look at the Applied Materials story and you look at your positioning in the market, what do you think investors maybe misunderstand or underappreciate about the Applied Materials story as you look out in the next 3 to 5 years, what are you most excited about? What do you think is the biggest opportunity for the company?

Yes. Good question. And of course, I'm 1.5 years in the company, so learning a lot about it, too. I think one of the most interesting things that maybe people don't know is how closely semiconductor equipment companies and Applied Materials has to work with its customers to identify the path for R&D. And so rather than it being speculative R&D, it's really you're working hand-in-hand with the customers pointing to specific inventions and developments that need to be made for their road maps that will be put in place. So when you think about the efficiency of the business model that comes through in the cash flow percentage it comes through in the operating cash flow percentage. But when you look at the business and what it's done this year, take ICAPS or packaging, I'll come back to ICAPS as the mature nodes, like IoT, communications, auto or sensors to older technologies. Applied Materials years ago worked with its customers on developing technologies to fit these road maps. And when we look forward and we think about gate-all-around or we think about backside power delivery or CFET or next-generation DRAM. Applied Materials is well positioned to serve a lot of the demand for those inflections. And I think back to the question, it's like, what's interesting to know about Applied Materials. You really have to have great relationships with your customers. They have to be sharing their road map. And you have to have the capabilities and the portfolio to solve their problems. So Applied has a broad portfolio. We have the ability to integrate solutions. So that lets you have a very close relationship with effective R&D.

That's perfect. Maybe kind of double-clicking a little bit on that. Let's kind of talk about the demand outlook. Obviously, the longer-term thesis you just kind of talked about. But as we kind of talk about next year, think about you guys just reported earnings a couple of weeks ago, maybe frame the discussion, help us kind of understand what are you seeing from a WFE perspective, looking at '24. What are maybe the biggest unknowns as you sit today, looking in that visibility?

Sure. So first, just to let you know how we think about it in the company, of course, we're not spending a lot of time really fretting about '24. I'll come back to '24 in a second because I know this room has to fret about '24. For us, what do we think about, we think that semiconductors are foundational to productivity investments globally, whether it's data centers, EVs, green power projects, we can go on and on, on the end markets that are using semiconductors. So we think semiconductors will grow faster than GDP, a multiple. Last 20 year semiconductors have grown at 7.5%, I think, CAGR. So it's a multiple of GDP. Then we think equipment companies will grow faster than semiconductors because it's becoming harder and harder to build leading-edge technologies. We think Applied will grow faster than that because we will gain share for the reasons that we articulated with R&D. And then we think our services business will grow faster than that, low double digits over the coming years. And that's the way we plan our -- when you ask, how do you play in your R&D investments, how do you plan your factory expansions? That's what we think about is that macro. For '24, we guided Q1 and we said Q1 is going to be strong again based on DRAM shipments to China, ICAPS is going to be strong in Q1. What we did for '24 rather than give a WFE number or a specific forecast, we just gave a sense of what the market -- what our feeling for the markets are. NAND has been very low. We think NAND will improve, but still be low, okay? So it will come off the bottom, if you will, a little bit, and we do see utilizations improving. Leading Logic, we expect to improve through the year. It's pretty low right now. Our Q1 guide has strong DRAM, strong ICAPS, pretty weak, Leading Logic. We think that will grow from the year. Anecdotally, our customers are telling us it will speed up through the course of the year. We see utilizations improving, and we know that gate-all-around will start ramping toward the second half of this year. ICAPS, we said won't be as strong as '23. That's our expectation despite a strong Q1 guide. And then DRAM, we think will still be a strong year, not as sure if it will be as strong as last year, but it's been strong for us on the equipment side.

Maybe just kind of double-click on that. It seems like the demand drivers are kind of -- this year, we kind of had a handoff right of leading-edge foundry/logic, maybe slowing, but that being more than made up by ICAPS, maybe we're kind of seeing that maybe change a little bit this year or into next year rather. Do you worry about like an air pocket of demand as like seemingly hand off of that demand profile like changing? Or I guess, how do you think about that as we get -- it seems like second half maybe is a little bit stronger for leading-edge as we get into the middle of this year, do you see that kind of a little bit of an air pocket potential? Or how do you think about that?

I guess we do think that leading-edge will be the stronger grower over the medium term. And part of the reason for that is there's been a significant amount of mature technology investments in China. We think those are operating at not perfectly high yields because a lot of them are new customers and new factories. So we expect the ICAPS demand, the way we think about it is it's fairly stable over the coming years, okay? Meaning it grew 40-plus percent in '22. It grew even faster in '23. We expect it to still be a strong market, but stable over the next couple of years so that capacity matures. And then we expect leading logic to grow, same drivers as historically. As far as an air pocket, we've been monitoring, we monitor utilization across all of the factories globally. ICAPS for the first 3 of our fiscal quarters was very healthy. And then this last quarter, we said it was lower utilization. And so I don't look at that as an air pocket because they're not stopping orders and taking a year off or anything. But there is lower utilization in the market, and I would attribute that back to end market demand. So lower budgets maybe on the industrial side is the market that we called out during our call. So putting those sensors in place, putting the IoT investments in place, that may have slowed down a little bit. But again, that may be a '24 thing. We don't view it as an indication of what's happening over the next 3 to 5 years.

And maybe just kind of staying with ICAPS. It's obviously been a strong several years there. I think the questions I get from investors a lot is kind of trying to parse out what is the strength in domestic China. How do you think about that continuing? It seems like the spending there has a pretty good runway, but then the nondomestic China is maybe the area that you're talking about seeing some cracks or signs of may be slowing. So how do you kind of break that out? I know that China is the largest market for ICAPS. But I guess how do you think about those drivers?

Yes. And we get a lot of questions about China, just to fill in a few blanks, 44% of our revenue in Q4 was China. A lot of that was driven by DRAM, catching up to some DRAM orders that we're allowed to ship under the trade rules now, and that's going to happen in Q1 also. Normal size for us is probably around 30%, which we'll return to after we finish these DRAM shipments. We expect China to keep investing, and we expect the ICAPS wafer starts and ICAPS capabilities in China to keep growing. And the reason is there's national strategy, we think, to become self-sufficient from a chip perspective. They consume -- China consumes about 30% of chip demand globally. And we expect they're trying to put in the capacity to serve 30%. So we don't think that's going to slow down. On the rest of the world, the rest of the world has been serving China. And so we do think that the rest of the world has to reevaluate if you're selling image sensors into China, and they're building their own capacity for image sensors, we'll have to sort that out globally. That could be a little bit of what's happening in the mature markets. But really, I think it's just back to GDP and where the end markets are. There's been some talk of EVs being a little bit slower, although they're great from a chip perspective, and some industrial projects being a little slower, it's probably those things to point to.

So when you think about -- you take a step back and you talk about the industry, the semiconductor industry pushing to that $1 trillion and you might pick your time frame, end of the decade, right? How do you think about ICAPS contributing to that in terms of the growth and just the contribution?

Yes. This whole theme of AI has really put this in relief for us, bold relief. AI relies on data. And our company has been talking about how the IoT investments that are being made in sensors and in creating data will really be leveraged by this AI capability and the AI capabilities that will be put in. So we view those both as tailwinds and they kind of work together. On the AI side, it's going to be complex systems, with graphics, accelerators, high-bandwidth memory, CPU, all put together on complex systems, and using advanced packaging, heterogeneous integration, advanced bonding techniques. That's going to be a tailwind for the business. But going with that, will be ICAPS investments, which will be the sensors and the data collection and that. So we actually think of both of those as tailwinds for the business.

Maybe just kind of going back to China a little bit. Given that they've been running at such a high level in terms of kind of revenue contribution and just spending in general and maybe put the memory piece of it to the side for a second. I mean -- how do you think about -- what's the biggest risk of like a slowdown of demand there? Is it fab readiness? Is it just kind of digestion of what they've spent in terms of focusing more on just trying to ramp those fabs that they've already put in place? Like how do you think about those dynamics?

Yes. First of all, we don't see a cliff, just to be clear. So when we look at the next few years and we talk to these customers, it's many more customers than we used to have. So there's tens of customers. A lot of them are new in the last few years. They're building factories. They have products. They have financial support, et cetera. . If I were to -- so just to be clear, no cliff. But if I were to speculate, then I think consolidation just in a business sense, would be something that would change the dynamic in the market. In other words, if you have -- let's just make it up a scenario, if you have 10 companies investing in 28-nanometer and making image sensors. And they're at various yield stages and various success in the market. If somehow those 10 companies were to consolidate, like what happened in a lot of normal business situations, then you improve yields and you improve the factory environment, and that may dampen demand a little bit. I mean I think other than trade or geopolitical disruptions, I think that's the only scenario in my mind that would slow down the investment path.

Okay. That's helpful. One of the questions that I get a lot is just given the demand that you're seeing in China, obviously, they're trying to develop their own semiconductor ecosystem kind of across the board, right? So I'm curious like how do you think about competition from Chinese equipment manufacturers and maybe just remind us of your differentiation and your competitive positioning?

Yes. So there's always going to be competitors, and there are a few Chinese companies that make semiconductor equipment. From our perspective, the number one competitive reaction for us is to keep advancing the road map from an R&D perspective. We talked about developing 20 new tools for the ICAPS space over the last few years with our ICAPS group. And I view Applied Materials as sort of a system. When we sell a system to the customer, we have a network of suppliers that can offer the spares and these are high-quality spares that can keep the tools running. We have a service capability that can help the customer ramp up quickly to high yields and keep the equipment running at high yields. We have a road map for how to improve the equipment over time. For example, better eco footprint, lower power, lower chemical utilization, better efficiency from a spares perspective, et cetera. So when you buy an Applied tool, first of all, you know that it's used by the leading world manufacturers and whatever node you're building. It's already there. It's demonstrated, it's operating. It can produce high-yielding products. And you've got that sort of ecosystem of products and services around it and supply chain, I think it's difficult to replicate that. So for us, it's advanced and deepen those capabilities as quickly as we can.

That's helpful. Maybe shifting gears a little bit, leading-edge, foundry/logic. Obviously, you talked about that being weak. The improvements that you're seeing in that demand kind of visibility, I guess, looking in the second half, how do you think about -- some of it is technology driven, right, architectural changes in terms of -- you talked about gate-all-around, backside power. But how do you think about just the improvement that your customers are seeing from utilization rates and better visibility into just kind of those transitions?

Yes. So it's all -- you're always reading the tea leaves. We do see our customers forecasting higher utilizations in this quarter that we're in right now. We do know that gate-all-around will start shipping from a technology perspective. We do get anecdotally from our customers that they're expecting a stronger year in '25 from a leading-edge logic perspective. We also know that if you're a leading-edge foundry and you're putting a node in place for the largest customers in the world, there's no negotiation on that timing. If you're landing a new node and the largest companies in the world are designing products that take 3 or 4 years to design on that node. You have to put that node in place. It's not really discretionary. It's not timing. It's meeting of the 2 technologies at the same time. So we know some of these investments will be made. It's just a matter of scale. So I think it's utilization is going up. Anecdotally, we're hearing that from customers. The end markets, I should mention just the AI excitement and the tailwind from AI in terms of HBM, accelerators, new systems that are being designed to invest in that. That is providing a lot of leading-edge excitement, if you will. And most of it is pointed towards accelerating through the back half of this year and into '25.

What about -- how do you think about the opportunity for 3-nanometer versus 5-nanometer? I know we haven't seen a lot of tool reuse for the last couple of nodes, and that question starts to really come up given where utilization rates are, even there may be starting to show signs of improving. So I guess help us understand like what is -- how do you think about that?

Well, we think 3-nanometer will be probably similar size node to 5-nanometer. And maybe 2 will be much larger. But it's interesting. I remember thinking about the reuse topic. I have maybe a different view on reuse than you might expect. And that is reuse is a great indication. If companies are so enthused by the next node that they're going to take their product, their device and redesign it on the next node because it's going to be smaller, it's going to be lower power and it's going to be higher performance -- performance power area then that's just a great indication of Moore's Law is working and the benefits that we see from the new technology are working and that reuse is going to make a bigger node on that next landing spot, if you will. So I personally almost view it as a good thing from the perspective of -- it's indicating the health of the process, if you will. Now in practice, the foundry customers typically don't move that equipment forward. They fill up the prior nodes. I don't know if that's going to change with 5-nanometer or not, we'll have to wait and see. But if it does, then we would view it as a -- I would view it as a good indicator.

Okay. Maybe we'll see kind of how numbers shake out with WFE where it falls as we get into next year. But I guess, it seems like you guys are going to take quite a bit of share this year just given your revenue is down much less than the overall market, much less than your peers. I guess, how do you think about that looking into next year, just given the kind of handoff of demand that we're seeing and just your position and some of the opportunities to maybe gain share or protect share?

Yes. So for investors, we do expect to gain share. And what I hope that it demonstrates is the company's capability to identify 3, 4, 5, 6, 7 years in advance. What are the fastest areas of growth, where do customers need the most help for new techniques, new tools, new solutions, and getting those in place in time to match the customers' road maps. So '23, DRAM ended up being a pretty strong year, and we highlighted that we've gained 10 points of share in DRAM. It's using more of the technologies in terms of high-K metal gate and patterning like logic did years ago. So we're not forecasting to lose share as we go forward. I mean, there are scenarios where that could happen, but we think because of gate-all-around, because of DRAM, because of backside power, because of CFET, because of the packaging growth, we're not anticipating losing share.