AMAT - Applied Materials, Inc.
Q4
Q4 2025
2025-11-13
+
Operator: Welcome to the Applied Materials Fourth Quarter of Fiscal 2025 Earnings Call. [Operator Instructions] I would now like to turn the call over to Mike Sullivan, Corporate Vice President of Investor Relations. Please go ahead.
Michael Sullivan: Good afternoon, everyone, and thank you for joining today's call. With me are Gary Dickerson, our President and CEO; and Brice Hill, our Chief Financial Officer. Before we begin, I'd like to remind you that today's call includes forward-looking statements, which are subject to risks and uncertainties that could cause our actual results to differ. Information concerning these risks and uncertainties is discussed in our most recent Form 10-Q and other filings with the SEC. Today's call also includes non-GAAP financial measures. Reconciliations to GAAP measures can be found in today's earnings press release and in our quarterly earnings materials, which are available on our Investor Relations website at ir.appliedmaterials.com. And with that introduction, I'd now like to turn the call over to Gary Dickerson.
Gary Dickerson: Welcome back, Mike. Applied Materials delivered fiscal fourth quarter results above the midpoint of our guidance to complete another record year. 2025 was our sixth consecutive year of growth. And over this period, we have grown revenue and earnings at annualized rates of approximately 12% and 20%. These results are made possible by our passionate and dedicated employees around the world. Over the past 12 months, we have built new capabilities, strengthened our product portfolio and streamlined our organization to prepare for the opportunities ahead. Applied is in a tremendous position to benefit as AI computing fuels secular growth in semiconductors and wafer fab equipment. As this is our year-end call, I'll begin with a brief review of our performance in 2025, then I'll provide our latest market outlook. And finally, I'll describe how our inflection-focused innovation strategy enables us to extend our leadership in the most valuable and fastest-growing areas of the market as next-generation technologies ramp in volume production in 2026 and beyond. Looking back at fiscal 2025, while it was a growth year for Applied, our growth rate was tempered due to increased trade restrictions and an unfavorable market mix. Over the past 12 months, multiple trade rule changes have reduced the size of our accessible market in China. Overall, China declined to 28% of our total systems and service revenues in fiscal 2025 and to 25% for our fourth quarter. In 2026, we expect wafer fab equipment spending in China to be lower, and we are not anticipating significant changes to market restrictions. In the areas of the market where we can operate, we are competing well and maintaining market share. Outside of China, the fastest-growing areas of the market in 2025 were segments where Applied had low or no share. In leading -edge foundry/logic, investment was more oriented toward advanced lithography. We believe this is a positive leading indicator for process equipment demand in 2026. NAND, where historically Applied has lower market share, is on track to approximately double in 2025, even though it remains a relatively small portion of the wafer fab equipment market. In DRAM, where Applied has strong process technology leadership, overall spending is tracking to be approximately flat for calendar 2025. Nevertheless, we strengthened our leadership position in DRAM, growing revenues from leading edge customers by more than 50% over the past 4 fiscal quarters. As we look ahead to 2026, we expect the spending mix to play more to Applied's strengths with leading-edge foundry/logic, DRAM and advanced packaging being the fastest-growing areas of the market. I recently returned from an extended trip to Asia. My discussions with customers and partners reinforce my view that opportunities for the semiconductor industry and Applied Materials have never been greater. Our customers are engaging with us to ensure we are ready to support significant production ramps in the coming years. AI has reached a tipping point that is accelerating investment in next-generation computing infrastructure and advanced silicon. Today, we are seeing a virtuous cycle of innovation and demand. Advances in performance, energy consumption and cost of AI computing open up new AI applications that in turn, significantly increase demand for AI compute capacity. Recent third-party forecasts predict that the semiconductor industry will grow at a compound annual rate between 10% to 15% over the next 5 years, driving a healthy increase in wafer fab equipment spending. We expect 2026 to be another growth year for Applied with our revenue being weighted toward the second half of the calendar year. AI computing is not only fueling growth, but also reshaping the semiconductor road map and changing the way chips are designed and manufactured. Foundational semiconductor technology plays a critical role in increasing performance and bringing down the cost of AI in the data center and at the edge. Today, major technology inflections are underway in 5 key areas: leading-edge logic, high-performance DRAM, high-bandwidth memory or DRAM stacking, advanced packaging for heterogeneous integration and power electronics. At Applied, our core strategy is inflection-focused innovation. We partner with our customers to see technology inflections early. We focus our research and development on the most critical and valuable challenges on their road maps using deep co-innovation engagement models, and we create highly differentiated solutions by connecting our broad portfolio of capabilities and technologies. The 3 products we recently launched at SEMICON West are great examples of how this strategy works. Our new Xtera epitaxy system enables higher-performance gate-all-around transistors for 2-nanometer and beyond. Xtera creates void-free source, drain structures that provide higher transistor speeds that are especially critical for AI computing. The Xtera system integrates epi, cleaning and etch, resulting in a 40% improvement in uniformity and 50% lower gas usage compared to traditional epi. Kinex is the industry's first integrated die-to-wafer bonder. Hybrid bonding enables significant improvement in performance, power consumption and costs for both complex multi-chip packages and die stacking. Kinex is a 6-step integrated system with onboard metrology that provides higher accuracy bonding, smaller interconnect pitches and higher yields for new logic and memory packaging architectures. PROVision 10 is designed to improve yield in 3D devices and further extends our leadership in eBeam metrology. eBeam metrology is critical for 3D devices as it can see through multiple layers of 3D chips and provide multilayer images to identify defects in buried structures. This system is the first to use cold field emission technology for metrology, which increases image resolution by 50% and imaging speed 10x compared to conventional thermal field emission technology. Overall, Applied is very well positioned at the most valuable technology inflections and in areas of the market that will grow fastest as AI is deployed on a large scale. The process tool of record positions that we have established over the past several years give us confidence that we will extend our strong leadership position in logic, DRAM and packaging as advanced technology nodes ramp in volume production. Another key theme we consistently hear from our customers and our customers' customers is that co-optimization of the technology stack is more critical than ever. We are expanding our deep multiyear co-innovation engagements that focus on system technology co-optimization. Our high-velocity co-innovation model provides chip makers and chip designers much earlier access to next-generation process technology to accelerate their new chip and system architectures. This is a core value proposition of Applied Materials Equipment and Process Innovation and Commercialization platform or EPIC. Construction of the platform's flagship facility, the EPIC Center in Silicon Valley is on track, and we are excited to begin operations next year. Our co-optimization strategies extend well beyond R&D. As our customers race to bring these complex new device architecture inflections to market, we are providing advanced service solutions that help them rapidly transfer new technology into their pilot lines and then rapidly optimize device performance, yield and cost in volume production. In 2025, our core service business delivered another year of double-digit growth with more than 2/3s of our service revenue generated from subscriptions. In AGS and across Applied, we are rapidly adopting AI and digital tools. This enables us to drive higher velocity and productivity, innovate the way we work and streamline our organization to meet the opportunities ahead. Before I hand over to Brice, I'll quickly summarize. Fiscal 2025 was our sixth consecutive year of growth even as trade restrictions and an unfavorable market mix trimmed our growth rate for the year. As we look ahead, large-scale AI adoption will drive substantial investment in AI computing infrastructure, including advanced semiconductors and wafer fab equipment. Applied's inflection-focused innovation strategy positions us for another record year in 2026 as we gain share at the highest value technology inflections in the fastest-growing areas of the market. As next-generation technologies ramp in volume production over the coming years, we will extend our leadership in logic, DRAM and packaging. Brice, over to you.
Brice Hill: Thank you, Gary, and thanks, everyone, for joining today's call. I am pleased that Applied delivered record annual revenue, gross margin dollars, operating profit and earnings per share in fiscal 2025. Looking ahead into 2026, I believe we are in great position to benefit from favorable market trends. Based on growing demand for AI data center capacity, we forecast that leading-edge foundry/logic, DRAM and high-bandwidth memory will be the fastest-growing areas of the semiconductor equipment market. We have strong leadership positions in these segments today, and we have targeted our R&D investments to create new products and technologies that will enable even faster and more energy-efficient transistors, chips and systems and drive growth for Applied. In addition, we have been working closely with our customers to better understand their longer-term demand expectations and align our supply chain and manufacturing slots to meet their needs for advanced capacity. Based on our conversations with our customers and other industry players, we are preparing our operations and service organizations to be ready to support higher demand beginning in the second half of calendar 2026. Next, I'll briefly summarize our fiscal 2025 results versus fiscal 2024. Revenue grew 4% to $28.4 billion with growth across all of our segments. Semiconductor Systems revenue was up 4%, growing even as the impact of trade restrictions significantly reduced our access to the market in China. The impact of these restrictions was equivalent to around 10% of the China market in fiscal 2024 and more than double that amount in fiscal 2025. On a global basis, we generated record foundry systems revenue, along with record DRAM sales outside China, and we posted record revenue in both Taiwan and Korea. Applied Global Services revenue grew 3% to a record $6.4 billion. The recurring parts, services and software portion of AGS grew by double digits in the year, while the 200-millimeter equipment business declined. We grew display revenue by 20%. I am pleased that we increased non-GAAP gross margin by 120 basis points to 48.8%, the highest level in 25 years. We shipped a richer mix of advanced systems and increased prices broadly, helping to more than offset cost increases. Non-GAAP operating expenses grew 5%, primarily driven by a 10% increase in R&D investments. At the end of the year, we announced actions to reduce headcount and enable us to scale Applied more productively as we capture the growth opportunities we see in 2026 and beyond. We continue to shift spending to strategic areas, adding people in fields like advanced analytics that are critical to the speed and efficiency of our R&D programs and operations. Non-GAAP earnings per share increased 9%. We generated nearly $8 billion in cash from operations. Free cash flow of $5.7 billion included elevated capital spending of $2.3 billion, over half of which was used in building the new EPIC Center in Silicon Valley, which will open next year and become the most advanced collaborative semiconductor equipment and process innovation facility in the world. We distributed approximately $6.3 billion to shareholders. We paid $1.4 billion in cash dividends and the quarterly dividend per share was increased by 15% during the year to $0.46. Operating income from Applied Global Services more than covered the dividend payment. We allocated $4.9 billion to our share repurchase program and reduced shares outstanding by more than 3%. Turning to fiscal Q4. We delivered revenue and non-GAAP EPS above the midpoint of guidance. China revenue declined to 29% of total company revenue, which is in line with our longer-term average and well below a peak of 45% in the first quarter of fiscal 2024. Non-GAAP gross margin was at the midpoint of guidance and up 60 basis points year-on-year, while non-GAAP operating expenses were slightly higher than our expectation and up 3% year-on-year. Turning to the segments. Semiconductor Systems and AGS revenue exceeded our expectations for the quarter, while non-GAAP operating margin for both segments declined along with revenue on a year-on-year basis. Lastly, Display revenue exceeded our expectation for the quarter and was up 68% year-over-year. Next, I'll share several reporting changes we are making that will help us drive further efficiency gains and also give investors more visibility into our semiconductor and services businesses. First, as of Q4 fiscal 2025, our display business is being reported in corporate and other. There is no change to our display strategy. Next, as of Q1 fiscal 2026, we are moving our 200-millimeter equipment business from Applied Global Services to Semiconductor Systems. This change will increase our operational efficiency and enable investors to see all of our semiconductor systems revenue in one place. Also, as a result, Applied Global Services will consist entirely of recurring revenue. This will make it easier for investors to track our subscription-like growth in services. Finally, as of Q1 of fiscal 2026, we are fully allocating corporate support costs to our businesses. This change will have the effect of reducing Semiconductor Systems and AGS operating margins, but also give our teams better visibility and opportunity to optimize these costs. Now I'll share our guidance for Q1, which includes the reporting changes I just outlined. We expect company revenue of $6.85 billion, plus or minus $500 million and non-GAAP EPS of $2.18, plus or minus $0.20. Within this outlook, we expect Semiconductor Systems revenue of around $5.025 billion. AGS should generate revenue of around $1.52 billion. Corporate and other revenue should be around $305 million composed primarily of display revenue. We currently expect non-GAAP gross margin to be approximately 48.4% in Q1 and remain at that level until volumes ramp to support higher demand beginning in the second half of the calendar year. Non-GAAP operating expenses should be around $1.33 billion, which is up only slightly from fiscal Q4 because the actions we recently took are mostly offsetting the increase we normally see in Q1 due to the timing of annual merit increases and equity compensation expenses. Finally, we are modeling a tax rate of around 13%. In summary, our customers are indicating to us that wafer fab equipment spending is likely to accelerate beginning in the second half of calendar 2026. In addition, we see a positive fab equipment spending mix developing for Applied. AI data center investments translate to strong demand for our most enabling products in leading-edge foundry/logic, DRAM and high-bandwidth memory, along with advanced services that help our customers accelerate ramps and yields. Thank you for listening. And now Mike, let's begin the Q&A.
Michael Sullivan: Thanks, Brice. [Operator Instructions] Operator, let's please begin.
Operator: Certainly. And our first question comes from the line of C.J. Muse from Cantor Fitzgerald.
Christopher Muse: I guess, Gary, the world has clearly changed since NVIDIA reported August 26 and discussed $3 trillion to $4 trillion in AI infrastructure spending. Curious, given your trip to visit with clients over the very near term, how your conversations have evolved in the last few months? How has your visibility changed? And how are you preparing your supply chain for this likely tremendous growth?
Gary Dickerson: C.J., thanks for the question. Yes, I have been spending a lot of time with customers and just came back from a long trip to Asia. And I would say, I was meeting with the R&D leaders and CEOs of our largest customers. And just like you said, AI is the biggest focus for all of our customers. It's driving the WFE mix to segments driven by AI, including leading-edge foundry/logic and DRAM, where Applied has strong #1 positions. And Applied is in deep, high-velocity co-innovation relationships with all of these different customers. We have very high share. If you look at the transistor for gate-all-around or backside power in leading-edge foundry/logic, we have very strong visibility and co-innovation relationships with customers over 4 technology nodes, a decade out in the future. So very high visibility in terms of our positions in leading-edge foundry/logic and in DRAM and high confidence we're going to outperform as those advanced chips ramp going forward in the future. The other thing that I heard and we've been seeing over the last couple of months is a major improvement in customer demand visibility. So customers, just like you said, they're planning large ramps of advanced factories, and they want to make sure our supply chain operations and service teams are ready to deliver. So we're getting more than 1-year visibility, in some cases, 2 years visibility with a number of these different customers because as we talked about in the prepared remarks, in the second half of '26, we will see significant ramps for these advanced factories. And again, customers want to make sure that especially our supply chains are ready to deliver. And of course, the improved visibility is critical to our ability to ensure on-time delivery to customer needs. But I'd say that's, C.J., the biggest thing that's changed in the last couple of months.
Christopher Muse: Very helpful. And then maybe, Brice, you announced the headcount reduction over the last quarter. I'm curious how we should think about that and the implications to gross margins and OpEx into first half calendar '26 perhaps versus second half given the ramp that you've talked about?
Brice Hill: So on the reduction, if you look at our Q1 spend in our guide, you'll see that we don't have the uplift that we typically have in Q1 for our annual pay raises and the share-based compensation increases that typically happen. So you'll be able to quantify the rough change in that quarter from that perspective. And just to highlight, that was a year's long program that we worked on to increase velocity and productivity across the company. And for the balance of 2026, we'll add back some skills that we need to fill in for the company. So it was really a wholesale evaluation of all the staffing model we have across the entire company.
Gary Dickerson: Yes. C.J., maybe I'll add one -- another aspect to this. In strategic planning, we had a big focus on innovating the way we work, including AI and digital technologies and driving, like Brice said, higher velocity and productivity. That's really at the foundation of how companies compete. So huge focus, especially on velocity across all of Applied Materials and then streamlining our organization to optimize future performance. And as we went through these changes, we also wanted to make sure, as I talked about, we see significant demand coming from our customers. So we wanted to make sure as we're doing this overall company and workforce optimization to improve the performance of Applied, we're also ready to meet those major customer ramps in the second half of '26 and then going forward.
Operator: And our next question comes from the line of Krish Sankar from TD Cowen.
Sreekrishnan Sankarnarayanan: Gary, my first question is, as you mentioned, you're clearly gaining share in new technologies like gate-all-around and backside power delivery. But when I look at your leadership products, it feels like PVD is moving to ALD, on the CVD, CMP and etch side, besides the usual U.S. and Japanese competitors, there's increasing domestic China competition. So I'm kind of curious how to think about the momentum in those leadership products over the next 2, 3 years as you see increasing competition, both globally and from China? And then I have a follow-up for Brice.
Gary Dickerson: Okay, Krish. There's a lot in that first question, but thank you for the question. So we have very strong positions, gate-all-around backside power. Besides being #1 in process equipment for leading logic and foundry, we're also #1 in DRAM and advanced packaging, especially for high-bandwidth memory. And those are the most important segments for AI energy-efficient computing. And leading-edge foundry/logic and DRAM will be the fastest-growing segments in '26 and for the next several years. And Krish, we're performing well across all our products. Again, I am extremely confident that we will grow share, especially in these segments that are enabling AI energy-efficient computing. Again, we have deep relationships with all of these different customers across multiple technology nodes. We see strong demand for our products, increasing demand for integrated systems for transistors, wiring and now Kinex for packaging. So all of that, I'm very optimistic about. And I'll come back to your PVD comment in just a minute here. The biggest change we've seen in our competitive position in the near term is trade restrictions. We used to serve the entire global WFE market before restrictions were implemented. By '24, 2024, we were restricted from serving around 10% of China's WFE market, mainly in leading-edge logic and the domestic NAND market. And then in the last month of 2024 and the first month of 2025, the restrictions significantly increased for us, and we could no longer serve China's DRAM market and some of the ICAPS market. So our impact grew to well over 20% of the China WFE market, which I think everybody knows that the WFE market in China has been elevated this year and also over the past few years with China approaching 40% of total WFE. So this change in restrictions was a very big impact, especially in 2025. And non-U.S. equipment companies don't have the same restrictions. And so restricted customers can buy from those companies even if they would rather buy from Applied. And we've put a lot of time and studied this topic very carefully. And what we can see is that where we can compete, we are doing very well. And then if I look at China going forward, our business has returned to normalized levels. We talked about kind of mid-20s on our semi business, our systems and AGS. And looking ahead, we don't anticipate significant new restrictions. In fact, we believe that our share in China ICAPS, where we can compete was flat from 2024 to 2025. And that's across, Krish, all of our different leadership products. And we think we can continue to hold and have an opportunity to even gain share going forward. And I say this because we have several major ICAPS new products coming for China and non-China that add to the strong products we have today enabling us to enter new ICAPS markets and also better compete in cost-sensitive areas of the market. So we will increase our ICAPS addressable market opportunity in China and worldwide. And again, where we can compete, I'm actually very optimistic with our position. And then back to PVD, your question on PVD. So I'll tell you, Krish, PVD is doing great where we can compete. We grew PVD in '25. We have a positive outlook in '26 and into the future. And I talked earlier about AI, PVD is enabling low-resistance wiring, and that's critical for faster and more energy-efficient chips. And with advanced chips, you have hundreds of miles of wiring in a single chip. And if you could imagine, that long wire keeps getting thinner and thinner and moving data through that wire at high speed with low power is incredibly hard, almost like magic, and Applied is the clear leader in wiring innovation. So Krish, we see strong demand for PVD, strong growth in 2026 and into the future. We're also innovating. I talked about that pipeline of innovative products. We have new PVD innovations, specifically for ICAPS beyond all of the other things that we're doing that will help us in performance in that particular segment. And then the last thing, on the leading edge, we are -- we do have an increase in demand for the integrated products, including selective ALD and selective moly, which we just had some big wins in the most critical applications. So long answer, Krish.
Sreekrishnan Sankarnarayanan: Gary, no, it's extremely helpful and very informative. Just a quick follow-up for Brice. Brice, how do you think about the impact of this 200-millimeter movement from AGS to semi? Is there a way to quantify it for Q1? Or how much was it in FY '25?
Brice Hill: Yes. It's approximately $125 million, Krish, for Q1 and approximately the same number in Q4. So we're moving that from our services business to our equipment business, and we think that will be easier for us to manage and easier for investors to understand those 2 different reportable segments.
Operator: And our next question comes from the line of Vivek Arya from Bank of America Securities.
Vivek Arya: You mentioned you expect fiscal '26 mix to work in your favor. You also mentioned the significant growth in the back half. I'm curious, do you think there is a potential for WFE to grow high single digit, close to double digit next year and for Applied to outperform that because you do seem to have better visibility and you also expect the mix to work in your favor. So I'm just hoping that if you could give us some sense of what the market could be doing? And how is Applied positioned with respect to that market?
Brice Hill: Vivek, it's Brice. Thanks for your question. We do expect strong growth in '26, led by leading edge and DRAM. The headwind that we have, we do still expect some digestion in China and in ICAPS. So it's kind of a similar situation relative to 2025. But we think leading edge will be very strong. DRAM will be very strong. These are pulled by the headline of the AI solutions in those spaces. And then a little bit of digestion on the ICAPS side, but we do think it will be a growth year.
Vivek Arya: On the clarification side, Brice, how much of the $600 million BIS headwind is in Q1 and the rest of the year? And if I could squeeze in the real question for Gary. One of your peers suggested an $8 billion or so, I think, WFE for every $100 billion or so in data center build-out. Do you agree with that number? Do you have a different number? And how do you see the mix within that $8 billion playing to Applied's strength?
Brice Hill: Okay, Vivek. So starting with the affiliate rule that came out and affected our Q4 and then was suspended. So we had shared that $110 million would be affected in our Q4. We didn't ship that in Q4, but we will ship that in Q1. So the first answer is that's included in our guide for Q1. And then the $600 million is still a good estimate for what's in the rest of 2026. And we didn't share any linearity for that. And in fact, it's still being closed in terms of delivery dates, et cetera. So I would think of that as through the rest of the year. And then on the WFE per gigawatt, we spent some time thinking about this. We think the best way to think about this is about 15% of leading-edge wafer starts and DRAM wafer starts are allocated towards AI data center solutions. So if you think about capacity planning, that's growing at a mid-30 CAGR across the industry. So today, 15% of those 2 end markets growing at mid-30s. And so for companies that are planning capacity for 2 and 3 years out, they make that projection and that much WFE should be allocated towards data center. So we can talk more about that in a follow-up, but I think that's a good way to think about how much is being allocated.
Gary Dickerson: Vivek, you asked about what does that mean the AI demand mean for mix and for Applied. So I talked about this earlier that the 2 fastest-growing segments for '26 and going forward are leading-edge foundry/logic and DRAM, including high-bandwidth memory. And in both of those cases, again, we're clear #1. We're gaining share, high visibility in gate-all-around, and that will become clear as those advanced factories ramp in the latter half of '26. Again, we have very strong visibility, very strong positions. And the same thing is true with DRAM. And I think longer term, if you look at gate-all-around backside power, we're positioned to capture more than 50% of our served market. So I have high confidence that we are gaining share in gate-all-around. I mean that's all of the discussions I had with these R&D leaders when I was in Asia in the last month. And then DRAM, high-bandwidth memory, we're in really great position there. We've gained a significant amount of share over the last several years. As FinFET ramps, we have strong leadership in FinFET. That's going to be adopted in DRAM, 4F-square. We also are in a great position. And then in HBM, future generations will adopt hybrid bonding, where we also have clear leadership. We announced Kinex at SEMICON West. So the mix is definitely working in our favor, and we're well positioned going forward.
Operator: And our next question comes from the line of Stacy Rasgon from Bernstein Research.
Stacy Rasgon: For the first one, Brice, this second half lift in demand. So what does that imply for the trajectory in the first half? Like are you thinking revenue stays kind of in this ballpark until we hit the second half? Or does it grow a little bit and then it grows a lot? And maybe even if you wanted to quantify like second half versus first half split or something like that. But like how do we think about the first half relative to the second half given the lift in the second half?
Brice Hill: Yes. For the semi business, Stacy, and thanks for the question. For the semi business, we think it will be flattish until we see that growth. We will see a little growth in the AGS business. We expect that, as we've shared, to be growing at low double digits, and that's fairly continuous growth through the course of the year. But for the semi business in the short term, it will be flattish.
Stacy Rasgon: Got it. Until like Q3, I guess, you're thinking or?
Brice Hill: Yes. For us, it will really be our Q4 and our Q1 of next year. So that's when the calendar -- that's when it matches the calendar. So we'll see a -- we expect a significant uplift there from leading edge, especially. And until then, it will be slower growth.
Stacy Rasgon: And I guess for my follow-up, you sort of mentioned gross margins kind of staying in this range until we see that lift. So do you expect like a material lift in gross margins when that material lift in revenue comes through? Or like how do we expect like the gross margin trajectory on the back of those revenues as that comes through second half of next year?
Brice Hill: Yes. The 48.4% guide for Q1 is good for this level of business. And I'll just highlight in Q1, we're also expecting a normal share from a China shipment perspective. So it should be -- for the whole company, it should be in the order of 29% and less than that for AGS and our semi business. So that level without a significant uplift in the business, that's good for this size business. And then when we do get to the second half, added volume will help with cost improvements. And then over time, we continue to work on our pricing and cost programs. So you can see with the 120 basis point uplift that we had in 2025, most of that was driven by price improvement. And so we'll continue with that program, but it takes time to affect the margins.
Gary Dickerson: Stacy, this is Gary. One thing I'd say in terms of margins, I do think that we will be able to drive sustainable improvements in margins over time. I mean one thing in the industry, you see a tremendous amount of profits being generated with the AI demand throughout the entire ecosystem. And so our customer profitability has improved significantly. And as I mentioned earlier, also, we're enabling these incredible innovations. I mentioned the hundreds of miles of wiring in an advanced chip. I mean, again, for AI, this is incredibly valuable. So as we continue to work with our customers, enabling those innovations that are critical for AI, bringing new products to market, I believe that we can sustainably drive our margins higher going forward.
Operator: And our next question comes from the line of Timothy Arcuri from UBS.
Timothy Arcuri: Brice, I'm trying to understand China a little better. So all these affiliates were banned and then they got reinstated. And the way that China usually operates is that they -- when they were rushing to get in front of the ban and now -- I would think that they're going to come back and they're going to just try to get as much equipment as they can as fast as they can. So is the add-back not going to be more than 600? And in fiscal Q1, like are you taking slots away from the other customers to give to them? I guess I'm trying to figure out how we don't add more than 600 back the way that these Chinese customers typically operate.
Brice Hill: Thanks for the question, Tim. So yes, for the first quarter, 110 is in the quarter. Those tools are built and will ship. For the balance, we weren't building those tools. We didn't know we would be able to ship them. And so it's going to take time to do the supply chain and actually make the builds and we'll have to finalize the date with the customers. So they may have interest in working that process as quickly as possible. But I would just say the good way to think about this is cycle time, and we'll spread it through the year.
Timothy Arcuri: And then I guess, Gary, I wanted to go back to PVD. So this is like 1/3 of your systems revenue per Gartner. So it's a super important piece of your business. And it does seem a little bit like ALD is eating into this franchise a little bit. I know it's super important. It's been growing. But how do you sort of protect that franchise from -- you have one franchise you're very strong in and one where your share is not as high. And do you think that we're all kind of maybe too concerned about the degree to which ALD is eating into PVD?
Gary Dickerson: Tim, thanks for the question. I absolutely think you're too concerned. Again, as I mentioned, I just spent a lot of time in Asia with our R&D leaders for our top customers. And we have deep visibility into their technology road maps for 4 generations. And I can tell you, again, we have high confidence. PVD is going to continue to ramp. I mentioned the hundreds of miles of wiring in an advanced chip. And we have innovations that are really amazing innovations, and we are the leader in wiring. So if you think about an AI server, again, how you move the data at a very high speed with low power is enormously important. That's within a chip, chip to chip. There's so much innovation that's happening there. But I don't know where this comes from on the ALD eating PVD share, that type of thing. All I can say is that we have extremely high visibility. We have unique technologies, including PVD, and we have unique ability to combine technologies and integrated platforms. One of those integrated platforms, we combine selective ALD together with PVD and 5 other technologies. That's part of what's enabling those hundreds of miles of wiring. And Applied is really unique in being able to deliver those kinds of innovations. And that is enabling 50% improvement in resistivity. It's a $1 billion business for us every year. And so that is going to continue. And again, deep visibility into every -- the N+1, N+2, N+3, N+4 technology nodes. Another innovation that we just won with leading foundry/logic companies is integrating selective moly with PVD into an integrated platform for the most critical applications for our customers. So again, Tim, very high visibility. I agree with you. It's a great business. It's going to keep growing in '26 and keep growing as we go forward.
Operator: And our next question comes from the line of Atif Malik from Citi.
Atif Malik: Welcome back, Mike. Brice, I'm just trying to understand the China commentary you're expecting China to kind of stay at this mid-20s level for the silicon products into next year. So if the assumption is that the domestic China spending is coming down next year, why wouldn't this number be lower than mid-25% and if the rest of the China is growing? I'm just trying to understand the moving pieces.
Brice Hill: So for Q1, that just happens to be what's in the mix. But we do expect it -- if we do have the digestion that we're expecting in China, then we will see it lower during the course of the year. And that will be more than offset -- to get a growth here, that will be more than offset by the strength in leading edge and DRAM that we highlighted earlier. And we'll just admit, I think we've been wrong for 2 years in a row forecasting a digestion related to China, and it's been stronger each year. So we could be surprised on the upside as we go through. But right now, Gary mentioned this a few minutes ago, China has been almost 40% of WFE. It's been elevated. They've been investing heavily to get to self-sustainability from a production perspective. So it's -- we don't have as good a visibility because of a large number of customers there and continuing creation of new customers. So it's not as easy to forecast as the rest of the mature market.
Gary Dickerson: Yes Atif, the one thing I would say also, longer term, what we think is that ICAPS remains about 1/3 of WFE, and then you have memory and leading-edge foundry/logic. So China is really mostly an ICAPS market and there has been this elevated spending over multiple years. We think that -- if you just look at end market demand, we think that moves back into that ZIP code over a longer period of time. Now when that happens, hard to anticipate exactly when that's going to happen, but that's what we think about over a longer period.
Atif Malik: And then, Gary, one for you. You visited several customers. Are your memory customers constrained by shell capacity? I mean why are we talking about a second half inflection or the first half? I mean I understand the first half is being dragged lower by China. But is there a situation where the memory guys are constrained by shell capacity that explains why second half is going higher?
Brice Hill: Well, I can take the question, Atif. So first, we would say that factory schedules, of course, matter. So each company -- each customer has their own schedule for installation and ramp and qualification and the like. Our information on the macro level, we had several people asked this question, but our information on the macro level suggests that there is factory capacity from a space perspective to ramp across the industry. And then you'll just have to ask each customer. That's a macro answer. You'll have to ask each customer for their situation. But we don't think the capacity is limiting the ramp at this point.
Operator: And our next question comes from the line of Charles Shi from Needham & Company.
Yu Shi: I have a follow-up on China. You did mention outside of restrictions, you are not losing share. But your U.S. peer, the closest one, reported much stronger China revenue growth this year. I think they are probably growing at 20% year-on-year. You're probably declining more than 10% year-on-year. So just really trying to reconcile how are you not losing share? Is there some nuances that we may be missing and that would suggest you actually did not lose share in China despite the numbers would suggest otherwise?
Brice Hill: Charles, I'll start on this one. So I think the nuance, we clarified that or we made an adjustment in there and said, if we look at the accounts that we can support, we're holding our share and competing well in China. And if you just take the macro number, we've certainly lost share in China. I think Gary commented that if you go back to 2024, just over 10% of the market was restricted from -- for U.S. companies. And if you go to 2025, it's more than doubled. The restriction has more than doubled. So we've certainly lost share in China because a larger portion of the market is unaccessible -- inaccessible to us. But when we look at the accounts that we can sell to, we feel we're competing very well.
Gary Dickerson: Yes, Charles, this is Gary. Yes, I think we have done a lot of work on this. And there's lots of different sources of data that can cross-validate the numbers. And again, when we look at -- we can't comment on other people, their ability to serve global customers that are in China versus domestic customers. Everybody is going to have a little difference in terms of what they're serving in China. Certainly, for us, NAND, the global NAND in China is not a significant business. DRAM, however, which really was restricted at the beginning of our fiscal year, we have very high share in DRAM. If you look at the top -- the leading DRAM companies in this last year, I think our business was up something like 50%. And so for Applied, DRAM going away had a really big impact on us. So again, Charles, I think we have pretty good confidence in maintaining share where we can compete, but there's going to be a different mix for every company. I don't know, Brice, if you want to add anything else.
Brice Hill: No, I think that covers it. Yes. The other thing that I would add to that, Charles, is the other part of that is not only the DRAM, but it's the NAND. So if you think about it, what was shipping? You have multinational NAND companies that are in China, and they were presumably pretty strong. And as you may know that in our mix, that's the lowest area for us. So Gary talked about the DRAM impact. The other impact is NAND being strong. And both -- and that equation in the future on a global basis, we think is changing.
Operator: And our next question comes from the line of Joe Quatrochi from Wells Fargo.
Joseph Quatrochi: I was wondering if you could help us understand just what was the size of the advanced packaging business in fiscal '25? And how did that grow?
Brice Hill: Joe, it's Brice. It's a little bit lower than it was in 2024. So 2024 was buoyed by the shipments, the high shipments in the end of the year for HBM. So it's essentially been running at the same level for 2025, minus that extra shipments at the end of that year. And then as we look forward, Gary is going to comment about the changing technologies for advanced packaging over time. We think we're well positioned for that. We still think that business will be growing significantly along with AI data center and all the different packaging capabilities that will be required over time.
Gary Dickerson: Yes, Joe. So Applied is #1 in high-bandwidth memory. And so that business was not as strong in '25 as it was the previous year. So that was one thing that impacted our relative growth rate, but it's pretty close to flat year-over-year '25 versus '24 with the HBM coming down versus where it was. What I would say is that we're still on track to what we've discussed before. We've grown the business around $500 million to about $1.5 billion today. And we're on track still to double this business to $3 billion or more over the next few years. And we're in a very good position. As I said earlier, we're #1 in HBM. So as HBM ramps over time, that puts us in a good position. We have a great portfolio of products. And especially, this is -- I mentioned the trip in Asia and meeting with a lot of these R&D leaders. There's an enormous focus for our customers to go to larger packaging sizes so they can connect more GPUs, CPUs and all these different computing components to improve performance and power an enormous amount. And so Applied is extremely well positioned in HBM, and we're also well positioned as these new technologies ramp for the future. So I think high confidence in being able to double this business again over the next few years.
Joseph Quatrochi: That's helpful detail. Maybe as a follow-up on the China entity list of $600 million that maybe comes back here. Any help on what the split of that was for like AGS in terms of services? I assume that would basically just kind of turn right back on given you're able to service those customers.
Brice Hill: Yes. You're right. There was some AGS in that. We didn't offer that, and I don't have it at my fingertips. So there was a component of AGS, but I think the majority was equipment.
Operator: And our next question comes from the line of Shane Brett from Morgan Stanley.
Shane Brett: So first question is, you mentioned that your revenue from leading-edge DRAM customers is up 50% in fiscal '25, which is stronger than those customers' DRAM WFE spending to my knowledge. Where specifically are these share gains with these leading-edge customers coming from? And now that you're past extremely tough China DRAM comps, just how do you expect the DRAM business to track in 2026?
Brice Hill: Yes. I'll start, Shane. So when you do the year-over-year comparison, it looks flat for us from a DRAM perspective. But what was happening in 2024, we shipped a significant amount of business to China customers. And so we basically had the same business in 2025 without the China customers. So all of that business went to the internationals, and that equated to approximately a 50% growth for them. And we do expect DRAM to strengthen as we go into '26 and should be healthy growth in the -- across the customers for DRAM investments.
Gary Dickerson: Yes, Shane, just in terms of where are we gaining? Certainly, I mentioned earlier, high-bandwidth memory, that's one part -- one segment where Applied is clear #1, and that is helping us. The DRAM companies are driving innovation in I/O and Applied, that really leverages a lot of our leadership products in foundry/logic. In fact, they're moving to FinFET going forward. That puts Applied in a good position even to gain more share as we go forward. We have a strong position in capacitor scaling. We have a really strong etch share in DRAM, and we've achieved patterning share gains. So those are a few areas. And again, I would say, going forward, as these DRAM companies adopt hybrid bonding for HBM, as they adopt FinFET, as they adopt 4F-square, the vertical channel transistor architecture in all of those areas, we're positioned to gain share as those new technologies ramp.
Shane Brett: Got it. And my second question may be a little bit critical, but your fiscal '25 foundry/logic revenue is up about 3% year-over-year, but your U.S. peers who should be on a level playing field with you are recording extremely strong double-digit growth through the first 3 quarters of the year. Just what's your self-assessment on this performance gap with those U.S. peers? And just how should we have confidence that Applied can outperform foundry/logic WFE in 2026?
Gary Dickerson: So foundry/logic includes both leading edge and ICAPS customers. So those 2 things are mixed together. If you look, we had record revenue in Taiwan, record revenue in Korea. And so our performance is extremely strong. And the leading edge, our revenue was up a significant amount. So you have all those things kind of mixed together, I think, in those -- in that data because everybody is combining that.
Michael Sullivan: And operator, we have time for one more question, please.
Operator: And our final question for today comes from the line of Jim Schneider from Goldman Sachs.
James Schneider: Relative to the strength you're expecting in '26 across leading-edge logic and DRAM, can you maybe guess which one might grow stronger than the other and which one might start to recover first?
Brice Hill: Sure, Jim. We think leading-edge will be the strongest grower with DRAM second. So that's our forecast.
Gary Dickerson: Yes. The one thing I would say, Jim, I think both of them are going to grow at a pretty strong rate. But it really in terms of when we see the revenue depends on fab timing. And so that's why we talked about kind of second half of calendar '26. And those were the discussions I had with a lot of the Asian customers when I was traveling and the dramatic improvement in visibility. So that also you have to take into consideration when you think about the timing of when those ramps will happen.
James Schneider: And then maybe as a quick follow-up, gross margins, you're expecting to increase in the back half once you get better absorption on the costs. I understand that. But are there any other underlying initiatives you're doing to improve gross margins operationally that could cause kind of a further tailwind once we get there?
Brice Hill: Sure. Thanks for the question. So on the gross margin side, if you look at '25, up 120 basis points from '24, the majority of that really has been improvements in our pricing processes, a good change that we made. We highlighted in previous calls that we revamped our entire pricing program across the company. But we're also working on cost reductions. The cost reductions were offset to some degree by tariff headwinds and by some of the inventory management, et cetera. So the real driver during the year was the price -- process improvements. And those will continue as we go into '26. So at the calendar second half, as we get more volume and you get some time for those to work, we should be able to improve the gross margins.
Michael Sullivan: And Brice, would you like to sum up today's call?
Brice Hill: Thanks, Mike. I'm excited that the investments we're making in leading-edge foundry/logic, DRAM and high-bandwidth memory give us technology leadership. At the same time, our customers are signaling an inflection in the AI-related fab equipment spending in the second half of calendar '26. I wish everyone a nice Thanksgiving, and I look forward to seeing many of you soon at the UBS Conference in Scottsdale. Mike, please close the call.
Michael Sullivan: All right. Thank you. And I'd like to let everybody know that a replay of today's call is going to be available on the IR page of our website by 5:00 p.m. Pacific Time today. In the meantime, thank you for your continued interest in Applied Materials.
Operator: Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.