Q3 2026 Earnings Call Transcript

Good day, and welcome to the Lam Research Corporation's March 2026 Earnings Conference Call. [Operator Instructions] Please note, this event is being recorded. I would now like to turn the conference over to Ram Ganesh, Vice President of Investor Relations. Please go ahead.

Thank you, and good afternoon, everyone. Welcome to Lam Research Quarterly Earnings Conference Call. With me today are Tim Archer, President and Chief Executive Officer; and Doug Bettinger, Executive Vice President and Chief Financial Officer. During today's call, we will share our overview on the business environment, and we'll review our financial results for the March 2026 quarter and our outlook for the June 2026 quarter.

The press release detailing our financial results was distributed a little after 1:00 p.m. Pacific Time. The release and the accompanying presentation slides for today's call can also be found on the Investors section of the company's website. Today's presentation and Q&A include forward-looking statements that are subject to risks and uncertainties reflected in the risk factors disclosed in our SEC public filings.

Actual results could materially -- differ materially from those expressed in such forward-looking statements. Please see the accompanying presentation slides for additional information. Today's discussion of our financial results will be presented on a non-GAAP financial basis, unless otherwise specified. A detailed reconciliation between GAAP and non-GAAP results can be found in the accompanying presentation slides.

This call is scheduled to last until 3:00 p.m. Pacific Time. A replay of this call will be made available later this afternoon on our website. And with that, I'll hand the call over to Tim.

Thank you, Ram, and good afternoon, everyone. Lam is off to a solid start in calendar year 2026, with revenues and profitability in the March quarter at the upper end of our guidance ranges and earnings per share exceeding the top end of our guided range. Revenues were at record levels, highlighted by the first $2 billion quarter from our Customer Support Business Group. Our guidance for the June quarter points to Lam's strong momentum in an accelerating AI-driven semiconductor demand environment.

In January, we shared our outlook for 2026 WFE in the $135 billion range. Since then, spending projections from customers have moved higher across all device segments. We now expect WFE of $140 billion with a bias to the upside as the industry continues to work through various constraints. We believe this sets the stage for another year of compelling WFE growth in 2027. For Lam, the AI-driven demand environment is creating an ideal setup for continued outperformance.

Semiconductor technology inflections required to meet escalating AI compute needs are driving higher deposition and etch intensity. In 2026, we see Lam's served available market or SAM, percent of WFE, expanding to slightly more than the mid-30s percent level, well on track towards our stated goal of high 30s percent over the next few years. Lam is prepared for this moment by transforming how we innovate, build and support. The semiconductor manufacturing equipment needed to address the industry's most critical challenges.

Our commitment to R&D and the velocity with which we have scaled our development capabilities have enabled us to create the broadest, most competitive product and services portfolio in the company's history. This is fueling our current outperformance and puts us in an excellent position to deliver on our future growth ambitions. Across all device segments, we are seeing greater opportunity for Lam. In NAND, AI transformation is moving beyond compute and into the storage layer. [ Hoken ] economics are driving changes to the memory hierarchy used in AI data centers including rising adoption of higher layer count QLC-based NAND devices for SSDs.

We expect total data center bits this year to be greater than both PC and mobile segments combined with growing -- continuing growth in data center mix into the future. The growing device performance requirements of AI data centers are driving an acceleration of NAND technology upgrades. As you may recall, we said in early 2025 that roughly $40 billion in conversion spending would be required over several years to enable existing NAND installed wafer capacity to produce devices with more than 200 layers. We now anticipate that this conversion will be pulled forward with the majority of spending occurring before the end of calendar year 2027. In parallel, we expect growth in bit demand will drive greenfield capacity investment, especially considering that overall industry installed wafer capacity is expected to decline more than 20% from prior highs by the end of this year.

Looking further ahead, we see continued adoption of NAND in the AI memory stack, driving even higher layer count NAND devices. With the largest installed base of tools for 3D NAND, Lam is uniquely positioned to benefit from this trend. As manufacturing complexity scales with layer count, we see an expanding set of deposition and etch opportunities, all rooted in our established leadership in high aspect ratio cryo etch, dielectric stack deposition, [ Worldline ] metallization, backside stress management and gap fill technologies. In dielectric etch, our Vantex and [ Flex ] tool sets delivered the industry's highest power density and productivity for dielectric channel hole edge applications, where we have a market-leading position.

In conductor etch, we are also seeing momentum for our [ Kio ] systems as customers collaborate with us to maximize device yield in a constrained capacity environment. In a recent win, a customer switched to [ Kio ] in the middle of their production ramp due to superior defect performance and better yield. In deposition, we are seeing the transition to higher layer count NAND, also driving greater demand for our Strata, [ Altus Halo ALD ] and [ Vector DT ] products. Altogether, we believe the production proven strength of our portfolio puts Lam in a great position to outperform overall NAND WFE growth as AI demand accelerates over the next few years.

In DRAM, AI's power and efficiency requirements are driving an industry transition to 1C generation devices. As feature dimensions shrink, the industry is shifting from traditional silicon nitride base dielectric films deposited using furnace, the more advanced ALD silicon carbide [ loc ] layers to achieve bit line capacity in production. Studies have shown the re-architected device structures, combined with low [ KBitline ] spacers can reduce capacitance by over 60%. Lam's Stryker carbide solution with its unique plasma source enables capacitive scaling by depositing dense, conformal and tunable low dielectric films with high productivity.

As a result, our Stryker based solutions are the tools of record at all leading memory makers for bitline spacer applications. As the industry moves to 1C nodes, we see our total dielectric deposition SAM and DRAM growing more than 20%. With innovations like Stryker ALD, we believe Lam is well positioned to gain share within this expanding opportunity. In foundry/logic, calendar 2025 was a record year for Lam.

We are carrying that momentum into 2026 as we capture more opportunities from inflections at the leading edge. Most notably, this quarter, we achieved both dielectric etch wins at a key founder logic manufacturer. Our first dielectric edge wins at this customer. And finally, we see growing demand for our advanced packaging solutions where we bring unmatched experience in equipment design and process technology for copper plating and [ TSC Edge ].

Lam's advanced packaging revenue growth is expected to exceed 50% in calendar year 2026. Turning to our Customer Support Business Group. We delivered our first $2 billion plus revenue quarter. Demand was strong across spares, upgrades and services. As customers look to improve fab output in a space-constrained environment, more opportunities are being created for CSBG to deliver innovations that increase productivity and enhance yield for our customers.

Our services business posted mid-teens growth over the December quarter. Highlights included a new agreement with a leading foundry/logic customer to deploy our equipment intelligence services for critical deposition applications. The top memory customers also set to utilize our Equipment Intelligence capabilities in R&D to enable faster ramps of new nodes for NAND and DRAM production. We are also gaining momentum with our Dextro cobots, which deliver an unprecedented level of automated tool maintenance precision and repeatability.

Customers using Dextro and production are benefiting from higher output and in some cases, improved yield from existing capacity. In the March quarter, we expanded Dextro coverage to 8 Lam tool types up from 6 last quarter. We also introduced the next generation of Dextro, which packs 10x more compute power than the first generation into a smaller footprint. This quarter, we will ship our first Dextro cobot for a deposition product further increasing our ability to create value from our overall installed base more than 100,000 chambers.

It's an exciting time for the semiconductor industry and for Lam. In an accelerating demand environment, we see rising deposition and etch intensity creating a multiyear outperformance setup for Lam. We have made strategic investments across the company to capitalize on this opportunity. increasing the velocity of both our technology development and our operational execution. Our progress can be seen in our strong March quarter results, our higher June quarter outlook, and our expectation that second half calendar year revenues will exceed the first half.

In short, we are delivering on the tremendous opportunity in front of us with more to come. Thank you, and here's Doug.

Excellent. Thank you, Tim. Good afternoon, everyone, and thank you for joining our call today during what I know is a very busy earnings season. Lam is off to a solid start in 2026, building on the momentum we delivered across 2025. In the March quarter, revenue gross margin and operating margin came in above the midpoint of our guidance ranges, while earnings per share actually exceeded the high end of the range.

We also achieved our third consecutive record revenue quarter. March quarter revenue came in at $5.84 billion, which was up 9% sequentially and up 24% from the same period in 2025. The deferred revenue balance at quarter end came in at $2.22 billion, which was flat sequentially. Within this balance, however, customer down payments came down by roughly [ $300 ] million, while the other line items increased with the growing business levels. I just mentioned that down payments are now at the lowest level we've seen in nearly 4 years.

From a market segment perspective, foundry accounted for 54% of our systems revenue in the March quarter, which was down from 59% in the December quarter. Revenue in dollar terms was approximately flat sequentially, and it was up 35% year-over-year. Foundry saw strength in investments at the leading edge, as well as ongoing mature node spending. Advanced packaging within foundry continues to be an area of solid growth for us.

Memory was 39% of systems revenue, up from 34% in the December quarter. Within memory, we delivered record DRAM revenue accounting for 27% of systems revenue which was up from 23% in the December quarter. High-bandwidth memory investments remained strong. The profile of spending is also gravitating towards the 1C node and beyond enabling the ramp of DDR5 and LPDDR5. Nonvolatile memory contributed 12% of our systems revenue, up slightly from 11% in the December quarter.

As Tim outlined, AI workloads are accelerating demand for higher capacity NAND and Lam continues to benefit from strong leadership within this segment. We expect to see growth in NAND investments throughout the remainder of the year as the industry converts to 256 layer and above class devices. And finally, the Logic and Other segment came in at 7% of systems revenue in the March quarter, in line with the prior quarter. Let's turn to the regional breakdown of our total revenue.

China came in at 34%, which was a slight decrease from the prior quarter level of 35%. We expect that China revenue in the June quarter will decline from these levels. Korea and Taiwan each came in at 23%, which was both up from 20% in the prior quarter. Both the Korea and Taiwan regions represent record revenue level in dollar terms in March.

And I just mentioned that this regional mix was generally in line with our expectations from the beginning of the quarter. Customer Support Business Group generated a record $2.1 billion in revenue in the March quarter, which was up 6% sequentially and up 25% from the same period in 2025. Sequential growth was driven by a large and expanding installed base and the continued expansion across our spares, upgrades and services business, partly offset by Reliant. Growth in spares and service is benefiting from strong factory utilization across the industry. Let's take a look at profitability.

Gross margin in the March quarter was 49.9%, which was at the high end of the guidance range, driven by multiple factors, including favorable customer product mix as well as improved factory efficiencies. Operating expenses in the March quarter came in at $866 million, up from the prior quarter's level of $827 million. The increase was driven by seasonal employee-related costs as well as higher headcount to support our growth. R&D accounted for 68% of total operating expenses.

We will be growing R&D investments throughout the remainder of the year. March quarter operating margin was 35% at the high end of our guidance range due to the higher revenue and the improved gross margin. The non-GAAP tax rate for the quarter was 9.2%, which came in lower due to benefits from higher equity compensation vesting, which is deductible on the taxes during the quarter. We continue to see the tax rate below the mid-teens for calendar year 2026. Other income expense in the March quarter was $8 million in expense compared with $10 million in income in the December quarter.

The variance in OI&E was primarily the result of small losses in our venture portfolio as well as lower interest income. Interest income decreased due to the lower cash balance in the quarter. And as we've talked about in the past, you should expect to see variability in OI&E quarter-to-quarter. For capital return in the March quarter, we allocated approximately $800 million to share buybacks through a combination of open share repurchases and a $200 million accelerated share repurchase transaction.

Our average buyback price was approximately $211 per share. We also retired $750 million of unsecured notes that reach maturity using cash from the balance sheet. Additionally, we paid $326 million in dividends. In the March quarter, we returned 139% of our free cash flow.

Our plans remain to return at least 85% of free cash flow to our shareholders over time. The March quarter diluted earnings per share came in at a record of $1.47 which was above the high end of our guidance range. The diluted share count was 1.26 billion shares, which was flattish with the December quarter and consistent with our guidance. And I just mentioned that we have $4.3 billion remaining on our board authorized share repurchase program.

Let me pivot to the balance sheet. Cash and cash equivalents totaled approximately $4.8 billion at the end of the March quarter, which was a decrease from $6.2 billion at the end of the December quarter. The decrease was primarily driven by capital return activities that debt paydown as well as capital spending. Days sales outstanding was 64 days in the March quarter, an increase from 59 days in the December quarter.

Inventory turns improved to 2.9x from 2.7x in the prior quarter. These were our highest level of inventory turns in over 4 years. As a company, we remain focused on our strong asset utilization and return on invested capital. We're pleased with the sustained performance we continue to deliver here.

We will be managing our inventory and supply chain to align with the growing demand that we see in front of us. Noncash expenses in the March quarter included approximately $97 million in equity compensation, $103 million in depreciation and $13 million in amortization. Capital expenditures in the March quarter was $332 million, which was up $71 million from the December quarter. Spending was higher to support the strong demand environment that we're seeing.

Investments are enabling a second manufacturing facility in Malaysia as well as lab-related investments in the United States and Taiwan. Looking forward, we continue to expect capital expenditure to be in the 4% to 5% of revenue range. We ended the March quarter with approximately 20,600 regular full-time employees which was an increase of approximately 900 people from the prior quarter. Headcount increases were primarily within the manufacturing and field organizations to support volume growth. as well as in R&D to support our long-term product road map.

As we scale the organization, we also undertook a small workforce optimization focused on efficiency. You'll see this in our non-GAAP reconciliation. Let's turn to our non-GAAP guidance for the June 2026 quarter. We're expecting revenue of $6.6 billion, plus or minus $400 million.

Gross margin of [ 50.5% ], plus or minus 1 percentage point. We're expecting this expanding gross margin despite slight headwinds that we're seeing from customer mix. Forecasting operating margins of 36.5% plus or minus 1 percentage point. And finally, we're forecasting record earnings per share of $1.65, plus or minus $0.15 based on a share count of approximately 1.255 billion shares.

So let me wrap up. We're executing well against our financial objectives and driving operational efficiency while increasing R&D investments to extend our technology leadership. With our expanding installed base, the strength of our product portfolio and our disciplined approach to capital allocation, we remain confident in Lam's setup for continued outperformance. Operator, that concludes our prepared remarks.

We would now like to open up the call for questions.

[Operator Instructions] Our first question comes from Timothy Arcuri with UBS.

Doug, I wanted to ask about gross margin. The guidance is great, it's [ 50.5% ]. It sounds like despite the mix being against you. So you're kind of already edge at your target model, right, because you were saying above 50. And I'm not asking you to update that model.

But sort of can you like deconstruct how you got here so fast? And maybe also, I think people want to hear how much capacity you have? I know you mentioned that you're adding another site in Malaysia. So can you just speak about sort of what the puts and takes are going to be on margin going forward?

Yes. No, Tim, it's a great question. Yes, I think we're pretty pleased with where we're at from a gross margin standpoint. And it's been a lot of real hard work from the company, honestly.

I think you'll remember, I don't know, 4, 5 years ago, we talked about expanding our factory footprint to be closer to where our customers were. And that has delivered efficiencies from just a proximity standpoint, from shorter frame logistic lanes. From slightly lower cost from a labor standpoint, a better supply chain set up all of those things. Those were self-help activities that we undertook and frankly, we've delivered on it.

So when I think about the global operations part of the company, they've really done a wonderful job. On top of that, we're working on everything we can do to get paid for the value we're delivering to customers. That's something we're always doing. And I think we're doing a reasonably good job with that, Tim.

So anyway, when you put all of that together, I think we're pretty pleased with all of that, and I'll let Tim add a few things here.

Yes. No, I was just going to add one other part that's pretty important in showing up is very important in this constrained period, which is the performance of our tools. We embarked -- Doug talked about some of our higher R&D spending. A lot of that was to ensure that all of these new tools that we have hitting the field enter the level of maturity that's beyond what we had probably delivered in the past.

And that's very important for our customers in a period of fast ramp. That also yields benefits for us in terms of installation and warranty spending, which flows through to gross margin. And so you're seeing some of that as well. So that's something that, again, we're focused on going forward is reliability of systems, maturity of tools as they hit the fence.

And let me just add one more thing. Yes, let me add one quick thing. I know there's going to be a question [indiscernible], how should our model gross margin for the rest of the year? I would encourage you to kind of keep it roughly in the levels that we just guided you to in June.

This is going to kind of level out at where it's at, I think, for the rest of the year. So as you build your models, keep that in mind.

Awesome Doug. And then I guess just as a follow-up. So there's been a big like massive new fab project. I mean you guys have obviously seen this news bigger than anything we've ever seen before.

I mean I think that this customer would have to get in the queue given how booked out things are. Are you seeing -- I mean, I don't want to ask just about that one customer. But like are you seeing these signs of these huge new fab projects, sort of the customer base expanding? And if you did want to comment on that particular fab project if they're sort of coming to you was like a new opportunity, that would be great.

Yes. Obviously, we can't comment on any specific customer, but clearly, the environment right now is such that there is -- there's just not enough to do. There's not enough memory in the world. And so -- and people are worried about supply.

And so I don't think it's a surprise that more companies around the world will start to enter into the semiconductor space. And that's why when we talk about the longer-term outlook, it's a combination both of the increased demand, but really increased demand at some of the most compelling leading-edge opportunities that are presented to Lam. And so I think this is -- when we talk about WFE, there's so much -- only so much that can be executed in this year. But again, you see a lot of these projects starting to line up that I think represents opportunity in the future.

Our next question comes from C.J. Muse with Cantor Fitzgerald.

I just wanted to touch on your commentary around '27 visibility. And can you kind of speak to your discussions, conversations exceeding 18, 24 months, whether you're starting to see real slotting and desire to lock in time frames for delivery? And I guess as part of that, how are you kind of working your supply chain for readiness for that ramp?

Yes. So clearly, we're about halfway through '26. And so given our lead times, of course, we're having conversations with customers about '27. And in some cases, for planning purposes, like getting resources ready engineers hired and trained in the right locations, those -- some of those conversations even extend beyond that. We have customers who clearly announced fabs with openings in 2028. There's no reason not to start having conversations with them about what the tooling is it's going to be required based on the node that we run, kind of the size, the resourcing requirements. So I'd say we're in various stages of those conversations, but the more visibility we have, the better we can get our supply chain and our own capabilities ready.

I think it is a case where today, our view on WFE, as I said, for 2026 really has a lot to do with what we believe can be executed. We talked about this upward bias. We're working a lot with customers on near-term constraints, things they can do within their existing fabs. But at the same time, preparing for those new fab openings and true kind of greenfield shipments as they roll out later this year and through next year.

Yes. [indiscernible], I'd just add, it feels like it's setting up to be a pretty good year in '27 right now based on what we can see.

Excellent. And then maybe a question on CSBG. Obviously, tremendous focus on trying to get every bit out the door in this very tight environment. Curious if kind of the upgrade business that you're seeing is sustainable? And is there kind of [indiscernible] we should be thinking about for full calendar year '26 revenue growth in that bucket?

Yes, [indiscernible] that's a great question. Look, I think we're feeling really good about CSBG, industry utilizations are high. So spares was quite strong in March. Service was quite strong in March.

Tim talked about the new Equipment Intelligence and cobots that we're rolling out. We're excited about that. Our customers are excited about that. So when you see how strong it was in March, I think it popped up, I think it's going to kind of sustain roughly at these levels as we go through the remaining quarters in the calendar year, maybe up a little bit.

But I think we're feeling pretty good about the strength that we're seeing here. And frankly, we're innovating here, too. So I think we feel pretty good.

Our next question comes from Harlan Sur with JPMorgan.

When you -- when I speak with the process development and integration engineers, obviously, of your customers, they're very focused on next-generation technologies and architectures and that's what we hear on these calls, right? How Lam is enabling 3D device architecture, cell structures, driving high aspect ratios, new materials, et cetera. But then when we speak with the manufacturing and operations teams, it's a very different focus, right? And the vocabulary set is very different.

It's all about throughput, uptime, defectivity, overall fab cycle time. And then especially with the tight supply situation and constrained premium space environment that we're in today, any incremental improvement in high-volume productivity could unlock like literally millions of dollars of incremental wafer output. You've talked about things like the Dextro cobot, but any other enhancements that you're driving, Tim, to the installed base on productivity and manufacturability and more importantly, like, how are you guys monetizing this? I assume it's maybe primarily services and upgrades?

Yes, it is a too-focused world, as you talked about. And the good news is we've got the company organized in a way that we can focus on both with significant intensity. So clearly, leading edge being in front of those inflections, a number of years, we said sometimes 5, 6, 7 years, you're working with the customer in advance of that node ever reaching production. But at the same time, especially in the environment we're in right now, I mean, production output, uptime yield, those things are really what are most critical to customers in the immediate term.

Plus, I would say, really identifying the bottleneck tools within the customer that's limiting output and helping them with those workstations. Equipment Intelligence, if you think about what it does is it allows us to look at massive amounts of data coming from our tools on every single wafer, and that shortens troubleshooting time if there is a problem with the tool. It helps us with the time to ramp those tools either on new process or as they start up, helps us to match tools, better tool to tool chamber to chamber, all those things can yield -- lead to those tiny little improvements in yield that really do matter for the customer.

On the Dextro cobot, we've talked about the fact that at some customers, the precision and repeatability of the maintenance has actually yielded improvements in both output and yield and does through -- that through better first time right. You do the maintenance, it comes back up and is back into production more quickly. And also just the improved repeatability of, like, let's say, the new part placement inside the chamber, actually has had positive effect on the yield need. So that's something we're really focused on.

How do we monetize it. Yes, it's through services and obviously, in some cases, new tools.

Yes. Okay. I appreciate that. And for Doug, your OpEx grew 5% sequentially in the March quarter, implied OpEx growth in June is 7% and given the leverage, it's allowing you to actually exceed your long-term operating margin target of 35%.

So how should we think about the OpEx growth through the remainder of this year? And I guess when is the team going to update its long-term targets? Because as the year unfolds on more revenue growth, you're clearly going to drive margins above the 36.5% op margin range that you guided to for June, right? So when is the team contemplating like updating this long-term targets?

Yes. Harlan, it's a great question. First, let me talk about the spending trajectory for the year. Listen, I think at the end of the day, this management team likes to see the top line growing faster than spending so that we can deliver leverage and that's absolutely how we're thinking about things this year.

Having said that, we're going to grow spending this year because, frankly, we can afford to do so, and we have some things that I think are quite innovative that we've been thinking about that we've wanted to put a little more money towards. So we're going to do that. We've decided we're going to do that this year. And yes, we're talking internally about the fact that we're above the previous model that we gave.

And yes, I know we need to give you an updated framework and we will do that later in the year. And we haven't bottomed out on exactly when or exactly how we're going to do it, but we know we need to and we will be doing that, Harlan.

Our next question comes from Atif Malik with Citi.

My question is on the NAND market. It seems like near-term NAND is still low, like 12% sales, but something has changed versus 90 days ago, you guys are talking about NAND growing through the year and the pull forward in that the $40 billion number. So can you [indiscernible] has changed in the NAND market? Are you seeing signs of capacity additions? Or what has changed maybe with [ KB Cash ]?

Yes, we didn't mention [ KB Cash], but I think it's a good example of exactly what I was referring to when I talked about it, it's increasingly important role in the AI memory hierarchy. And so clearly, there is increased demand for NAND coming from AI data centers, and that's helpful. But also, if you think of the -- on a relative basis, what under investment in that area, partly as customers make choices about clean room allocation and obviously some other devices like HBM were so hot during that period. Also, going back to what we said early last year, the installed base had gotten a little bit behind in terms of the state-of-the-art technology.

And so most of the installed base at that time, early 2025, about 2/3 of it was still running in the [ 1xx ] 100-plus layer technologies really when you need to get those incremental bits out now, you need to be 200-layer plus. And so that's what's caused this acceleration is you need more bits. You need those bits to be more capable, you need QLC to meet AI data center demands and so you've started to see the push for accelerated conversions in the technology. And that is -- that's what caused a lot more activity in the NAND space.

As people push forward, we didn't also said, look, the conversions are going to happen because that's very -- the quickest way to get to the high capability, but you'll also need greenfield because those technology improvements like in Lam's case, to go above 200-layer, we talked about the number of new tools you need to add to manage the complexity of higher layer count stacks that in itself reduces total wafer output capacity of the industry. And so eventually, you need to add greenfield back to continue to get the big growth we need. So that's the reason we started talking about it is it's materializing as a significant opportunity now on the revenue side for Lam and looks to be so for quite some time.

Doug, you talked about customer down payments at [indiscernible] level in 4 years. And you're also talking about WFE growing in next year. Can you reconcile those 2 comments?

I guess what I would tell you, Atif, is the group of customers that generally provided on payments aren't the ones that are growing the quickest, and that's absolutely what we're seeing going on right now.

The next question comes from Melissa Weathers with Deutsche Bank.

I had a more thematic question maybe for Tim or Doug, if you want to take a stab you can do. We've heard a lot of about reasons why this memory cycle is different with HBM and trade ratios and new applications like [ SOC ] and it does seem like AI is driving memory demand growth a lot faster than what we've seen historically. So I guess, do you ascribe to the view that this memory cycle is -- I won't say the D-word, but there's a change this time around? And then what kind of actions are you taking to derisk the cyclical side of things while still being able to capture the upside?

Okay. Well, it's a great question. And maybe I won't use the D-word either, but I think it's -- or maybe I will. I think it's different for Lam in that -- and there was an earlier question that talked about how so many of these new devices have different architectures, 3D scaling.

And so I think the most important thing about this memory cycle is it is a cycle in which you're seeing dramatic improvement and change in the etch and dep intensity. And so the complexity of 3D scaling has created a lot of new opportunities for Lam. And so that is driving both SAM expansion plus share gain for us through those new applications. So I feel like compared to prior upturns in memory, we are and we're doing even better just because of that extra layer of etch and dep intensity scaling.

How do we prepare if there is ultimately that peak, which we're not -- we're certainly not calling right now given the tremendous demand that's out there. But it is we operate very flexibly. I mean, Doug talked about a lot of our operational investments we've made. And in many cases, some of the things we talked about, Dextro cobots, Equipment Intelligence.

These are all kinds of capabilities that in many ways, allow us to support our customers without so much of the fixed cost scaling that we had to make in the past. And so we always have an eye on what's it going to look like if the business were to slow down. And I think if you look at our track record, in those periods, we've also outperformed.

And Melissa, maybe I'd just add. I mean, the way I'm looking at this right now is memory is just so critical in all of these accelerated compute architectures to feed the parallel compute, you need just data coming in to keep the machine going. And so the criticality of it maybe is more than it's ever been from my point of view. And I observe -- maybe I'll use a different D-word disciplined investment, right?

I mean, everybody likes profitability that they're generating right now. Everybody is just kind of lugging into where demand is. And I think that's a good thing for all of us in the industry.

Our next question comes from Srini Pajjuri with RBC Capital Markets.

My question is on China. Doug, I think your comment about prepayments being down. I'm guessing that's related to China. Can you talk about what you're seeing in terms of the demand environment in China? And as you go through the next few quarters, what are your expectations?

Yes. I think, Srini, what we described a quarter ago is still the way I would describe it this year. I think WFE in China is flattish year-over-year from '25 to '26, maybe it's up a little bit. But you're just seeing so significant growth from the global multinational set of customers that China as a percent of the overall revenue is coming down.

The other dynamic in China is you're starting to see some of the global multinationals in China spending a little bit more, too. So when you look at that overall geographic distribution in China, it's also broadening out in that regard. And yes, you're right about the fact that down payments are down -- down payments tend to come from smaller customers, and a lot of them are in the China region, and so those 2 things are correlated together.

Okay. Great. And then my next question is on the CSBG. So obviously, I think it grew at a double-digit pace for the last several years.

And I think last quarter, if I recall correctly, I think you were expecting high single digits because of the reliant slowdown here. But it does seem like the clean room issue is not going to get resolved. Demand is very strong. So my question is, should we -- I mean, are you seeing any acceleration in terms of your services and spares business?

Is this something structural in your view going forward?

Listen, Srini I'll let Tim comment after I give you a little bit of data. What drives a lot of spares and service, frankly, is utilization in the overall industry. Utilization right now and in the March quarter, is very, very high. And so a lot of the growth at least contributing to some of the sequential growth in CSBG was the uptick in spares and service from that utilization.

I don't know that utilization can get any higher than it is. Frankly, it's pretty full out right now. And so when you think about growth sequentially over the next couple of quarters, those components of CSBG are probably kind of plus or minus where they are. Now Tim talked about advanced service and cobots and [ EI ], that layers on top of that to a certain extent.

And then also, if you think about what's going on in mature node spending, a lot of that is what drives Reliant and that's flattish this year. The real growth is coming from stuff at the leading edge, which we're really benefiting from move to etch and dep intensity. So Anyway, that's just a few things to think about relative to CSBG. Anything you'd add, Tim?

No, not really. I'd just point out that you're trying to work on constrained workstations within a fab. Again, this is where things like the Equipment Intelligence, how to get those tools up faster for production. There's a lot of focus on that.

That's the short-term prove out. And I think that long term, that then has a real benefit because once the value has been seen in this kind of constrained environment, I think that it will be more likely that new fabs get built with all of those intelligent services and automated maintenance capabilities built in right from the start.

Next question comes from Vivek Arya from Bank of America.

To many of your memory customers are talking about long-term contracts, LTAs, pricing arrangements and whatnot. How is that translating into your visibility and pricing power? Should we expect customers to start putting down payments to secure your capacity also? And if not, why not?

Well, it's a good question. I would say that it's translated into a longer visibility for us. As I mentioned in an answer earlier, clearly, we're having conversations with customers now at around the time that they're starting to construct these fabs, it means we have much longer visibility. And I think the most important thing there is to be ready with the resources that are needed and our own capacity to be needed to support those shipments.

And so I would say, we're working with customers today short term in their existing fabs. We're working with them with these long-term fab plans and being ready. And in many ways, that's allowing us to be more efficient. As Doug talked about disciplined build out in our operational capabilities, our manufacturing, our supply chain.

I would say that it is translating into financial benefit for Lam as well by having those longer visibility conversations.

And Vivek, I mean, listen, we're having very long-term conversation with customers, but we don't need down payments. We generate ample free cash flow from the business we run the commitments we're going to get from customers are important and significant and they're happening, certainly, but it doesn't require down payments for us.

Got it. I guess maybe the subtext of my question is the gross margins that you're seeing, right, the 50.5%, how durable are there? So let's say if memory pricing goes down next year for whatever reason, do you still think these gross margins are sustainable? And maybe you can even expand from these?

Or do you think these gross margins are because the industry is so tight today. So I'm not asking for a gross margin forecast per se. I'm just trying to understand that if you're customers are getting assurance of their pricing? Is there anything Lam can do to help get assurance around your pricing and your -- the sustainability of your margins over the next 1, 2 years?

You know what -- actually, we're not going to give you a gross margin forecast longer term. But I think that what you can see and what we've said is we have been building the gross margin improvement in our company around fundamental capabilities, either our own through our own operational efficiency or through the value that our equipment delivers. And that can be technically as the manufacturing becomes more complex, it can be the unique capabilities our tools provide from a technical or a productivity perspective. And so we have moved at a pace where we feel like the improvements we're making are sustainable because they're rooted in real value or real efficiency.

And they're not a -- they're not leveraging sort of the opportunity, and they're not transactional in nature. They're really founded in fundamental value delivered to the customer. And when I talk about things like cobot, for instance, the value of a cobot is rooted directly in the value being delivered to the customer through better uptime, better yield, and we get paid for that. And I think those types of things are sustainable.

When we deliver technology that enables the move to the next technology node. We think those are sustainable regardless of the cycle because it is delivering value to the customer. And that's -- we're in this for the long term with our customers, and that's why we look at, at all of this.

Our next question comes from Jim Schneider with Goldman Sachs.

I was wondering if you could maybe comment on in terms of the WFE uptick you expect which of the product areas do you expect the most kind of incremental leverage? Is it kind of split across all of them? You talked about advanced packaging, but which was driving the most upside to the overall spending envelope this year do you believe? And is that being driven mostly by early fab clean room pull-ins or something else?

Yes, Jim, I'll comment and then if Tim wants to add I'll let him do that. I think the reality of it is everything is a little bit stronger. I think everybody in the industry is working on finding a little bit of clean room that they've been able to just accelerate to a certain extent. Demand has always been there.

Demand is as strong as I can remember it. Frankly, it was strong 90 days ago. It continues to be maybe even a little bit stronger right now, and everybody found a little bit more clean room and so they were able to take a little bit more equipment.

Our next question comes from Stacy Rasgon.

For the first one, I wanted to push a little bit more on the services growth. So I understand the drivers around utilization topping and the Reliant weakness. But I mean, you also talked about the $40 billion in upgrade spending pretty much all happening by the end of '27. I don't get the feeling that we've had like tens of billions of that upgrade spending happening already. So it almost feels like we should have tens of billions of upgrade spending happening between now and the end of next year.

And from what I understand, I thought that all goes into your services business. So why shouldn't that be a pretty big driver of services growth, I guess, between now and the [indiscernible].

Yes. Stacy, I would point out a couple of things to you. In that $40 billion number. Yes, there's upgrades for sure, but there's also new equipment purchases, right?

There's some new things, right? When you upgrade the installed base, you need to buy new equipment to break bottlenecks and constraints. There's also some new equipment as the industry moves to moly. So it's not all just upgrades.

And the other thing I would say relative to upgrades is upgrades were actually quite strong last year in '25 and are going to continue to be for the next year or 2. So that's part of the upgrade story. And then the other components, like I said, spares and service. It's already pretty darn strong in March and, frankly, Reliant with the mature node spending being a little bit softer than everything else, that's the puts and takes to get you to kind of quarter-by-quarter plus or minus flattish as you go through the rest of the year.

Okay. That makes sense. If I could ask a follow-up. So you guys are seeing WFE growing this year on the order of, what, $30 billion, like you said, 110 last year to now 140 plus this year.

And that's very strong, but it strong that it is, as you know, it is a constrained growth because of clean rooms. And those clean rooms start to come online into next year. Does that suggest to me that the sequential growth of WFE next year ought to be even stronger on a dollar basis than it is in '26 because you'll have some were to actually put the tools, whereas you don't really have that this year. Like what's wrong with that logic?

How would you push back on that?

[indiscernible] to decline to comment on the exact magnitude of WFE next year, but we do firmly as we sit here today, look at clean rooms are going to be more available next year and where we believe demand to be WFE is going to be nicely growing next year. And it's too soon for us to give you a number, but we feel pretty good about the growth trajectory into next year.

Yes, I'd also point out that every year that goes by, as technology advances, etch and deposition intensity rises. And so as those new clean rooms come on and they're targeting more advanced technology nodes, that's better for -- certainly better for Lam's position within whatever the term I used compelling WFE growth is.

Our next question comes from Krish Sankar with TD Cowen.

I just want to follow up on an earlier question on the upgrade to the WFE numbers, the $135 billion going to $140 billion plus. Is there a way to segment was the bigger driver NAND? Was it CPU tightness? Or was it just AI strength?

Krish, what I said is everything got a little bit stronger because everybody got the little bit of [indiscernible] clean room. So it's not any one component of the customer base. Everything is just a little bit better.

Got it. Got it. And then as a quick follow-up. It looks like the third-party market share data came out and you folks gained share in PECVD quite a bit last year. I'm curious which vertical drove that PECVD share gain? Was it DRAM or foundry/logic or something else?

Do you want to take that Tim? You want to [indiscernible]?

Sure, go ahead, Doug.

Listen, I think PECVD is such a broad pervasive tool. It shows up in every component of the customer base. One area I think that sometimes is underappreciated is the use of PECVD and underfill in advanced packaging, honestly. And that was a key contributor. Tim talked about we see packaging this year growing 50%. We talked about real strong growth last year. PECVD benefited from that, obviously.

Yes. I think PECVD also shows up. It's challenging because you think about the old traditional PECVD applications. But even as I mentioned, as we move forward in NAND, for instance, even like our Vector [ DT ] backside stress management actually is a PECVD application. So in many ways, it's such a pervasive technology, and so we see that improvement in PECVD.

Our next question comes from Joe Quatrochi with Wells Fargo.

I was wondering if you could talk a little bit just about where your lead times sit today? And then also, I think you talked about the second Malaysia factory opening. When is that ramp? And can you remind us like what is the size of that relative to, I think it was a pretty large first facility that you have like 700,000 square feet.

Joe, first saying, we don't specifically put numbers around our lead times, but they are stretching out a little bit as demand is obviously quite strong. So [indiscernible] I'm going to give you a number though. Second, the second [indiscernible] facility, we'll come on the second half of the year. And yes, you're right.

The first one was our largest factory in the network. This will be nearly the same size or maybe approximately the same size as the first one. So it will give us the opportunity to scale into the next year's demand, I think.

That's helpful. And then I was just curious, I was wondering if you could talk a little bit about just your position for high band with flash. And just any thoughts around that? What does the SAM potentially look like for you guys there?

I'll let Tim should take that one.

Well, I think in any of these cases where you are talking about device architectures that require 3D scaling. I mean, obviously, our SAM opportunity just grows. I think these devices in the exact process flows and [indiscernible] still being worked through. But the types of systems we have, whether it's high aspect ratio conductor etches, higher-spec dialectric etches, the depositions ALD, it will be a great opportunity for us if it -- when it comes to fruition.

Operator, I think we have time for one more question.

Our next question comes from Vijay Rakesh with Mizuho.

Just a quick question on the DRAM side. It looks like it grew very nicely, up 45% year-on-year. On the -- when you look at HBM3E going to HBM4, with the higher layer count, I think, 50% higher. Is there a way to look at what your content uplift is per 100,000 wafers or something HBM3E goes to HBM4or 4E? And I have a quick follow-up.

Vijay, maybe I'll comment and then maybe let Tim talk about the technology. Yes, clearly, it goes up. We haven't given specific numbers around it. But obviously, the higher stack required.

I'm getting a little feed back. The higher stack requires more equipment, a little more challenging for the industry. So you clearly need more equipment. We haven't given a specific number on it in terms of dollar per 10-K.

All right. And just on the follow-up on HBF. I mean are you seeing book SanDisk and Hynix talking about it, I guess, but outside of that, when you look at high bandwidth flash, are you seeing investments or CapEx picking up there? Is that something you're seeing into '27? How would you look at that ramp?

Yes, I'd probably leave it to our customers to talk about their timing on these kinds of new technologies. But as I mentioned earlier, on any new technology, we're engaged with customers quite well ahead from a technology perspective of any production ramp. And then it's very much up to them ind. The one thing that's true, and we talked about it is that these are being driven by the growing importance of NAND as we see it within the AI memory hierarchy.

And so again, we think it's something that in a matter of time, this kind of capability is likely needed and [indiscernible] technologies that will support it very well.

This concludes our question-and-answer session. I would like to turn the conference back over to Doug Bettinger for any closing remarks.

Listen, I think Tim and I, and Ram would just like to thank everybody for your time and attention during what I know it's a super busy earnings season. I know we're going to see lots of you as the quarter unfolds at different conferences and road shows. So we're looking forward to that. And again, thank you for your interest in the company. We appreciate it.

This conference has now concluded. Thank you for attending today's presentation. You may now disconnect.