Lam Research [LRCX] Conference call transcript for 2023 q3
2023-10-18 21:56:04
Fiscal: 2024 q1
Operator: Good day and welcome to the Lam Research September 2023 Financial Conference call. All participants will be in listen-only mode. [Operator Instructions] After today's presentation, there will be an opportunity to ask questions. [Operator Instructions] Please note, today's event is being recorded. I would now like to turn the conference over to Tina Correia, Corporate Vice President, Chief Accounting Officer, and Investor Relations. Please go ahead.
Tina Correia: Thank you, and good afternoon, everyone. Welcome to the 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 September 2023 quarter and our outlook for the December 2023 quarter. The press release detailing our financial results was distributed a little after 1 o'clock PM Pacific Time this afternoon. The release can also be found on the Investor Relations section of the company's website, along with the presentation slides that accompany today's call. 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. Please see accompanying slides in the presentation 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 slides in the presentation. This call is scheduled to last until 3 o'clock PM 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.
Tim Archer: Thanks, Tina, and welcome, everyone. Lam produced solid results for the September quarter. Revenues came in above the midpoint of our guidance. And for the second quarter in a row, our gross margin, operating margin, and earnings per share, all exceeded the high end of the guidance range. Our revenue and earnings per share are expected to improve further in the December quarter, demonstrating our continued strong execution in a cyclically soft calendar year 2023. Turning to the wafer fabrication equipment environment, we see spending for calendar year 2023 in the $80 billion range. The adjustment in WFE from our prior view of mid $70 billion is based on updated checks on non-Lam related markets, as well as restricted fab spending in China. It does not change our assumptions on Lam revenues for the year. On the device segment side, NAND weakness continued in the quarter as customers adjusted spending levels down and further lowered utilizations to drive a faster path to supply-demand balance. While NAND WFE is down significantly in 2023, supply actions are starting to have a positive impact. Customers have recently indicated that pricing trends have stabilized and NAND bit demand has increased from high single digits per cent year-over-year growth to high teens, as certain consumer markets are demonstrating greater demand elasticity in per unit content. DRAM spending is modestly up relative to our prior view, driven by better trends in high bandwidth memory related demand, as well as further upside from domestic China customers. Meanwhile, the foundry/logic segment is down slightly versus our prior baseline due to weakness in both leading edge and non-China based mature node investments. Looking forward, it remains hard to call the timing and pace of WFE recovery, but we believe Lam is in a good position to benefit from both cyclical and structural drivers of demand. When memory investments begin to recover from current cyclical lows, we expect to see early benefits in our installed base business as fab utilization improves, driving increased demand for spares, services, and equipment upgrades. Longer term, Lam's growth story is strong and is underpinned by the fact that etch and deposition are fundamental enablers of higher performance, more scalable semiconductor device architectures. To address emerging technical challenges, customers continue to identify new innovative use cases for vertical scaling. Backside power delivery is a good example, as it is an emerging device architecture being developed to address the scaling limitations of traditional back end of line integration schemes. Etch and deposition play a critical role in enabling this transition, and backside power delivery is expected to add close to $1 billion of incremental SAM opportunity for Lam per 100,000 monthly wafer starts. Today, Power interconnects increasingly compete for space in the complex back end of line wiring, while also taking up considerable area at the transistor level. Additionally, managing power loss between the external source and the transistors is increasingly challenging due to resistance. A backside power delivery architecture enables the separation of the signal and power delivery paths to free up valuable wafer real estate and minimize power loss. Furthermore, customers are implementing changes, including the use of thicker metal layers in order to efficiently integrate backside power with their advanced packaging schemes. New etch and deposition capabilities are needed, and the trends are favorable for Lam. Due to our existing strength in back end processes, we've been able to quickly extend the capabilities of our copper electroplating and [PCVD] (ph) deposition products to address the throughput and productivity requirements of backside power applications. We now have two record positions at a leading foundry/logic customer and expect these positions to continue to grow. As we approach the end of the year, our installed base is closing in on 90,000 chambers. As semiconductor manufacturing is becoming increasingly complex, our customer support business group is seeing more opportunities to deliver innovation, productivity, and yield enhancement. In the September quarter, we expanded our equipment intelligence offering and multiple customers to include the first big data application of high resolution optical emission spectroscopy or OES. The equipment intelligence capabilities we are delivering with OES are highly differentiated due to the complexity of collecting and interpreting plasma spectra in manufacturing over time and across a large fleet of tools. Our solution allows customers to resolve performance issues that would otherwise remain undetected. Recently, our CSBG team also put the industry's first collaborative maintenance robot, or COBOT, into a production fab at a leading customer. COBOTs help execute complex maintenance tasks with precision and reliability, leading to improved tool-to-tool performance matching and higher equipment availability. Also, we believe COBOTs as a new service offering can play an important role in addressing anticipated skilled labor shortages as semiconductor manufacturing expands and becomes more regionalized. Overall, we see tremendous vectors of growth ahead for the semiconductor industry and for Lam. Scaling and complexity challenges are driving multiple inflections towards 3D architectures and in turn greater etch and deposition intensity. Lam has a strong track record of execution and we are committed to making the strategic investments needed to position the company to outperform as the industry and our markets grow. Over the last two years, we've been laying the groundwork for greater scale and efficiency with the expansion of our manufacturing, supply chain, and warehousing capabilities in Asia in order to better serve our customers in that region. We are also increasing our R&D efforts to extend our technology differentiation and expand our product portfolio to capture new inflection driven applications. While the current business environment remains challenging, secular industry trends play extremely well to land strengths, and we are excited by the breadth of opportunities we see ahead for the company. Thank you, and I'll now turn it over to Doug.
Doug Bettinger: Excellent. Thank you, Tim. Good afternoon, everyone, and thank you for joining the call today. We delivered strong results in the September 2023 quarter. Our revenue came in above the midpoint of our guided range, and gross margin, operating income, and earnings per share all exceeded the high end of guidance. We're pleased with the company's execution during a year where memory WFE investment has declined by unprecedented amounts. Let's look at the details of our September quarter financial results. Revenue for the September quarter was $3.48 billion, which was up 9% from the prior quarter and down more than 30% from a year ago. Our deferred revenue balance at quarter end was $1.69 billion, which was a decrease of approximately $150 million from the June quarter, mainly related to revenue recognized tied to customer advance payments. We continue to have a higher deferred revenue balance versus historic levels given these advanced payments. We expect to recognize revenue in the December quarter for a portion of these deposits, which is comprehended in our guidance. Within calendar year 2024, I believe the deferred revenue balance will trend to more normalized levels. Let's now look at the segments. From a segment perspective, September quarter systems revenue and memory was 38%, which is an increase from the prior quarter level of 27%. The growth in the memory segment was driven by DRAM, which increased, sequentially coming in at 23% of systems revenue compared with 9% that we saw in the June quarter. As we noted in prior quarters, non-volatile memory spending is at historic lows in 2023. And for the September quarter, the segment represented 15% of system for revenue, which was down from the 18% that we saw last quarter. The spending levels in NAND are at dollar levels we have not seen since [planar] (ph) NAND was the predominant technology. The foundry segment represented 36% of our systems revenue, lower than the percentage concentration in the June quarter of 47%. The decrease is related to timing of leading edge investments within calendar year 2023. We performed well in this segment during the year with this quarter's spending coming mainly from mature node customers. And finally, the logic and other segments was 26% of our systems revenue in the September quarter, which was flat with the prior quarter level. Investments in this segment were heavily focused in the specialty device areas, including sensors, analog, and power devices. I'll now discuss the regional composition of our total revenue. The China region came in at a high water mark of 48%, up from 26% in the prior quarter. The majority of the China revenue this quarter was from domestic Chinese customers. And we currently expect we will have another strong China geographic concentration profile in the December quarter as well. Our next largest geographic region concentration was Korea, had 16% of revenue in the September quarter and that compares with the 24% that we saw in June. Our customer support business group generated revenue in the September quarter, totaling approximately $1.4 billion, which was down 5% from the June quarter and 25% lower than the September quarter in calendar year 2022. Memory customers continue to operate their fabs at very low utilization rates. And customers are holding off on upgrading tools until there's more digestion of the outstanding inventory that is in the industry. The specialty technology market has been a bright spot this year, and we see that part of our business up year-over-year as we close calendar year 2023. Spares and the Reliant product line continues to be the two largest components of CSBG. Let me now turn to the gross margin performance. The September quarter came in at 47.9% above our guided range and higher than the June quarter level of 45.7%. Our strong gross margin performance compared to the prior quarter was driven primarily by favorable customer mix. We've improved elements of our cost structure during the year and are on track with our plan to improve gross margin from the March quarter level by approximately 1 percentage point as we exit calendar year 2023. September quarter operating expenses came in at $622 million, up from the prior quarter amount of $590 million. R&D as a percentage of spending was somewhat higher versus the June quarter, coming in at over 68% of our spending. The increased investment was focused on key technology inflections and development engagements with our customers. We will continue to invest in programs across multiple market segments to support our long-term strategic objectives for continued company outperformance. Operating margin for the current quarter was 30.1%, higher than the June quarter level of 27.3%, and more than 100 basis points over the high end of our guidance because of that strong gross margin performance. The non-GAAP tax rate for the quarter was 13.4%, in line with our expectations. Our estimate for the December 2023 quarter as well as for calendar year 2024 is for the tax rate to be in the low to mid-teens range. Other income and expense for the September quarter came in at $7 million in income compared with $7 million in expense in the June quarter. The favorable fluctuation in OI&E was due to a variety of factors, including rising interest rates, generating income on our cash balance. OI&E will continue to be subject to market-related fluctuations that will cause some level of volatility quarter by quarter. Let me now pivot to the capital return side of things. We allocated approximately $830 million to open market share repurchases and paid $230 million in dividends in the September quarter. I'll highlight that in September we announced a 16% growth in our dividend in line with our plan to deliver disciplined annual dividend growth. Since paying our first dividend in 2014, we have now raised the dividend amount 9 times. We returned over 120% of free cash flow in the quarter, and we have $2.7 billion remaining on our board authorized share repurchase plan. Calendar year to date, we've returned 83% of our free cash flow to shareholders. September quarter diluted earnings per share was $6.85 over the high end of our guided range. Diluted share count was 133 million shares, on track of our expectations and down from the June quarter. Let me pivot to the balance sheet. Our cash and short-term investments at the end of the September quarter totaled $5.2 billion, down from $5.6 billion in the June quarter. The main driver of the cash decrease was obviously our capital return activity. I’ll just mention, we also purchased buildings at our company headquarters as well as our Bay Area California factory for approximately $250 million, retiring the leases that were on the balance sheet. [indiscernible] cash was somewhat offset by improvement in days sales outstanding, which were 73 days in the September quarter down from the 80 days that we saw in the June quarter. Inventory turns were flat with the prior quarter level at 1.5 times. We continue to work to bring our inventory down, but as we've noted in prior quarter, we expect this to occur at a slower pace than we've done in the past. Our non-cash expenses for the September quarter included approximately $67 million for equity compensation, $76 million in depreciation, and $14 million in amortization. Capital expenditures for the September quarter came in at $77 million, which was flagged with the June quarter. Spending in September was primarily centered on product development activities and lab expansions in the United States and Asia. We ended the September quarter with approximately 17,200 regular full-time employees, which was a decrease of 200 people from the prior quarter. Most of this decrease is related to the restructuring actions we took earlier in the calendar year with the timing of the headcount reduction occurring in the September quarter. Let me now turn to our non-GAAP guidance for the December 2023 quarter. We're expecting revenue of $3.7 billion plus or minus $300 million. Gross margin of 47% plus or minus 1 percentage point. This level of gross margin reflects a continued favorable customer mix, albeit not quite as favorable as we saw in September. Operating margin of 29.5% plus or minus 1 percentage point. The operating expenses embedded in this guidance increase from the September level due to growth in R&D. I'd also just mention that the June 2024 quarter will be higher as it includes an extra week in fiscal quarter which occurs every few years. It's going to be a 14-week quarter in March. And finally, earnings per share of $7 plus or minus $0.75 based on a share count of approximately 132 million shares. With our December quarter guidance, we see solid performance in both revenue and profitability. Lam is delivering strong financial results and technology leadership to our customers as we develop solutions for the next industry inflections. And before I wrap up, I'd just like to mention two things as you think about modeling our business into 2024. The first is that we're currently experiencing favorable customer mix that may not continue at the same level going into next year. This may create near-term headwinds for gross market. Second, given all the opportunities we see in long-term technology inflections, like date all around, dry resist, advanced packaging, changing metalization schemes, and continuing the evolution of other 3D structures like DRAM. 2024 may be an R&D spending growth year to take advantage of these future opportunities that we see. As a result, it's possible the historic leverage we've delivered takes a temporary pause. We will obviously continue to aggressively drive the operational efficiencies that we always have, and our longer-term profitability objectives remain unchanged. Operator, that concludes our prepared remarks. Kim and I would now like to open up the call for questions.
Operator: Thank you. [Operator Instructions] And today's first question comes from Timothy Arcuri with UBS. Please go ahead.
Timothy Arcuri: Thanks a lot. I had a question on China. So obviously, you've had a huge second half in China. Doug, it sounds like you think it's going to remain pretty strong in December. It sounds like maybe it's going to remain close to 50% of the mix. But it sounds like you're a little worried about -- sorry, not worried, but you think that it could actually come off a bit during the first half of 2024. Can you talk about that? What are the puts and the takes? I ask because, if you sort of back into where they're running in terms of WFE, they're running probably in the high 20s if I just take the back half of the year. So I'm wondering if that can sustain.
Tim Archer: Yes, Tim, let me start on that and then let Doug add. I think that when we think about China investment, clearly it has been strong as we've messaged. Part of the mix story and why it's such a high percentage of our mix right now also has to do with the fact that other customers are not spending. We all know that memory and NAND in particular is at extreme lows. As we look into next year, and it's a bit too early for us to give 2024 WFE, so we're not going to do that. We think about longer term China and this overall move towards regionalization that you see people investing for long term demand in mature nodes. And so, we're not going to comment on whether we think it's sustainable in the first half or the second half of next year, but long-term, we do believe that there's growing demand in mature nodes that will drive China investment in a rather sustainable manner for the next several years.
Doug Bettinger: And yes, Tim, just to parse my comments a little bit. I said, we believe China will continue to be strong, but I also said, albeit not quite as strong perhaps as we saw in September. And as Tim alluded to, the China investment is not going away. I don't know if it's up, down, sideways next year, but it's not going away. They're investing for opportunities in the market that they see. I think we're hopeful that rest of the market begins to recover at some point, because it’s at pretty low points, and that will mitigate the China mix to a certain extent.
Timothy Arcuri: Thanks a lot for that. Can you just talk also, your major peer that makes litho talked this morning about there being a handful of fabs now with these new restrictions that they can't ship into. It's predominantly a litho thing, but can you just talk about sort of where we are in terms of restrictions? And then also, when you're talking about that what's the commonality in the etch and dep tool set for, let's say, 28 nanometer versus let's say 7 nanometer because technically you can buy tools for 28 nanometer and you can use them to pattern 7 nanometer from a dep/etch point of view. It's not as efficient but you can do it. Can you just talk about that? Thanks.
Tim Archer: Yeah, Tim, I guess what I'd say is, we've reviewed the details of the regulations and our early assessment is we don't see any material impact to our forecasted business. Now, some of that has to do with the fact that we've already been quite restricted into what we can ship into China relative to technology engagement. And I think to your other point about tools being purchased for one node and used for another. I mean, that's something that obviously we're -- we have to follow very strict regulations to adhere to the US regulations. So I don't know that that's something that might be quite as common as what you're saying. So, but it's something that we make sure that we're fully compliant.
Timothy Arcuri: Of course, Tim. Thank you so much.
Tim Archer: Thanks, Tim.
Operator: Thank you. And our next question today comes from Harlan Sur with JPMorgan. Please go ahead.
Harlan Sur: Good afternoon. Thanks for taking my question. CSBG down 8% year-over-year through the first nine months of the year. It did decline, as you mentioned, 5% sequentially. I assume the sequential decline was driven by continued utilization declines by your customers, especially your NAND customers, as you pointed out. You did talk about improving bit shipments for these customers in second half of the year, pricing stabilization, it sort of seems to be reflective of this sort of steadily improving supply demand environment. So does the team believe that utilizations have bottomed the cost of memory and foundry/logic customers? Have they started to at least stabilize at current levels?
Tim Archer: Yeah, Harlan. What we said was that, those are the comments that obviously our customers are out talking about their business. There's clearly some time lag from when they start to see improvement in bit demand and in pricing before they start to bring some of that fab utilization back online. Our CSBG business, as we said, is affected by a couple of things. One is, clearly fab utilization are at levels that we really haven't seen in terms of how much capacity has been taken offline in the NAND space. And that's affected spares. But also, when you don't need to add bits and you're really trying to conserve your own spending, there's been quite a significant delay in technology upgrades to the install base. You have those tools offline, you're not really upgrading them at this point. So that has also hit the CSBG business in terms of our upgrades component. We anticipate that as the memory business starts to improve, which is what it seems to be the leading indicators are planning to, we would think spares would start to come back and the technology upgrades would be done. Because in that next leg of growth, customers are going to want to be able to scale on that next technology node for their own efficiency of manufacturing. So we're not seeing that yet, but the leading signs are that it will come.
Harlan Sur: Thank you. And then you talked about this on the last earnings call, but you've got a strong position in [stat] (ph) chip architecture, advanced packaging. You guys have talked about this segment as being $200 million per year for Lam, potentially with the path to $1 billion business. On HBM specifically, where you have very strong share, right, HBM to HBM2 to HBM3 to HBM4, I think that there's a doubling of TSVs per chip, every new generation of HBM the DRAM stack height is also growing from like 8 to like 24 at some point. So this is -- this all is sort of a very strong tailwind for the team. Do you guys anticipate strong growth next year for your HBM/advanced packaging business? And some of your DRAM customers are also talking about HBM driving 10% of overall industry DRAM wafer starts. Is that how you guys are seeing it as well?
Tim Archer: We'll let them talk about how much it affects their business. But from what it affects ours, I mean, everything you just said, I think we're generally aligned with, which is, we are trying to drive performance generally leads to more equipment needs and more sophisticated equipment needs. And so, we do see HBM. I talked about the uptick in our business that we've already seen from it. This appears to be an area right now that is still under-supplied, and we're seeing strong demand. And so, I don't anticipate, given the interest in AI that you hear so broadly in the industry right now, that our HBM business would grow pretty strong in the next year. To characterize our overall advanced packaging, you mentioned a couple hundred million. We basically said that we think that this could be a market in which our revenues actually exceeded $1 billion over the next -- sometime in the next fiscal years. So it is a rapidly growing part of our business. And 1 in which we have quite high share.
Doug Bettinger: Thanks, Harlan.
Harlan Sur: Thank you.
Operator: Thank you. And our next question today comes from Atif Malik with Citi. Please go ahead.
Atif Malik: Hi. Thank you for taking my question. Tim, my first question is on equipment spending. I know you guys don't guide WFE until January, but your lithography peer was talking about maybe a softer first half and then things sticking up into 2025. And curious, if you're seeing a similar profile where first half is more in like a lamb and out like a lion.
Tim Archer: In like a lamb, out like a lion. You know, I think that it's -- what we would say and without getting 2024 at this point, I think it is reasonable to my prior comments that I think spending will come back cautiously. And so, even though we may be seeing some of the leading indicators in some of our markets, I think people are going to want to see sustainability of that condition before we start to see significant spending. So if we were to think, and I believe that the general feeling is that 2025 is probably -- a lot of fabs opening, a lot of perhaps demand out in that timeframe. It's not unreasonable to think that you would ramp towards that as you move through 2024.
Atif Malik: Thanks. And Doug, on the gross margins, you're still guiding to year-over-year growth but down sequentially because of the mix that you talked about. Can you walk us through which innings are we in terms of the structural gross margin improvement that you guys have been talking about this year?
Doug Bettinger: Yeah, if you take yourself back to when we were talking about this in the March call, you know, we're at roughly 44% gross margin and we articulated a view that we would be able to drive 100 basis point improvement in that from the operational efficiencies that we were undertaking during the year. And I'm super confident that that's absolutely already happened, frankly, and we will continue to be in a good spot as we exit the year.
Atif Malik: Thanks.
Operator: Thank you very much. Our next question today comes from Krish Sankar with TD Cowen. Please go ahead.
Krish Sankar: Hi, thanks for taking my question. I have two of them. First one, either Doug or Tim, I don't know if you guys quantified. In the past you said like, the export restrictions with the $2 billion impact on your revenue. Is there a way to quantify with the new export restrictions? And I noticed that ALE was added to it, a way to quantify the impact in calendar 2024 or incremental dollar value? And then I have a follow-up.
Doug Bettinger: Krish, there really was nothing material incrementally in what was clarified. I guess it was yesterday, right? So you're right. When we came into the year we said $2 billion to $2.5 billion. Now we've got clarification. We could ship a little bit. We took it on to $2 billion. That's still kind of what we see this year. And nothing incremental really came out yesterday.
Tim Archer: And the reason for that, Krish, is that, in many of those cases, while specific tools might have been now, or technologies now, called out, they were already the types of tools that were used to produce the technologies that were below the limits that were already allowed. So we had already recognized that in our initial statement.
Krish Sankar: Got it, got it. That's very helpful. And then now just to follow-up and to my -- I asked this question last time too, kind of like with respect to the whole cryo etch, the [indiscernible] etch versus [Tel] (ph), announcing a similar product, can you give us like an update on where you are market share wise on the [indiscernible] side? And is the cryo etch having any impact? Because my understanding is cryo etch does help the throughput, but it's probably a negative from selling number of tools for [indiscernible] etch NAND applications. I'm just trying to figure out how to handicap that.
Tim Archer: Yeah, I mean, there's no change from what we've said before. I mean, Lam is a leader in high aspect ratio etch without a doubt. And I think that also on the last call I did point out the fact that while a lot has been made of cryo or these very cold etching conditions. This is already standard Lam condition. And so to a certain extent, many of these when we talk about our business and any impact that you talk about with throughput, those are already factored into our commentary. So it's not something that's new to us. And I think that in general, our focus in NAND has been and always will be, and memory in general, driving productivity at every technology node. That goes into our projected growth outlook that we always talk about.
Operator: Thank you. And our next question today comes from Joe Moore at Morgan Stanley. Please go ahead.
Joe Moore: Great. Thank you. I wonder if you could talk about the DRAM uptick that you saw in the quarter. Can you kind of help us just understand qualitatively how much of that is coming from China? How much of that is from advanced packaging? Are you seeing kind of a resumption of technology spending? Just that kind of thing. What are you seeing that's driving that improvement?
Doug Bettinger: I guess, Joe, one I would point to, and I'll let Tim add. Yes, China was a part of it, but it's not all of it. There was an earlier question about, hey, where we are with bandwidth memory? Yes, it's getting pulled in. I mean customers want it sooner, we can ship. You've got a transition from DDR4 to DDR5 with these new CPUs that are out there. So that also is a little bit of a bright spot. So I guess I'd point to both of those things as part of what we saw in DRAM.
Tim Archer: Yes. No, I think that's pretty much it. I mean, as we've said before, you also have this transition to higher buy sizes, which ultimately will be a driver of wafer outs and therefore, equipment demand and probably starting to see some of the initial phases of that right now as well.
Joe Moore: Great. Thank you. And then in your upfront remarks, Tim, I think you had talked about trailing etch, mature node somewhat weaker outside of China. Can you -- is there something there that could be a trend? Or is that just kind of more of a one quarter phenomenon?
Tim Archer: Well, obviously, since we're not guiding future quarters, I mean, I don't -- I think it's just somewhat natural in general. I mean, we've seen in certain segments, very high spend rates as companies have looked to bring on capacity to meet demand in these mature nodes. And our industry goes through digestion phases where those tools have then started going up, and you get output and you sort of figure out whether demand is there, that requires more and it's what makes our investment cycle somewhat lumpy. And I think that's what we're really looking at right now. And if long-term demand, which we believe long-term demand for semiconductors is growing, eventually, you come back and put in more capacity into those fabs as well.
Joe Moore: Okay. Thank you.
Tim Archer: Thanks, Joe.
Operator: Thank you. And our next question today comes from Toshiya Hari with Goldman Sachs. Please go ahead.
Toshiya Hari: Hi, guys. Thank you so much for taking the question. I had a follow-up question to Joe's question on non-China trailing etch. And Tim, I apologize if I missed this, but can you point to any end markets or any device types that's driving the near-term weakness? And can you remind us how big or small of a market this is for you guys today?
Doug Bettinger: Toshi, I'll take it and then let Tim add on. It's a broad set of customers investing when you look at this, the specialty nodes. I mean, it's across multiple customers unlike like leading etch foundry as an example. And so I can't really point to any one or another for you. And quite frankly, I think we all know there's inventory out there in a lot of these device types, but these are long-term investments as well, right? This isn't something that comes in one quarter and then goes away just because of what's happening in the near-term marketplace. But it is in any one segment or another, Toshi, just kind of the broad set of customers.
Tim Archer: Yes. And Toshi, the only thing I would add is, I think this is going to be an area that I think we're going to have to sort of acceptably a little harder to forecast from the standpoint. It's also a part of the market that is impacted by a number of the different chip stack type government support activities around the world. And so, you may see as certain regions try to build up their capabilities we may not be able to point in that moment to the demand being greater than the supply, and that's why they're investing. As Doug said, this is about long-term build-out what I think everybody sees is a much bigger demand for these types of devices over all these various types of applications in automotive and IoT and CMOS image sensors, et cetera, et cetera, over time.
Toshiya Hari: Got it. I appreciate the color. And then as my follow-up, one for Doug, on the 2024 model you talked about mix normalizing and you guys potentially experiencing some headwinds in gross margin. You also talked about 2024 potentially being a growth year from an R&D spending perspective. So I guess on gross margins, I guess what's the baseline that we should be working off of? Is the Q4 rate a good starting point? Or is that still kind of high given where China is. And then from an R&D spending perspective, I guess, relative to answer if you can kind of hold our hand and quantify how low leverage could be in 2024 versus history that would be super helpful.
Doug Bettinger: All right, Toshi, I'll give you a couple of data points, maybe for a little bit of color, but I'm not going to guide next year. Yes, we're still at an elevated level of gross margin from customer mix relative to where I think things normalize. Maybe a good way to think of it is go back to the June quarter, which was before we saw a lot of this China favorable mix. That's not an unreasonable baseline to start from for gross margin. So anyway, I don't know if that's helpful. And then I guess what I described from an R&D standpoint. R&D, quite frankly, has to follow a cadence independent of the level of revenue sometimes. I don't know what WFE next year is going to be. I don't know what our top line is going to be quite yet. We'll give you some color next quarter. But I do know we see an enormous number of opportunities around these technology inflections that if we don't invest right now three, four years from now, we'll look back and say, why didn't we, right? I'd be all around. Tim talked about backside power. There's high bandwidth, there's so many things that play to the strength of what we do well. So I think if you look at the December quarter and compared to September, spending up, the March quarter independent of the fact that we're going to invest more in R&D is a 14-week quarter. So you got to comprehend that. And then we're going to grow R&D as we go into next year, maybe a little bit independent of what revenue turns out to be. I guess what I would want you to think about is, historically, when business grows at Lam, you've seen nice leverage in the model. It's maybe going to flatten out a little bit, honestly, is how I'm trying to like, come to you to think about it a little bit. And again, that's because we see a lot of opportunities that we think played at the strength and is going to set us up to win in the longer term. I don't know if that helps, Toshiya.
Toshiya Hari: Yes, it makes sense. I appreciate the color, Doug.
Doug Bettinger: Okay. Thank you.
Operator: Our next question today comes from Vivek Arya with Bank of America Securities. Please go ahead.
Vivek Arya: Thanks for taking my question. For the first one, I'm trying to understand the usual delta between the recovery in your CSBG business and your memory systems. So how effective of a leading indicator is CSPG recovery? And what is it telling you right now about conceptually when memory system orders recover? Like is it in Q1, Q2, Q3, of next year, like what has been that delta historically? And what is that telling you right now about when your memory system orders can recover?
Tim Archer: Yes. I mean we've -- first, I'd point out the fact that it's been a long time if ever that we've seen fab utilization is quite this low. I mean our prior commentary had been that spares generally grows every year because the installed base continues to get bigger. One piece of that is true is the installed base continues to get a lot bigger. And in fact, we've said it's up more than 40% since the last cyclical downturn. And so the fact that we've seen spares and upgrades and all of these things sort of be off all at one time is pretty unique. We are anticipating that as fab utilization starts back, we'll see spares come back. And I think the one thing that will come back, and it's a little hard to predict is certain is the technology upgrade portion of this. It's been now a couple of years since any of these tools have been upgraded, and that will have to happen. In terms of your comment about offset, I think we'll -- typically, as customers start to turn the tools back on, you're probably a couple of quarters away from seeing further investments in the technology upgrades. And then I think when you're really talking about capacity adds, it's hard to predict what that time frame is because it really depends on many other factors about our customers' use of long-term demand. So I think the one thing we're certain is that we're at very low points now. We would anticipate things like utilization and spares and upgrades to start improving next year. And beyond that, I think we'll wait until January to give you a better view.
Vivek Arya: All right. And for my follow-up, another China-related question. Part of your second half strength came, I believe, from clarification of some rules. And I was hoping you would quantify how much of that strength you saw in your shipments came from just that clarification? And does that kind of spill over to early 2024. I'm just trying to tease apart how much is sort of sustainable China strength versus how much is potentially from one-off clarification in rules? Or maybe those were not one-off, right? Maybe they're also sustainable. So just if you could give us some way to guide us to how we should think about China conceptually in your first half of next year.
Doug Bettinger: Yes, Vivek, we're not going to guide you next year quite yet. But the clarification of the rules we described isn't changing, right? So we understood one node that a certain customer was doing was okay to ship to. The rules didn't change. We just had to do a little work to understand that. Collectively, as an industry, that's not going to go away. Having said that also, what our commentary on the call so far has been, I don't know if China is up now on our sideways next tier, but it's not going away. When we talk to our customers in China, they all communicate roadmaps that have multiyear horizons in front of them. Nothing new came from the regulations that you saw yesterday. So I see a level of sustainability in China as we go into next year and frankly, beyond. They have long-term objectives.
Vivek Arya: Okay. Thank you.
Operator: Thank you. And our next question today comes from Stacy Rasgon with Bernstein Research. Please go ahead.
Stacy Rasgon: HI, guys. Thank for taking my question. For my first question, you talked about in your WFE uptick on non-AM markets driving some of that. Am I oversimplifying? Are you just talking about litho or do you have something else in mind when you made that statement?
Doug Bettinger: Stacy, it's primarily litho. It's litho and these restricted fabs in China that we didn't have complete visibility into what they were doing, frankly. It was those two.
Stacy Rasgon: Got it. You have better visibility now?
Doug Bettinger: I think we do. That's why we updated the number. We never get this exactly right, but we try to tell you what we think of it now.
Tim Archer: Stacy, some of our visibility comes from the fact that our peer companies are reporting on the business and it’s really their markets. So as the year goes on, we try to give you a view of the whole market, but obviously, we're most accurate on the Lam business.
Stacy Rasgon: Got it. Got it. But to be clear, I think you said there's no change to your forecast for the Lam business for the year, whatever that forecast is.
Tim Archer: That's right. Because we understand that WFE quite well.
Stacy Rasgon: Got it. Got it. And so for my follow-up, I want to go back to the leverage question for next year. So I understand you're talking about like leverage, like maybe flattening out. Was that just a statement you just think OpEx, I mean, just to put on table OpEx growing with revenue, whatever revenue is? Or given the gross margin compression that we'll probably see at least from the current level, like do you think operating margins year-over-year could actually go -- could decline next calendar year? Or like just how do we think about those different pieces?
Doug Bettinger: Stacy, if you look at what Lam has done over the last, I don't know, a decade, frankly. We've expanded margin as revenue has grown. At this point, I'm not sure what revenue is going to be next year, but I know we're going to invest more in R&D. That's what I'm trying to describe because we see all of these opportunities. And yes, I referred to the fact that we've got pretty favorable customer mix that likely mitigates somewhat next year. And so when you think about those two things, it's possible to think about kind of margin flattening off for a period of time. Our long-term profit objectives are unchanged. So I do want to reiterate that point as well.
Stacy Rasgon: Got it. That’s helpful. Thank you guys.
Tim Archer: Thanks, Stacy.
Operator: Thank you. And our next question comes from Srini Pajjuri with Raymond James. Please go ahead.
Srinivas Pajjuri: Thank you. I have a couple of longer-term questions, Tim. I guess if I look at the last five years, your logic and foundry business has been growing almost at a 30% rate. And historically, your memory was close to 60%. And now I think we're at the bottom kind of moving from 27% to 38% or so this quarter. I'm just curious, how do you think about the mix longer term, I guess, when things normalize for you? And what is, I think, in your view, what do you think is the ideal mix for you? And what implications, if any, that might have on your top line growth going forward?
Tim Archer: Yes. It's a good question. It's a hard question to answer because the actual. Our view is we want more of everything. So we're not looking to reduce our position in memory just to make the mix look better. So we try hard every day. We have a fantastic position in memory, and we think there's still more to come there as over the next decade, I mean, NAND is going to scale customers saying to 1,000 layers, and that's a tremendous opportunity for Lam. DRAM going to 3D around the end of the decade, tremendous opportunity for Lam. So I'm afraid the memory side will keep growing simply because it's so well suited to our strengths. But you have heard us talk a lot about the fact that we see huge opportunity in the foundry logic side as well. And when Doug just talked about spending, I mean, I've tried to lay out for you in the last several quarters, the breadth of opportunities that are ahead of the company, many of which -- not all, but many of which are on the foundry logic side. I mean just to kind of recap some of those, I mean, the dry EUV photoresist and develop. When we said that's $1.5 billion opportunity over five years. And when you get towards the tail end of that five years, I mean, that's a revenue that's growing with the number of expanding EUV layers at every technology node after that. That's primarily foundry logic business. We talked about gate all around, about $1 billion incremental opportunity for Lam introduces opportunities to win new tools and selective etch and ALD. And so that's same expansion for us in foundry logic and opportunity to grow. We talked about advanced packaging. I mean you've already seen what's happened with not only the HBM side of AI, but also the entire formation of these big AI systems using interposers. Lam's invested now in panel processing as a way to ultimately bring down the costs of some of this chiplet -- these chiplet applications. And then finally, today, I just talked about backside power distribution as a new way of being very creative about how you use that backside of the wafer is additional real estate and it opens up a lot of new opportunities for us, and that's primarily a foundry logic application as well. And so really, what we're talking about is Lam has a long way to go to expand our SAM, especially on the foundry logic side. That's what we're investing for. And each of these is a $1 billion-plus opportunity for Lam over the next several years. And so we're pretty excited about that. But we're not giving up on our strong memory positions.
Srinivas Pajjuri: Thank you. I appreciate that answer. And then my quick follow-up on the HBM, I guess, technology itself. A lot of questions have been asked already. Just on the capital intensity of HBM. Tim, I mean, I know you said the die size is larger and I guess, cycle times are longer, et cetera. But is there a way to think about capital intensity per wafer or per bit? How we should think about HBM versus traditional DDR?
Tim Archer: Well, I think it's a -- I don't know that we've quantified that number, but it's -- it obviously is a much higher performance die than our device, and it does -- it is bigger and takes more capital. And so therefore, it's a performance-driven application. From our perspective, though, what really is interesting is that, many of the new tools that get added to enable HBM or Lam tools or tools that are in our market, things like silicon etch and copper plating for the TSV formation. And so that's what really makes it an even better transition for [indiscernible] company like Lam.
Srinivas Pajjuri: Got it. Thanks, Tim.
Tim Archer: Thanks, Sri.
Operator: And our next question today comes from Brian Chin with Stifel. Please go ahead.
Brian Chin: Good afternoon. Thanks for letting us ask a few questions. So about a week ago, the U.S. relaxed its licensing requirements for some foreign companies that operate more advanced fabs in China. I know it's fairly recent, but have you seen a positive impact yet from this change in the licensing policy. I did note that in talking about China remaining a good concentration in the December quarter. It sounds like maybe there could be some shifting there between local and maybe foreign domiciled companies?
Tim Archer: Yes. It's pretty -- that comment is pretty recent. But I think relative to multinationals, wherever they operate, whether China or certainty of being able to make the investment and benefit from that long term is very important. So obviously, in the last couple of weeks, we haven't seen any movement that we had talked about. I think long term, it allows people -- it allows our customers to ultimately make the right decisions for them about where to invest. And that is especially true when you think about our installed base and the upgrades to the installed base and a customer willingness to sort of move forward with those upgrades with certainties.
Brian Chin: Got it. That's helpful. And then a lot of questions have been asked about sort of service spares and utilization improvements. But just curious, if memory companies are kind of talking as is they'll realize some of that utilization improvement, not just through increasing wafer starts, but also through some reduction in wafer start capacity as they emphasize newer nodes or capital efficiencies on some wafer loss there. How are you thinking about how that impacts maybe the trajectory of your service and spares revenue growth in calendar 2024?
Tim Archer: Well, I think that we have to see and we -- as we said about the -- it's difficult to predict the pace of the recovery. But what happens there is we're just trading off 1 part of our business that CSBG business for another. If upgrades happened before spares, you're right, it brings utilization down. But there's also something that we've said, which is as technology moves forward, many of those applications become more spares intensive because the processes are longer and more demanding. And so I would just say that we're going to see a rise in both parts of our business as fab operations recover and customers start to fully utilize the equipment in those fabs.
Operator: Thank you. And our next question today comes from Mehdi Hosseini with [SIG] (ph). Please go ahead.
Unidentified Analyst: Yes. Thanks for taking my question. Just two quick follow-ups. Opportunities with the back side power rate. Should I think about it more of a 3-nanometer or extension of 3-nanometer? Or are those opportunities materializing when we migrate to 2-nanometer. And I have a follow-up.
Tim Archer: Yes. I think that it's still a little bit out in the future, I mean, you could go look at our customers' road maps and what they've said publicly, and I think that's probably the best way to put timing on it. I think what we're trying to highlight is the fact that our -- if you think about the challenges across almost every device, we're becoming more and more convinced that the technology solutions to those challenges involve vertical scaling, they move to 3D and you're seeing that backside power, advanced packaging, gate all around. These are -- it is something that is playing extremely well to our strengths in etch and deposition. And so timing is hard to predict, but the certain -- relative certainty of those changes happening is quite high.
Unidentified Analyst: Got it. And a quick follow-up for Doug. What should we think of a normalized or normal deferred revenue level?
Doug Bettinger: Yes. Mehdi, in the past I've suggested maybe something around $1 billion is a normalized level and we're somewhat elevated from that. the longer we remain elevated, maybe the normalized level picks up a little bit, but I'll leave that statement, what I said, maybe roughly $1 billion.
Unidentified Analyst: Okay. Thank you.
Operator: Thank you. And our next question today comes from [Chris Caso] (ph) with Wolfe Research.
Unidentified Analyst: Yes. Thank you. Good evening. This question is about where your lead times delivery times are now and generally your ability to react when demand ultimately returns. You made some comments before that your customers you think would need to see some significant improvement before they started spending. Does this mean they have some -- a little bit of luxury of time to see things get a bit better before they start calling you and increasing orders.
Tim Archer: Yes. I think that if we were to go back and talk about lead times compared to pre-COVID, I would say that generally still extended, but for a variety of reasons. I mean I talked about the investments we've made in our supply chain and our manufacturing facilities. A lot of that has to be more responsive, but I think we will be able to respond when realistically when demand starts to come back. Now in certain areas, it's a little tighter than others. And I think that's where, again, we still continue to work very closely with our customers to make sure we have good forecast as we talked about. High bandwidth memory has been an area that I think has been in tighter supply. And the good thing about the investments we made in our global operations is that in certain cases, we're able to respond a little bit more quickly than customers have urgent needs.
Unidentified Analyst: Got it. Thank you. As a follow-up. If you give a little more clarity on the extra week you're expecting, I think it's in the March quarter, both on -- do you expect a revenue impact from that extra week? And what do you expect the cost impact to be?
Doug Bettinger: Yes. Usually, you don't really see much of a revenue impact, frankly. You kind of manage based on what customer wants when. But I know for sure, with an extra week, 14 weeks instead of 13, you got more salary expense, you've got more time to use project materials and whatnot. I don't know. I didn't give you a specific number, but I think it's pretty well chronicled about there. When people have a quarter coming in like this, how much spending grows just because of it. I'm not going to put a number on it, but just think about 14 versus 13.
Unidentified Analyst: Got it. We can do the math. Thank you.
Tina Correia: Operator, we have time for one more question, please.
Operator: Yes. Our final question will be from Sidney Ho with Deutsche Bank. Please go ahead.
Sidney Ho: Thanks for squeezing me in. In the past, you guys talked about memory spending recovery will go in several phases from increasing utilization to tech upgrades and then capacity expansion. One of the memory customers talk about converting some of this excess capacity to address the advanced note. Are you seeing that dynamics happening across other memory suppliers? And how does that change your view in terms of the timing of capacity expansion?
Doug Bettinger: Sidney, you were breaking up a little bit. I think you were asking about the cadence of when utilization comes back, what we think will happen, I think that was your question. And I think what I would tell you, first, spares comes back. Second, you'll see upgrades, and upgrades will have to happen obviously, some of the customer base has taken some things offline. So there's a spend that needs to occur to get that back online and up to speed and then eventually new equipment gets purchased. That hasn't changed.
Sidney Ho: Okay. So my question was more about some of your customers actually trying to convert some of the excess capacity to address the advanced notes, meaning that it seems like the timeline is compressed. Are you seeing all the memory suppliers doing that?
Doug Bettinger: We don't talk about any one customer or another, but that will show up in that upgrades commentary that I was talking about, the stuff that gets taken offline eventually needs to get upgraded. That's the upgrade spend in CSBG.
Sidney Ho: Okay. That's it. Maybe last question for you. Given your position in gate-all-around, has your timeline changed at all in terms of when do you expect to see gate-all-around related revenue as compared to three months ago? Maybe just remind us when that is going to happen? And what are the leading indicators you watching to gauge that timeline. Thanks.
Tim Archer: Well, yes, it's -- I don't know that our timing on gate-all-around around has changed much from three months ago. I think that, again, it's a means of scaling device performance and device performance is something is important to our customers and their [indiscernible]. So we're just engaged with customers to make sure our tools get qualified into those new nodes and when they decide to ramp them, then we'll be ready as well.
Sidney Ho: Okay. Thank you.
Tim Archer: Thanks, Sidney.
Operator: Thank you. And this concludes our question-and-answer session. I'd like to turn it back over to the management team for closing remarks.
Tina Correia: Thank you, operator, and we appreciate everyone for joining. Thank you for your time today.
Operator: Thank you. This concludes today's conference call. We thank you all for attending today's presentation. You may now disconnect your lines, and have a wonderful day.
