Ladies and gentlemen, please stand by. We're about to begin. Good day, welcome to the June 2019 quarter financial call for Lam Research. Today's conference is being recorded. At this time, I'd like to turn the conference over to Ms. Tina Correia, Corporate Vice President of Investor Relations. Please go ahead, ma'am.
Great. Thank you. 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 review our financial results for the June 2019 quarter and our outlook for the September 2019 quarter. The press release detailing our financial results was distributed a little after 1:00 P.M. 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 includes forward-looking statements that are subject to risks and uncertainties reflected in the risk factors disclosures of 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 today's earnings press release. This call is scheduled to last until 3:00 P.M. Pacific Time. A replay of this call will be available later this afternoon on our website. With that, let me hand the call over to Tim.
Thanks, Tina. Good afternoon to everyone on the call. In the June quarter, Lam delivered strong results. Revenues, gross margin, and operating margin exceeded the midpoint of the guidance we provided on our last earnings call, while EPS exceeded the high end of the guidance. Our ability to deliver these results was made possible by the support of our customers, our employees, and our partners. I would like to thank them for their ongoing commitment to Lam's success. I would like to start by offering some perspective on Lam, given the uncertainties surrounding trade and other influences on the market environment. Through the first half of calendar 2019, we have executed at a high level by focusing on what is in our control. Despite a difficult memory market, we have delivered on or exceeded our commitments.
As you will hear from Doug later, we generated another $880 million in cash from operations in the most recent quarter. At the same time, we have continued to prioritize investment in innovation and product differentiation, shifting a higher percentage of our total OpEx to R&D than in any prior quarter in our history. We remain committed to our long-term growth vectors of served available market expansion, market share gains, and increased revenue from our installed base. Specific proof points are emerging that validate our progress in each of these areas, and I will touch on them later in my remarks. First, let me update our industry outlook. Our view on total WFE remains directionally unchanged, with calendar year 2019 down mid to high teens percent from 2018.
Since our last earnings call, memory spending is incrementally lower, while foundry spending is tracking higher, we believe due in part to the acceleration in 5G development, leading to strong demand for 7 nm and 5 nm products. We now view foundry logic WFE to be second half-weighted versus our prior baseline of first half-weighted for calendar 2019. Within memory, customers have continued to take meaningful actions to restore supply and demand balance, reducing investment and lowering utilization levels for both NAND and DRAM as we progress through the June quarter. As a result, we see year-over-year bit supply growth for both NAND and DRAM continuing to decline, with bit supply growth rates exiting the year well below the long-term demand trend lines. On the demand side, we are encouraged by early signs that NAND price declines are leading to an acceleration of content growth in devices.
For example, SSD penetration in PCs is expected to reach nearly 60% by the end of 2019. More importantly, the average density per drive is also growing. The predominant SSD configurations in the PC market are now 256 GB and 512 GB versus 128 GB and 256 GB a year ago. The demand impact is amplified as both units and content per unit grow together. Similarly, we have seen acceleration in bit content growth for NAND in mobile devices. This is evident not only in the premium phone segment, but also in the mid-tier and lower-end phones, where 128 GB offerings are seeing the fastest growth versus 64 GB models a year ago. We believe the combination of factors influencing supply and demand creates a favorable setup for memory as we enter 2020.
Our actions are focused on ensuring Lam is in the best possible position to benefit from the anticipated recovery in memory spending. This focus is contributing to gains in both served market and market share, as evidenced by wins this quarter across various applications as well as in roadmaps for emerging 3D architectures. In NAND, we are working closely with customers to develop differentiated solutions for potential limiters to 3D scaling. An excellent example is the wafer processing challenge created by stress-induced wafer bow as the number of 3D NAND layers scales to 100 and beyond. Given Lam's leadership position in critical 3D NAND etch and deposition applications, we are best positioned to address this problem. A new Lam tool, which deposits a counter-stress film on the backside of the 3D NAND wafer using an innovative single-step process, has been introduced to leading-edge customers with multiple repeat orders received.
Importantly, the learning we gain through our investments to scale 3D NAND will be broadly applicable as new 3D architectures emerge in other segments. For example, we believe Lam's leadership in enabling 3D scaling will become increasingly valuable as logic devices migrate to a 3D gate-all-around structure at three nanometer, as new memories such as PCRAM scale vertically for cost and bit density improvement, and as 3D heterogeneous integration is adopted as a preferred packaging option for high-performance system solutions. For 3D heterogeneous integration, several leading companies across foundry and logic have announced architectures to connect different IP blocks using high-density interconnects with through-silicon vias. Lam's SABRE 3D electroplating and Syndion etch tools offer best-in-class technology, backed by years of high-volume production leadership in the TSV market.
During the quarter, we secured an important win at a leading logic customer for our SABRE 3D electroplating system for 3D chip stacking applications. This is a significant validation of Lam's ability to leverage its industry-leading position in 3D scaling to new and emerging manufacturing inflections. We're also seeing successes from our customer-focused investments in DRAM. We are collaborating with customers on critical new technologies required to scale to the 1Z and 1A nodes. For instance, as DRAM devices shrink to 1Z and beyond, the performance impact of resistance capacitance delays becomes a challenge that must be addressed. This quarter, we won critical spacer applications at multiple leading DRAM manufacturers for the 1Z node due to our ability to deposit highly conformal low-k films that help reduce RC delay.
In addition to our progress on critical applications, you might recall that I have spoken previously about the NAND semi-critical process space that is an area of increased focus for Lam as we drive to gain market share. In the most recent quarter, we began to deliver on this opportunity with significant conductor etch wins at multiple NAND customers for mask open applications. Specifically, Lam's ability to create a highly productive single-step etch process to replace our competitor's multiple-step approach allowed us to differentiate on system throughput, a key decision factor in the semi-critical space. Another example was seen in dielectric etch, where we recorded a key win for a semi-critical metal contact application at a leading memory maker. As 3D NAND scales, the peripheral contacts become deeper and aspect ratios increase.
Our dielectric etch system demonstrated faster etch rates and superior profile control, leading to greater capital productivity and less frequent maintenance. For our customer support business group, we continue to see 2019 as another year of solid revenue growth despite lower WFE spending. In the June quarter, we achieved a second consecutive quarterly record for revenue from our Reliant systems business as customers invest to address robust non-leading-edge demand from end markets such as IoT, automotive, and power devices. The combination of our focus on leading-edge technologies and installed base performance is being recognized by our customers. In the most recent quarter, a top customer completed their annual supplier evaluation process, ranking us as their number 1 supplier as measured by a broad set of install base performance, cost reduction, and R&D engagement metrics.
With this newest rating, we are now in the number one position at more than half of our top customers. To wrap up, the near-term environment remains challenging. The long-term growth opportunity for both our industry and for Lam is compelling. We are focused on executing to our commitments and extending the differentiation of our product and services portfolio. I believe that our continued prioritization of customer-focused investment will yield lasting benefits. As memory spending returns to normalized levels and non-memory technologies increasingly rely on 3D scaling for performance and cost improvement, Lam will be in an excellent position to outperform. Now I'd like to turn the call over to Doug.
Excellent. Thank you, Tim. Good afternoon, everyone, and thank you for joining us today during what I know is a busy earnings season. Lam executed well in the June quarter, with our results exceeding the midpoint of guidance for all financial metrics. Earnings per share exceeded the high end of the guidance range we provided due to stronger gross margin as well as proactive management of operating expenses. Our EPS performance is, I believe, a testament to our continual focus on delivering on our ongoing commitments. As Tim noted, our expectation is that 2019 WFE will be down from calendar year 2018 in the mid to high teens percentage level. Since our last call, we're seeing some upside strength in foundry, which was partially offset by a reduction in memory spending.
From a segment perspective, our systems revenue for the combined memory segment increased slightly to 64% of total system revenue from 61% in the March quarter. We had an increase from the March quarter in the non-volatile memory segment from 40% - 46%, while DRAM decreased to 18% from 21%. Memory revenue continues to be mainly targeted towards conversion-related investments. In DRAM spending, it's targeted towards 1Y and some initial 1Z investment. In NAND, it's primarily conversions to 9x layer devices. Additionally, we are seeing initial investment on 128-layer structures. We're continuing to see healthy spending in the foundry segment, which came in at 23% of our system revenue for the March quarter. As Tim mentioned, this is heavily focused on the 5 nm and 7 nm nodes. This was slightly down from 27% last quarter, although still quite strong.
The logic and other segment was flat with the prior quarter level, contributing 13% of system revenue. We expect to see continued strength in the foundry and logic space throughout the remainder of the calendar year. It's worth noting that from a geographic perspective, 33% of our revenue was generated in the China region. The majority of this came from indigenous Chinese customers. We expect to see higher than our average concentration level of revenue in the China region for the September quarter as well. We executed well on income statement performance for the June quarter. Revenues came in at $2,361,000,000, which was above the midpoint of the June guidance. Gross margin for the quarter was 45.9%, which was better than expected, primarily due to customer mix and improved field resource utilization.
As we've stated in previous quarters, our actual gross margins are a function of several factors, such as overall business volumes, product mix, and customer concentration, and you should expect to see some variability quarter to quarter. Operating expenses in the June quarter declined to $450 million from the prior quarter. We are proactively managing expenses with our lower revenue levels. We continue to invest in our strategic R&D programs, however, and remain focused on our commitment to technology and productivity leadership. The percentage of R&D spend increased to approximately 66% in the June quarter, which was a high-water mark. Operating income in the June quarter was $635 million, and operating margin was 26.9% at the high end of our guidance range. The non-GAAP tax rate for the June quarter was approximately 11%, which is slightly lower than our long-term rate.
For the remainder of the 2019 calendar year, we expect a tax rate in the low to mid-teens. I'll just remind you should expect to see fluctuations in the rate from quarter to quarter. For June, other income and expense was a total of approximately $6 million of expense. This total includes interest expense for a full quarter related to the issuance of a $2.5 billion senior notes that we completed in the March quarter. The quarterly interest expense is partially offset by the interest income earned on higher cash balances for the company. I'll just remind you that total interest expense on all tranches of our debt is approximately $45 million per quarter. We continued to execute on our capital return program during the June quarter.
For the quarter, we allocated $1.3 billion to capital return, with $1.1 billion coming in share repurchases and $165 million in dividend payments. Our share repurchase activities were from a combination of open market as well as structured repurchases. The structured repurchase program that we entered into is intended to continue to execute throughout our December quarter. We've completed approximately $2 billion of the $5 billion authorization that we announced in the March quarter. Over the past year and a half, we've utilized approximately $6 billion in buybacks and lowered diluted share count by roughly 15%. I believe this demonstrates our continued commitment to return meaningful cash to our shareholders. Earnings per share was $3.62, which was over our guidance range for the June quarter. This upside was driven primarily by better gross margin and lower-than-expected Operating Expenses.
Diluted shares for EPS were approximately 154 million shares, which reflects a 5% decrease in quarterly diluted share count since the beginning of the calendar year. The share count includes a dilutive impact of approximately five million shares from the 2041 convertible notes. The dilution schedules for the remaining 2041 converts is available on our investor relations website for your reference. Let me now move to the balance sheet. Our cash and short-term investments, including restricted cash, decreased in the June quarter to $5.7 billion from $6.4 billion in the March quarter. The decrease is mainly due to the capital return activities within the quarter, offset by strong cash generation from operations of $880 million. This is the second consecutive quarter where we had cash from operations in the $900 million range. DSO decreased by five days to 56 days. Our inventory balance decreased by $82 million.
Inventory turns remained at industry-leading levels, coming in at 3.3 x. Company non-cash expenses included approximately $45 million for equity comp, $47 million for depreciation, and $18 million for amortization. Amortization was down from last quarter by 50%, as a portion of the intangibles from the Novellus acquisition have now fully amortized. Capital expenditures were $66 million in the quarter, which was a slight decrease from the $76 million that we saw in March. The June quarter ended with approximately 10,700 regular full-time employees, which is down somewhat from the prior quarter. Additionally, I'd like to remind you that we use temporary labor as part of our operating model. We have reduced this temporary labor by almost 40%, or 600 head count, in the last year. This flexibility is a critical part of our operating model, enabling us to deliver sustainable operating profits during a time of reduced revenue levels.
Now looking ahead, I'd like to provide our non-GAAP guidance for the September 2019 quarter. We are expecting revenue of $2 billion, $150 million, ±$150 million. Gross margin of 45%, ±1 percentage point. Operating margin of 24.5%, ±1 percentage point. Finally, earnings per share of $3, ±$0.20, based on a share count of approximately 150 million shares. Before closing my scripted remarks, I'd like to reiterate some of the tone that Tim shared in his script. We continue to not see a recovery in memory spending, which is our strongest market, for 2019. We do observe dynamics in those markets, however, that are positive signs. Some examples of these signs are demand elasticity, pricing trends, and management of factory utilization to bring inventory down.
We are tracking a double-digit number of new fabs ready to receive equipment shipments during this year, and we see plans for that to happen again next year. We continue to believe, as a result, that 2020 sets up as a better year than 2019. Operator, that concludes my prepared remarks. Tim and I would now like to open up the call for questions.
Thank you, sir. Ladies and gentlemen, if you'd like to ask a question, please signal by pressing star one on your telephone keypad. If you're using a speakerphone, please make sure your mute function is turned off to allow your signal to reach our equipment. Again, that is star one at this time. We'll take our first question from Harlan Sur with JP Morgan.
Good afternoon. Thanks for taking my question. A quarter ago, the supply side situation in memory was one of disciplined spending and sort of reining in supply. It was really the demand side that was still uncertain, but just even over the past few weeks, it seems that the demand side is starting to materialize here in the second half of this year. PC market looking seasonally stronger, cloud spending set to re-accelerate this quarter, and you even have several big sort of AI and deep learning programs that are starting to fire. Your customers are also talking about inventory starting to come down. Is the Lam team feeling more confident, and more importantly, are your customers feeling more confident about the prospects for a healthier market environment exiting this year relative to three months or six months ago?
Sure, Harlan. I'll take that, and then Doug can add what he would like to. Clearly, maybe I walk you back since you kind of set us as a baseline, three or six months ago. The very first call of this year, we laid out a view that memory spending really wouldn't recover for this entire year. You just heard Doug kind of reiterate that again. That didn't mean that through the year, we wouldn't see progress. Progress in sentiment, progress in both the supply side and maybe the demand side. I guess what I'd say is, incrementally, you're hearing commentary about the demand side. I talked about elasticity. You've heard that also from some others even closer to those markets than us. I think that you're starting to see some sentiment in NAND that are positive signs.
Back six months ago, we also said that given the timing in which NAND corrected versus DRAM, that NAND corrected earlier, and therefore would be likely the first market where we would see end demand increases as well as pricing and market improvements. What we've tried not to do, and is just challenging given the uncertainty of the market, is pin down exactly when that happened. So we've put an end-year supply growth rate number out there, which we said on the last call, and we would reiterate now, in the range of about 30% supply growth for NAND as we exit the year. That's about 10 points below what Lam sees as long-term demand growth. So, again, what we said is, at that point, it feels like the market will have tightened and investment could return. Obviously, pinning that exact timing is challenging.
I feel like the year in terms of supply improvement, demand improvement. Discipline is playing out very close to what we thought it would, with a whole lot of moving parts, but in general, and as I said, directionally, very much like we thought at the beginning of the year. I don't know if Doug has anything to add.
No, perfect. I don't really have anything to add. You're starting to see, Harlan, you alluded to it, some of the early indications that the market is getting healthier, is turning. Inevitably, I think we all know it's a question of when, not an if, that memory spending will recover, and that continues to be how we see it.
Great. Thanks for the insights there. On the heavier China domestic mix, just given the continued trade tensions, U.S. and China, which is actually motivating a number of the China-based companies to bring in more chip design domestically, especially given some of the recent component bans and entity list additions. Have you guys seen an increase in dialogue or programs that suggest a step up on China domestic activities to accelerate their semiconductor manufacturing capabilities?
Yeah, I think it's hard to say, to tie one directly to the other. What we said as we came into this year is that China domestic spending would be stronger this year than last year. I would say that at this point in the year, we feel maybe it's even somewhat stronger than that. Demand from domestic China or indigenous Chinese customers is strong, as Doug pointed out. I think to this point of exactly what's driving that, I think there is obviously long-term demand and a long-term desire to build more domestic capability.
When I think about China, the biggest challenge for us really is the uncertainty that a lot of the trade discussions probably put not only into the investment plans of indigenous Chinese customers, but also into global players who are a little less certain about how those issues might play into the demand environment. We're managing through this. We've said our position in China is strong, from a market share perspective, and I think as you start to see some of the indigenous Chinese customers move into memory, I would assume that our market share position should get even stronger. It's an important region for us, but we track what's going on there closely and manage it as we see best.
Yep, great. Thank you.
Thanks, Harlan. Yep.
We'll go next to Toshiya Hari with Goldman Sachs.
Hi, guys. Thanks for taking the question. Tim, you talked a little bit about your focus on addressing semi-critical applications in your prepared remarks, and I think you threw out an example, in conductor etch, where you managed to have some wins in the quarter. Just curious, just semi-critical applications in general, how meaningful is that part of the market as a percentage of etch and deposition? Where is your market share today, and how do you see that evolving over the next couple of years? I have a follow-up.
Yep, that's a great question. In fact, we said last time that, as we move through the year and perhaps at our investor day when we hold it next, we might break out a little bit more detail on the size of semi-critical as a percentage of our total. We're not prepared to do that today, but semi-critical is a meaningful part of the etch market. I think it's also important, and I really want to make sure that the message doesn't get lost. Our core strength, our core business is the critical market. If you have ambitions to grow and outperform the industry in the long term, you have to also be competitive in semi-critical.
I think that when we look at our performance in the last quarter, I didn't talk about it, but we successfully defended critical positions, in all markets in the last quarter. That was a statement, and that's kind of what we go into every quarter thinking we have to get done, is defend our critical positions. Where we have the opportunity to grow is by taking semi-critical positions away from the competition. There, the defining factor is much more about productivity. As I think I mentioned in the last call, or at least in one of our conferences, productivity is something that the company knows how to do by learning from what has been done for a long time on the deposition side, where I would say a larger fraction of the market exists within the semi-critical space.
What we tried to highlight in the prepared remarks today is refocusing some of our efforts, to ensuring that our productivity that we're delivering to customers is best in class, can yield wins for Lam in the semi-critical etch space. We gave you examples both in the conductor etch space as well as the dielectric etch space. Obviously that can be interpreted as against two different competitors. I think we're making progress there. I think just stay tuned and we'll continue to report how we're doing against those efforts.
Got it. Thanks very much. Then as a follow-up, I was hoping to get an update on how you guys view EUV. How fast, or how slow, rather, the technology is progressing, on the foundry and logic side, how you see that impacting your business into 2020. Then more specifically, I guess one of your customers recently sort of talked about potentially inserting EUV on the DRAM side of their business at the 1Z nanometer node. How do you see that evolving over the next 12 - 18 months, and how that could impact your business? Thank you.
Sure. Okay, great. Well, it's funny. Maybe I'll just come out with a very clear statement at the beginning, because when the EUV question gets asked, it's often, at least I interpret it as kind of like, is this going to have a negative impact on Lam's business? I guess what I'd like to start by just saying is, at this point, we view EUV as being good for Lam. Maybe I'll just run you through a couple of reasons for that. We're aligned to this idea, which I think is held by many of our customers, that EUV is part of the answer to cost-effective scaling. Cost-effective scaling is what's needed for new technology nodes. New nodes are important to Lam, as you can guess. I just said our critical application business is the core of our business.
Critical applications, new ones get created when technology advances from node to node. That's part of our growth strategy, is continue to win the next critical applications that are being developed. New nodes also create SAM expansion opportunities for Lam. When there's a new node, there's new materials and new architectures where Lam can use etch and deposition more effectively. We win new positions. In terms of a negative impact, multiple patterning, we've said also very clearly, continues to grow even with EUV. We've said, obviously, multiple patterning doesn't grow as fast as if EUV didn't exist. Our view is that's one of those hypothetical futures that it's hard to say exactly how many wafers get produced in the future of future nodes without EUV. It's a long way of saying, I think EUV, in total, is good for Lam.
I also said, I think on our last call, that EUV is a big technology transition, and you kind of pointed out the speed with which it gets introduced has a lot to do with productivity. We think that Lam has a significant role that we can play in helping improve. By using etch and deposition, we can improve the productivity of the patterning module as a whole. If we can do that, it creates new opportunities for etch and deposition, we're partnering with ASML on those opportunities, I think it's a big area of focus for us. Specifically to your question, for introduction in logic and foundry, I think our view is consistent with industry consensus. I don't have a lot to say there. Around the question of 1Z insertion, again, I think you know probably what has been said.
It's clearly not the general consensus that DRAM at D1Z will commonly use EUV. It's a node where at least one customer, as you said, is talking about putting it in as maybe the learning node. I think I don't have anything else to add to it than that.
Thanks very much.
Thanks, Toshiya.
We'll go next to John Pitzer with Credit Suisse.
Yeah, good afternoon, guys. Congratulations on the solid results. Tim, at least by our math, if you look at your June shipments into the memory market, they're down about 50%, plus or minus, from sort of peak. They're still up, if you look at almost 50% from the 2014- 2016 average. There's been a lot that's gone on with capital intensity, your SAM expansion, market share gains. You guys have been very clear about not calling a recovery in memory, but I'm kind of curious, how do you think about the current level of spending relative to prior troughs in memory spending when you adjust for capital intensity going up, SAM growth, market share? Do you feel like we're bouncing along a bottom here in memory, or how should we think about that?
The one thing I would say, John, it's Doug, I would say, and we've been saying this all year, is when I look at the spending in memory this year, it is almost entirely allocated to conversion-related investments, right? Which is always a cost-effective thing for the customer base to do. It lowers cost per bit. I don't know that I would say it's a maintenance level or it's a bottom level, John, but it's always something that economically it makes sense for the customer to do it. That's pretty much what we're seeing happening this year. Your observation about SAM intensity, capital intensity going up, and all of that is absolutely valid. It's gotten more expensive to put wafers in place because the complexity of architectures have grown. Obviously, that's part of the calculus as well. I don't know, anything to add, Tim?
Yeah, no. I think in terms of especially etch and dep intensity has changed in a pretty dramatic way since last time. I agree with everything Doug said. Our comments have been we're clearly at a point where we believe that supply growth spending is insufficient to meet long-term demand growth. However you want to call that bottom or trough, as Doug said, we don't really want to do that, but this feels like the majority of investment is really around technology at this point.
That's helpful. As my follow-up, I was wondering if you could just give us an update on your view of the service business for this year, especially in lieu of some of the utilization cuts we've seen through customers. I know the install base is growing, which will give a tailwind to service, is that still expected to grow? As you answer the question, one of your customers with the power outage this quarter, what kind of impact might that have had either on the services or quite frankly, even the shipment business?
John, it's Doug again. I still expect our installed base business to grow this year. Again, your observation is absolutely right. This business will ebb and flow somewhat with industry overall utilization. I think it's pretty well understood that some of our memory customers are reducing utilization to a certain extent, including from a power outage. As a result of that, things like spares consumption will decrease for a period of time. This business will still grow this year, and it grows along with growth in chamber count. As we've been saying, our view, even though WFE is down so much, chamber count will still grow this year. The tailwind of the business over the next several years continues to be pretty good.
Thank you.
Thanks, John.
We'll take our next question from Timothy Arcuri with UBS.
Thanks so much. Doug, I guess both of my questions are for you. The first one is, I'm wondering if you can update us on the comments you gave, I think, on second half versus first half loading. Obviously, the full year, WFE has not changed, but the mix has changed a little bit. It seems like maybe a little bit less in your favor. Can you update us on the second half versus first half of this year? Thanks.
Yeah, Tim. Yeah, you're right, the puts and takes. We're still suggesting WFE this year is down mid to high teens. Within that, there are puts and takes. Memory is somewhat softer than we were describing a quarter ago, and foundry is a little bit stronger. Maybe a decent amount stronger. When I look at the profile of that investment, the spending in memory is somewhat first half-weighted. The spending in foundry and logic is somewhat second half-weighted. When you put it all together, I think WFE this year will be a little bit weighted to the second half.
I guess, Doug, I was more talking about your shipments, because I think you had previously guided your shipments or your revenue second half versus first half, so I'm wondering if you can update that.
No, we've never guided revenue half on half. We've always, and maybe it was misinterpreted, talked about WFE.
Okay. Awesome. Okay. I guess my second question, Doug, is you're doing $12 annualized on a pretty nasty memory cycle. Systems for memory are cut in half, and total system shipments are down somewhere in the range of 40% from the peak. Obviously, it's much different than really anything in the past. I guess the question is, how do you think about how to optimize the capital structure and the balance sheet as you come out of this? You have bought back a ton of stock, but I'm wondering how you think about the right balance sheet leverage targets as you look out over the next few years. Thanks.
Yeah, Tim. I haven't communicated a merit target for leverage or total cash or anything like that. If you look at what we've done over the last several years, I think we've had an inclination to provide meaningful cash back to shareholders. The cash generation capability of the business continues to be amazingly strong, right? We're coming off two quarters now of approximately $900 million in operational cash flow. Our confidence in the ability to sustain cash generation is obviously much higher. If you look at a metric like net cash, it's also why we've been comfortable bringing that down, and we've done that through raising a little bit of debt and consuming the cash by both dividends and more towards share buyback. At some point, we'll probably have an investor event again and talk a little bit more about it.
I'm not ready to change what we've described in the past, which has been we're going to return at least 50% of free cash flow to shareholders, and what you've seen us do is a whole lot more than that over the last several years.
Okay, Doug. Awesome. Thanks so much.
Great. Yeah. Thanks, Tim.
We'll go next to C.J. Muse with Evercore ISI.
Yeah, good afternoon. Thank you for taking the question. I guess first question, can you speak to, on both foundry logic side, your revenue intensity at the 16, 14, 10 nodes, and how we should think about share gains and/or greater opportunities for you as we migrate down to seven and five? If there's any way to kind of quantify what the incremental revenues per wafer start or any sort of math like that would be very helpful. Thank you.
Yeah. C.J., I'm going to let Tim actually talk about the direction. We haven't quantified it, and we're not ready to do that on the call, but Tim is pretty well-versed in the trajectory, so why don't you cover that?
Good, he took the numbers off the table for me. I think maybe the first thing we've said, and we feel quite confident, is in kind of the logic foundry world, our share gains for a variety of reasons, both true wins, but also new applications that we've gained. Our share gains are gains between 10, 7, and 5. We feel quite confident in that. Part of that is intensity of etch and deposition. Part of that is new applications that get created. It's new processes. While we haven't quantified it, I would say that as I said in my EUV commentary, every technology transition is an opportunity for us to gain new applications and gain share. We feel really good about the progress we've made.
It sometimes gets lost in this overlying story about memory and memory spending and how much impact it has on our business. We're feeling quite good about our momentum in logic and foundry both.
That's helpful. As my follow-up, any update on your self-cleaning etch offering? I'd love to hear about the Kiyo module coupled with Corvus. Anything you can share with us would be great.
Well, not sure what I can share with you other than it continues to progress in the marketplace. I talked about semi-critical applications where customers are really focused on trying to optimize the productivity of existing fabs. One element of productivity of fab is how often you have to actually have technicians or people going and doing maintenance on tools. The Kiyo product with the Corvus R, the self-maintaining tool, is part of that answer. Obviously, there's not a lot of spending going on in some of those segments, but we feel this is another example of where we've introduced the right product that when spending recovers in the memory market, these should be perfectly targeted to the types of tools that customers want to put into all those new fabs they're building right now.
Thanks, C.J.
We'll go next to Krish Sankar with Cowen and Company.
Yeah, hi. Thanks for taking my question. I have two of them. First one for Tim. One of the things we've been hearing is that as you go to higher and higher layers in 3D NAND, the process time for etch keeps going up. Is there a way to quantify, in either absolute minutes or relative to prior nodes, what kind of increase in process times are we talking about, and what does it mean for Lam?
Okay. Well, I don't think that we're going to quantify that because it's kind of a competitive piece of information that we wouldn't want to divulge here. To your point, it takes a longer time to etch higher aspect ratio features, and I think that it's relatively well known that that increase is nonlinear with the number of layers, meaning the etch becomes longer nonlinearly with the number of layers that are being created. That's simply due to the etch process physics themselves, physics and chemistry. More layers has a positive impact for Lam on the number of etch tools that are required to accomplish that etch. What I would say is that there's a constant battle.
To keep the cost of ownership reasonable for customers, we're continuously working on productivity of every etch we deliver to the customer, including critical etches like the 3D hole etch. Not prepared to quantify today, but it is a positive grower for us.
Got it. Okay. As a follow-up, Tim, when you start gaining more share of refocus more on the semi-critical etch applications, how should we think about the margin structure? It seems like productivity is key in this segment. Is productivity a euphemism for lower price? I'm just trying to figure out how to think about margins when you get more share in semi-critical etch.
No, definitely in my mind, it's not a euphemism for lower price. Productivity is, in many ways, as much of a technology challenge as any other. When we're attacking productivity, we're attacking it fundamentally from equipment and hardware and process design in a way that we deliver increased productivity with a cost structure of the tool that allows us to deliver corporate average gross margins for those applications. That's our expectation. That sometimes takes time and will require us in many cases to think long and hard about how our tools are designed. That is the expectation that I have for winning in that space.
Thanks, Tim.
Thanks, Krish.
We'll take our next question from Weston Twigg with KeyBanc Capital Markets.
Hi, thanks for taking my question. First, I just wanted to probe a little bit about your comments on 2020. You said you think it's shaping up to be an up year, I was wondering if you could just walk us a little bit through the puts and takes, maybe both in memory and foundry, of what it would take to be an up year and sort of what your expectations are regarding those.
Yeah, Wes, too soon for us to get into specifics on this. Really, the commentary around setting up for a better year next year, it's all about memory recovering in terms of investment levels. We'll give you more color on it when we get a little closer to next year, and I'm sure a lot's going to move around in there. As we sit here today and look at the level of investment occurring in memory and the growth rates of bits exiting the year and whatnot, we believe the level of investment there needs to go up next year, and that's the nature of the comments.
That makes sense. I guess related, too, you talked about some new wins and focusing on market share gains and SAM expansion. Some of the stuff you outlined today, can you help us maybe put some numbers around how much that could expand the opportunity in 2020, the revenue opportunity maybe for you?
We're getting closer and closer to our market share targets that we put out for 2021. I guess you can think of these as, we've come out and we've said that we would gain 4 - 8 points of share within etch and 4 - 8 points within deposition by 2021. That was our last stated set of objectives. When you think about progress we already made in our share position in critical, that's where you move to semi-critical, and you say a fair bit of that share gain, 4 - 8 points, is going to come from those types of wins. That's why we wanted to highlight today that we feel we're just at the start of making some of the progress towards that type of share gain.
Okay. Fair enough. Thank you.
Thanks, Wes.
We'll go next to Joe Moore with Morgan Stanley.
Great. Thank you. I wonder if you could talk a little bit more about the strength in the indigenous China spending. How does that break down between foundry, DRAM, NAND, and just generally, your view on the sustainability of that spending into next year?
Yeah, Joe. We described in the past a view that plus or minus, there was investment levels of $5 billion from indigenous China relative to WFE this year. As we look at it now, a little bit stronger than that, quite honestly. The strength is coming from memory primarily, and I think you know what's going on there. There's several foundry customers there. The big one is SMIC. They continue to be a very important customer for us. You got YMTC investing in NAND, and then emerging DRAM investments. The uptick relative to prior communication, I think, has been primarily related to memory, both a little bit of NAND, a little bit of DRAM.
The maturity of the output that they're getting from that spending, is that something where you could see more kind of a larger capacity as we move forward? Or is it more pilots and kind of getting things figured out?
Yeah. Joe, I think they're going to continue to keep innovating on the technology. They're going to keep getting better and better at what they do, and it wouldn't surprise me if they continue to invest at higher levels as we go forward. I'm not ready to quantify it for you, but they're working very diligently on innovating the technology.
Very helpful. Thank you.
Yeah. Thanks, Joe.
We'll take our next question from Vivek Arya with Bank of America.
Thanks for taking my question. I had two as well. I wanted to also ask about China. Sales were up 20% in the last fiscal year. Non-China was down about that number. How do you measure utilization at your customers? Because their purchasing has gone up right around the time when trade tensions have increased. Is there a risk that there have been pull-ins and this becomes an issue later on?
Vivek, when I look at it, I don't think there's significant pull-ins occurring in what we're seeing happening. I say that because we sort of knew our customers' plans as the year began. When I look at it, for the most part, they're executing to those plans. It's also important, I think, to understand that a lot of the investment in China is not indigenous Chinese customers, right? You've got the Koreans building fabs, Taiwanese, U.S. companies. Understand that a broad swath of the spending in China isn't necessarily the indigenous Chinese customers. That's generally how I see that.
Okay. Then for my follow-up, Doug, on gross margins, you're guiding to 45%. I know you mentioned a few times that mix changes from quarter to quarter. What in the mix is driving gross margins lower? When I look at the revenue level, it's kind of back to where it was in March 2017, at that time, gross margins were higher. I recall at that time, foundry logic was a lower part of the mix. Is it foundry logic that has an impact on gross margin? I think you also alluded to the fact that you do expect to continue to grow in foundry logic . Just how should we think about the trajectory of gross margins over the next several quarters?
Yeah, Vivek, I wouldn't assume, or you shouldn't assume there's differential gross margin by end market necessarily, meaning foundry to logic to memory. That's not the way it works. Generally, it's what I always say. The larger customers, because they're buying more, may tend to get a little bit better pricing, simply because they're buying more from us, relative discounts sometimes, not always. Obviously, when you look at revenue down the way it is, that's probably one of the bigger contributors right now.
Yeah, I think you referenced back to a point in 2017. We look at those all the time to think about what's different in our business. What you have to look at is that as we transitioned through 2017, it was a massive growth cycle for us.
Yep.
While we have a very flexible operating model, some of the physical infrastructure that we had to put in place to meet a $3 billion quarterly revenue run rate, some of that physical capacity has not been taken offline, and it does affect gross margin to some degree. We are confident that that kind of physical capacity is what's needed to be able to respond when our customers do ramp as memory spending recovers. Flexible operating model. Doug talked about temporary workforce, but in some cases, there were costs put in which are still with us. Doug, I don't know if you kind of.
Yep. Perfect.
Okay. Very helpful. Thank you.
Yep. Thank you.
We'll go next to Patrick Ho with Stifel.
Thank you very much. Tim, maybe to follow up from your prepared remarks, you talked about some of the new emerging technologies such as gate-all-around and the industry transition in DRAM to 1Z and 1A. On the gate-all-around, given that it's a similar format to FinFET, are there any capital intensity increases for etch and deposition that would benefit you guys when the industry makes that shift? You guys clearly benefited from the transition to FinFET.
Yeah, I think that what I was trying to message is that at almost every node, our customers, the industry has realized that 3D scaling is a key part of the answer, both to device performance as well as cost scaling. For us, 3D scaling really means etch and deposition intensity tends to increase. Obviously, we're still a little ways away. We're highly engaged with three nanometer, but final decisions ultimately get made relative to structures and architectures in the future. It's really a message around etch and dep intensity to create 3D architectures. It's also why I pointed out, in case it was missed, the 3D architecture that's emerging in the heterogeneous integration or advanced packaging space, which again, is 3D chip stacking and how Lam's etch and deposition can play a role in that.
What I was trying to just translate is I believe that 3D and etch and dep intensity kind of is fanning out across all device types in the future.
Great. My follow-up question for Doug in terms of the installed base business . That's probably providing you a lot of support this year, given the pressures on the systems business. Can you give a little color if you're seeing a lot of upgrade business for, I guess, the trailing edge of fabs that are trying to upgrade, improve their productivity? Is that a key driver to the install base business and the strength that you're likely seeing this year?
Yeah, it's certainly a piece of it. In fact, last quarter, record level for our Reliant business, which is our refurbished equipment. The answer, Patrick, is yes.
[Thanks].
Thank you.
We'll go next to Sidney Ho with Deutsche Bank.
Great. Thank you. I got two questions on the memory side. You talk about NAND bit supply rate exiting this year at about 30%. Can you give us a sense how you think the CapEx or the equipment spend needed to support, say, every one percentage point of bit supply growth, from the current "conversion only" trough? Can you do the same for DRAM? I guess, you have given us the estimated cost for greenfield fabs before. Just trying to put those pieces of information together.
Yeah, Sidney. No, I don't think we've quantified this in the past, and I don't think we're ready to on the call right now. I'm going to decline to answer the question.
Okay. I'll move on to the next one. If I take your reported revenue by end market and compare it to what we think the WFE dollars for the various segments, I can see that your share has been moving up steadily across all the different segments in recent years. The one that stood out to me is NAND, which gets to be pretty high. Just take your revenue number, that's probably including services, and divide it by WFE, you get to somewhere around 40%, from call it 20%-25% a few years ago. Do you think that number could go higher as the market transitions to higher layer count? I guess this question is whether it's sustainable.
Well, I guess, what we've said is that etch and deposition are really key to continued scaling in 3D NAND. Etch and deposition intensity scales with number of layers. From that perspective, measuring share of customers' total spend on 3D NAND, we do believe it can and will go higher.
Okay, great. Thanks.
Thanks, Sidney.
Operator, I think we have time for one more question, please.
Yes, ma'am. We'll take our final question from Mitch Steves with RBC Capital Markets.
Hey, guys. Yeah, thanks for taking my question. I just have one extra one on kind of focusing on the share gain potential here. I don't expect you guys to know exactly what's going to happen to Japan and Korea, but if you look at the etch and deposition tools you guys sell, where do you guys think you have the most opportunity to gain share against Tokyo Electron?
Not going to get specific. Okay. Well, this is also why I talk about the importance of performing really well in the semi-critical space. Critical applications, the reason we like them is they're hard to win, and they take, often, a generation or two to sort of prove yourself out. Those are less likely to switch due to some short-term event. Semi-critical or less critical applications are much more driven by can you accomplish that task at a certain targeted productivity point. I think that there are a number of applications that we would be targeting where we're highly capable of doing those. If the customer is motivated to give Lam a try, we're motivated to jump in there and show what we can do.
Got it. Just one real quick high-level one. I don't know the exact math off my head, but roughly speaking, from the last downturns you guys have seen, do you guys see anything irregular in terms of buying patterns suggesting that people are buying ahead or buying additional equipment than normal down cycles? Do you think this is essentially a normal semi cap cycle that you've seen in the past?
No. Nothing from my perspective that's out of the ordinary. In fact, that was why I kind of started my comments with, for the most part, given some different puts and takes, the year's playing out not that differently than we had originally thought from the standpoint of moving through a down cycle in memory.
Perfect. Thank you.
Great. Yep. Thank you. Operator, I think that's the end of the call, if you want to sign us off.
Thank you everyone for joining.
Yes, sir. Ladies and gentlemen, thank you very much for your participation. You may now-