科林研發 (LRCX) 2020 Q2 法說會逐字稿

Good day, and welcome to the Lam Research Corporation's December Quarter 2019 Financial Results Conference Call.

At this time, I would like to turn the conference over to Ms. Tina Correia, Corporate VP of Investor Relations.

Ma'am, please begin.

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 review our financial results for the December 2019 quarter and our outlook for the March 2020 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 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 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, I will hand the call over to Tim.

Thanks, Tina, and welcome, everyone.

In the December quarter, Lam delivered revenues and diluted earnings per share above the midpoint of guidance, marking another quarter of solid execution and closing out a year of strong performance in calendar 2019.

Consistent with our comments throughout this year, we entered 2020 with increasing momentum and an improved spending mix environment, which we believe will lead to outperformance by Lam.

I have now completed my first year as CEO of Lam Research, and I am incredibly proud of what our people have achieved.

Underpinning our strong financial results this past year has been a company-wide focus on execution and an emphasis on our culture where all employees can perform their best.

I want to thank Lam's employees across our global organization for their efforts and, of course, our partners and customers for their valued support.

I also want to take a moment to address the coronavirus situation.

Our concern, first and foremost, is with the health and well-being of our employees, customers and partners.

And with this in mind, we have implemented precautionary measures within our global business operations.

Additionally, we are donating to Chinese relief efforts to support the people and communities impacted by the coronavirus outbreak.

Now turning to our business results.

In calendar 2019, we generated solid operating cash flows, and at the same time, we invested a company record in research and development dollars to fuel technology innovation and product differentiation.

Our EPS performance, the second best in the company's 40-year history, was especially noteworthy given that memory spending declined significantly year-over-year in 2019.

The impact of lower memory spending was partially mitigated by record revenue from our customer support business group, highlighting the importance of our recurring spares and service opportunity to the overall quality of earnings of the company.

In 2019, we also saw outstanding execution in our product organizations.

We grew our revenue share of WFE spend in memory and foundry logic, and we laid the foundation for additional gains as we recorded our best ever performance in net penetration and defense application wins as measured by revenue potential over the next 3 years.

We are winning by focusing on high-volume manufacturing solutions for emerging technology inflection challenges, including those associated with new scaling architectures, new memory technologies and new materials.

We are capitalizing on the learning from our installed base of tools that today enable some of the industry's most critical etch and deposition applications.

For example, for the 3 most critical applications in 3D NAND, our Strata, ALTUS and Flex tools, each process well over 1 million wafers per month in high-volume manufacturing, allowing us to partner early and in a unique way with our customers on next-generation needs.

In 2019, this led to the addition of new products to our portfolio, such as the VECTOR DT, which is designed to address wafer stress problems encountered during the high-volume production of 3D NAND stacks of 96 layers and above.

Similarly, we enhanced the capabilities of our Coronus product family to improve device yield at the wafer edge by depositing encapsulating layers for bevel protection.

These products expand our served share of WFE spend but just as importantly, they exemplify our commitment to our customers' success.

In 2019, we improved to the #1 or #2 position at all of our largest customers that provide a competitive ranking of their suppliers based on detailed quality, support and performance scorecards.

In part, this improvement is due to our investment in productivity-enhancing upgrades for our installed base.

For instance, Corvus wafer edge solutions have become key differentiators for our conductor and dielectric etch systems in high-volume manufacturing and have been instrumental in helping our customers reduce cost of ownership.

We announced in 2019 a Corvus-enabled self-maintaining etch tool that ran for 1 year in high-volume manufacturing without human intervention.

Multiple customers are now upgrading their installed base tools to include this high value-added capability.

In 2019, we highlighted heterogeneous integration and advanced packaging as new areas of opportunity to leverage our product portfolio and 3D learning to drive growth.

In the December quarter, we built on our growing momentum in this space with additional wins for our SABRE 3D electroplating system at multiple advanced packaging customers.

With these latest wins, we estimate that we have gained more than 15 points of market share in the last 2 years, and we have firmly established ourselves as the technology leader in the increasingly important through silicon via market.

Another area of growth for the company has been in atomic layer deposition or ALD.

Our ALD solutions are gaining significant traction in the market by delivering best-in-class film properties, along with high productivity and low defects.

Due to the inherent advantages of higher film quality and conformity, our Striker ALD oxide systems are replacing older process alternatives, such as spin on dielectric and SACVD for critical applications, pulling more WFE spending into our served market.

Similarly, RC scaling requirements are driving increased demand for our single-wafer Striker carbide and ALTUS ALD metallization systems.

When used for low-k carbide liner spacer applications, our Striker system has been able to achieve 30% better RC properties versus competing batch tools.

For metallization, we closed out 2019 on a high note, with significant wins in logic foundry, for our ALTUS systems as customers look to replace the conventional barrier fill sequence with an integrated ALD approach to lower device resistance in a cost-effective way.

Across all products, we exited 2019 with approximately 61,000 chambers in our installed base.

Our installed base revenues grew year-over-year and reached record levels in 2019, with significant contribution from our Reliant business, which also reached record levels.

Productivity upgrades and solutions grew more than 30% year-over-year, as we have worked to help customers enhance the performance of their existing assets.

Furthermore, we signed several new multiyear customer support contracts.

These multiyear contracts enable Lam to significantly reduce customers' running costs, while generating a recurring revenue stream for Lam.

Now turning to WFE.

We estimate 2019 WFE ended in the $46 billion to $47 billion range, slightly higher than our prior mid-$40 billion estimate.

The increase was driven predominantly by higher foundry logic spending.

For calendar 2020, assuming no material impact from coronavirus on our full year outlook, our view calls for WFE spend in the mid- to high $50 billion range, supported by sustained strong spending in Foundry and Logic, and significantly for Lam, improved spending in memory led first by NAND.

Overall, we expect spending in Memory and Foundry logic segments to be up year-on-year in 2020.

To wrap up, Lam delivered strong financial performance in calendar 2019.

And as our March quarter guidance suggests, we're in a great position to drive higher in 2020 with the improvement in Memory spending.

Lam's product pipeline is very strong, with more innovation on the way, and we look forward to sharing more with you at our upcoming Investor Day on March 3.

Thanks.

And now here's Doug.

Great.

Thank you, Tim.

Good afternoon, everyone, and thank you for joining us today in the middle of what I know is a very busy earnings season.

We concluded calendar year 2019 with a strong December quarterly performance.

Results ended up better than we expected, primarily due to a little bit stronger NAND investment as well as an uptick in our installed base business.

Our installed base business delivered another record year and continued to be a stable profitable business for us.

And as Tim mentioned, from an earnings per share perspective, it was the second best year in our 40-year history.

The December quarter results came in over the midpoint of guidance for all financial metrics, with diluted earnings per share essentially at the high end of our guidance range.

As Tim also noted, we're pleased with our performance in calendar year 2019 and we delivered solid profitability levels within a challenging memory environment.

As we discussed at our earnings call last quarter, there were continued strong investments in the Foundry and Logic segments in the December quarter.

We had the highest system revenue dollars for the Foundry segment in our history, and the revenue concentration was at its highest level since the September 2016 quarter.

Foundry spending continued its focus on the 7- and 5-nanometer nodes and it represented 36% of our December quarter system revenue.

The Logic and Other segment grew in dollar terms and was essentially flat with the prior quarter intensity level coming in at 12% of system revenues.

It was the highest Logic and Other revenue level in dollar terms in 2 years, driven by 10-nanometer, image sensors and other specialty markets.

For Memory, the combined segment decreased to 52% of system revenues from the September quarter, which was at 64%.

We had a decrease in the December quarter in the nonvolatile memory segment going from 38% to 35%.

Quoted from a dollar perspective, revenue in the segment actually increased.

The DRAM segment decreased from 26% to 17% of system revenue.

NAND investment continues to be focused on 64-, 96- and initial 128-layer devices.

And DRAM spending continues to be primarily focused on node transitions.

Revenues for the quarter were $2.584 billion, which was above the midpoint.

In addition to the leading edge foundry and logic strength I mentioned, we saw continued investments in the China region, with the majority, again, coming from domestic Chinese customers.

China geographic revenue came in at 29% of total revenue in December.

Gross margin came in at 45.7%, 70 basis points above the midpoint.

Strength in the December quarter gross margin is related to customer and product mix as well as increased factory utilization levels relative to the prior quarter.

I'll remind you, as I always do, that actual gross margins are a function of several factors such as business volumes, product mix and customer concentration, and you should expect to see some variability quarter-to-quarter.

Operating expenses for the December quarter came in at $481 million.

Our variable compensation spending was higher in the quarter as it's tied to the increased level of profitability.

Spending in the December quarter also increased due to the appreciation of the market during the quarter and the resulting impact on the cost of our deferred compensation plan.

As I've mentioned in the past, we do hedge this to mitigate the exposure to the income statement.

However, because of the accounting rules, the offset to this expense shows up in other income and expense.

It's basically a neutral impact to earnings per share at the end of the day.

We continue to invest in our critical research and development programs, and you'll hear more about our commitment to technology and productivity leadership at our upcoming Investors Day in March.

I'd also remind you that as we look ahead to the 2020 calendar year, we'll see the normal seasonal spending increases related to the March quarter.

Operating income in the December quarter was $700 million and operating margin was 27.1%, essentially at the midpoint of guidance.

Our non-GAAP tax rate in the quarter was 12.5%.

I would like to highlight that the difference between the non-GAAP tax rate and the December quarter GAAP tax rate, which was 23.5%, was related to the reversal of the tax benefit from the Altera stock-based compensation case.

I think you're seeing this from lots of companies in the technology space.

You should expect the fluctuations in the tax rate will occur quarter-to-quarter.

And as we look into the rate for the calendar year 2020, I expect it to be in the low-teens level.

Other income expense was up slightly from the prior quarter at a total of approximately $13 million in expense.

The main components of OI&E are interest income from our cash and investment balances offset by interest expense related to the outstanding debt.

You should expect that other income and expense will fluctuate quarter-to-quarter based on several market-related items, things like foreign exchange or what I'm talking about.

Moving on to capital return.

For the December quarter, a $167 million of cash was deployed in dividends and $1 billion in share repurchase.

As we've frequently done in the past, the share repurchases were done through a structured share repurchase program that cover repurchases through the June 2020 quarter.

We remain on track with our commitment to capital return.

For calendar year 2019, we completed $3 billion of our current $5 billion buyback authorization.

In total, our capital return activities represented approximately 158% of free cash flow in 2019.

Diluted earnings per share was $4.01.

Our diluted shares continue to decline, and we ended the December quarter with diluted shares at approximately 150 million shares.

This is the eighth consecutive quarter where our diluted share count has declined.

The share count includes a dilutive impact of approximately 5 million shares from the 2041 convertible notes.

And I'll remind you that the dilution schedule for the 2041 converts is available on our IR website for your reference.

Let me now move to the balance sheet.

Our cash and short-term investments, including restricted cash, decreased in the December quarter to $4.9 billion from $5.8 billion in the September quarter.

The decrease quarter-to-quarter was due to the share repurchase and dividend activity, offset by cash flows from operations of $308 million.

And as I've mentioned before, when you see business levels grow, working capital levels generally increase, which impacted our cash flow from operations in December.

We concluded calendar year 2019 with the second highest level of free cash flow in the company's history at over $2.3 billion.

I believe this truly demonstrates the sustainability of our business through a lower industry spend period.

DSO increased slightly due to the timing of collections to 72 days versus 69 days in the prior quarter.

Inventory turns improved to 3.7 turns from 3.2 in the September quarter.

Both receivables and inventory grew in dollars during December as business levels increased.

Noncash expenses included approximately $46 million for equity compensation, $49 million for depreciation and $17 million for amortization.

December quarter capital expenditures increased to $62 million from $39 million in the September quarter.

Ending headcount as of the December quarter was flat with the prior quarter at approximately 10,700 regular full-time employees.

We expect that as revenue levels are growing, we'll add headcount to support the increasing business.

So now looking ahead, I'd like to provide our non-GAAP guidance for the March 2020 quarter.

We're expecting revenue of $2.800 billion, plus or minus $200 million; gross margin of 46.5%, plus or minus 1 percentage point; operating margins of 28%, plus or minus 1 percentage point; and finally, earnings per share of $4.55, plus or minus $0.40 based on a share count of approximately 149 million shares.

We see continued strength in Foundry and Logic spending going into the March quarter.

And additionally, we see NAND spending continue to increase going into 2020.

The March quarter guidance reflects our current view of the business environment, including our assessment with the potential impact from the public health situation in China.

We see business disruptions potentially with both customers and suppliers, that are essentially extending the Lunar New Year holiday through February 9. Absent this situation, our numbers would have been somewhat higher.

We also increased the revenue and EPS ranges to take into consideration the uncertainty of the impact from these activities.

We believe this is temporary, but the issue is developing day by day.

I think we're taking a prudent approach to what we're doing with the numbers.

So to conclude, we're well positioned heading into 2020.

We're on a strong trajectory to outperform based on our product portfolio as well as operational strategies.

And that concludes my prepared remarks.

Operator, Tim and I would now like to open up the call for questions.

(Operator Instructions) And our first question will come from John Pitzer with Crédit Suisse.

My first question just is going to focus in on the domestic China market.

Doug, you were kind enough to tell us it was the majority of the China business in the December quarter.

I'm wondering if you can help us understand how much of a majority it was.

And what was the split between sort of Memory and Logic?

And then as you look out to your WFE forecast growth of at least 20% in calendar year '20, how important is the domestic China market?

And how should we think about the Memory versus sort of the logic foundry side of that?

Yes, I'll answer and then I'll let Tim add on.

Yes, John, I'm not going to get into the habit every quarter of giving a precise quantification except to tell you, the majority in the December quarter was from domestic China.

And that's been the case the last couple of quarters.

And the reason I've been mentioning that over the last couple of quarters is it's different than the last several years have been where the majority has actually come from the global multinationals.

Then your question relative to how we're looking into 2020 in China.

I think 2019 probably finished with WFE from the local China customers a little bit above $6 billion, something like that, and it's growing in 2020.

And as we look into 2020, I think, John, probably it's up $2 billion to $3 billion is my best guess from local China.

And that spending is broad-based.

It's NAND, it's DRAM, it's Foundry and Logic.

So it's not one or the other, it's a broad-based set of customer spending.

Tim, you want to add anything?

No, I think that pretty much covers it.

And then guys as my follow-on, just on WFE and kind of your share, Tim, as you talked about in your prepared comments, you've done a nice job kind of gaining your share of WFE.

Is 2020 a year where that becomes a little bit more difficult with sort of the addition of EUV?

Or how important is the EUV to your growth projections this year?

And as we think about EUV deployment, is that a good leading indicator for your future business?

Or how should we think about that?

Well, yes, I think that we've talked about EUV a number of times.

And so I guess, maybe I'll just repeat a few of the things I've said and maybe add a couple of comments.

In general, what we've said is that technology transitions themselves are very good for Lam.

We have highlighted a number of times that Lam's served market actually grows at each technology node within foundry logic, even with the introduction of EUV.

So really, what we want is we want the market and our customers to be able to keep moving their technologies forward.

Specific to patterning, as you move from, say, 7- to 5-nanometer, even in the case of EUV, what you're starting to see is the increased use of more high-quality hard masks moving away from spin on dielectric towards deposition methods like PECVD, where Lam has a very strong position.

And so in many cases, our SAM is actually increasing because we're pulling in applications that before were actually done with older processes and now they're coming into our more critical space.

And so with that, we see SAM increase.

And therefore, even with EUV, we see a growing market for ourselves.

Doug talked about highest dollars from foundry logic.

And that's just, I think, further evidence because we're seeing technology investments right now at those EUV nodes.

Now as far as the leading indicator, sure, what we said is that EUV is, obviously, those shipments are signs of these more complex technology nodes and again, our SAM grows where opportunity grows.

And then, John, the only thing I would add, as you think about wallet share of WFE going into 2020, I think, you know our SAM and share in our NAND is very good.

And clearly, 2020, our view is going to be a stronger investment period for the NAND industry as we've seen kind of pricing and profitability stabilize.

So our share of wallet, actually, in addition to the EUV commentary, Tim had, is actually going to have a tailwind from the fact that NAND spending, I think, will be stronger I would say.

Our next question comes from C.J. Muse with Evercore.

I guess, first question on WFE.

If you think about the roughly $8 billion to $10 billion increase year-on-year into 2020, can you tell us how much of that is Memory versus foundry logic?

Is that kind of a 80-20 split?

And then as part of that, are you including the material uplift in DRAM within that?

Or is that something that could be a source of further upside?

Okay.

Great.

C.J. great question.

I don't think we're going to give you the exact breakout, but I think we've characterized it as -- I think maybe you can look back to what we said as we were exiting last year.

What we said was that we were seeing very strong spending in foundry logic, and that we were just starting to see the early signs of improvement in NAND, and we felt that DRAM would come after the NAND market had improved.

And so I think what you're seeing in our outlook for 2020 is that starting to play out, except for, again, sustained strong spending in Foundry and Logic.

So maybe the emphasis on sustained and a strong uptick in NAND spending, which, as Doug just commented, is a significant tailwind for us in terms of our SAM as a percent of WFE.

Relative to DRAM, I think we characterized last year primarily as a technology investment year with very little capacity additions.

And I would say that, in general, our outlook this year is that, that doesn't change much, at least until it's later in the year.

So I'd say primarily sustained strong Foundry and Logic spending -- Foundry and Logic and a strong uptick in NAND.

Great.

And as a follow-up question on the service side.

Thank you for providing the installed base numbers.

Very helpful.

So up, I think, 12% in calendar '18, up 9% in calendar '19.

Can you help us kind of think about what the growth rate for service could look like?

And is it fair to say that service in 2019 reached roughly mid-30s of total revenues?

So I'll take the last one.

It wasn't quite that high, C.J., but it was comfortably above 30% in 2019.

I'll give you that much.

Yes.

We're -- I think that we're very confident.

We've said that because of the growing installed base, the overall customer support business grows every year.

How it grows each year, I've talked a little bit about the dynamic where depending on where you are in kind of utilization and spending and customers will spend on different things.

I highlighted a very strong year for us in 2019 on productivity upgrades, which is one of the things customers do when they're looking to squeeze a little bit more out of the assets they already have in their fabs, and we saw a lot of that last year.

I think that as you transition now to a very busy year and customers ramping aggressively, then you'll find yourself much more in these spares and some of the value-added services part.

But in general, given the dynamics of that market and our product portfolio, we feel very confident about its ability to grow year-on-year with the installed base.

Our next question comes from Harlan Sur with JPMorgan.

Good job on the solid business execution.

With the complexity of the manufacturing process, combined with tighter specifications, maybe longer install and qual times, on average, from the time that you ship a tool to the time that, that tool is put into full volume production, I think greenfield is around 6 months.

Is this lead time around the same time for conversion and incremental capacity expansion?

And the reason why I ask is, I just want to do a sanity check that for some of the memory shipment goodness that you're seeing hopefully, this is going to be layering on the in sync with the demand pull for memory, which is typically more kind of middle to kind of second half of the year weighted?

I don't think we've really broken it out.

In fact, I don't know that I could even really speak to a material difference in lead times.

I guess, if I caught your question, it was -- does a conversion have approximately the same lead time for the upgrade to take place and the product to be qual-ed.

And if that was the question, I would say that they're not dramatically different in terms of the cycle time.

Yes.

And Harlan, maybe I'm guessing what your question might be.

I mean, when we look at the way spending occurred in the second half of '19 in NAND, and I'll remind you, we've been saying this for a while.

We exited the year, we believe the industry supply growth 30% year-on-year.

Based on the investments we saw occurring in the second half of '19, we believe that rate will decline, supply growth will decline into the first half of '20, right?

And so you're starting to see an uptick in investment that will have a lag effect to it, and it won't immediately be producing output.

I think that's inherently what you're asking, right?

Exactly.

That's exactly right.

And I appreciate the response.

And actually, as a follow on to that.

So you guys had anticipated NAND bit supply growth exiting 2019 around 30% below the demand rate.

I think some of your customers in DRAM were saying bit supply kind of low to mid-teens, again, below kind of the demand rate.

So given the improving fundamentals in memory this year, kind of extrapolating from your customer equipment requirements this year, how do you guys see bit supply growth in NAND and DRAM this year from kind of the depressed levels exiting last year?

Yes.

So you've got it exactly right what we said.

We exited the year well below the long-term demand trend.

And actually, even with the guidance we've just given, we actually see ourselves under growing long-term demand this year in terms of bit supply growth, both in NAND and in DRAM.

Our next question will come from Atif Malik with Citi.

Good job on results and guide.

If I look at your top line growth historically, you outperformed WFE by 2x or so.

How should we think about your top line growth relative to WFE for this year?

Atif, the color I'd give you is, again, what I've already said is, I think it's an uptick in NAND investment.

Our SAM in NAND is really strong.

Our share in NAND is really strong.

And so to the extent that NAND is driving an uptick in WFE, Lam will do really well.

And you've seen that over the years, in times when memory is spending more, we do real well; and when it's going the other way, we do less well.

You should expect the same thing.

I'm not going to quantify it for you, whether history is a good indicator.

I'm not completely sure.

But I'll be very disappointed if we don't outperform this year.

Very helpful, Doug.

And as a follow-up, your WFE number for this year is even bigger than what we were thinking, $55 billion.

Can you just talk about the opportunity in image sensor, advanced packaging, things like that?

I mean how big that number is in your WFE number for this year?

Yes, we put that in our Logic and Other space.

And what Tim indicated, as we expect Foundry and Logic to grow 19% to 20%.

And I don't know the exact image sensor number, so I can't answer your exact question.

It's a secular grower for sure.

But in the grand scheme of how big Foundry and Logic is, it's actually not all that big, but it is something we do really, really well in terms of the TSV aspect of image sensors.

Yes.

All I would add is one of the reasons you've heard us talk a lot more about specialty technologies and some of the products that we introduced in 2019, really focusing on heterogeneous integration and advanced packaging and MRAM and the image sensors is because many of those items are somewhat growing year-by-year from applications that are a little bit decoupled from the larger WFE cycle.

And so again, even last year as WFE was down, we were setting records and what we kind of refer to as our specialty technologies business, which is primarily served by the Reliant tools.

And so, yes, I think you can continue to see us focus in those areas because they are places where our technology plays very well.

And as Doug said, we can -- by bringing the right technology and cost point, we can really gain quite a large portion of that market.

Our next question will come from Vivek Arya with Bank of America Securities.

And congratulations on the strong and consistent execution.

I have 2 questions as well.

The first one, I'm curious what goes into your WFE calculation just because it's such a strong outlook for this year?

Is it orders?

Is it Capex?

Is it memory pricing assumption, I'm just curious what goes into it?

Then as part of that, if you could also give us some color on how you think the shape of WFE looks like first half versus second half?

Yes.

I'll go through it, and then I'm sure Tim will have comments on that.

I mean, we go after this in a couple of different ways.

We talk to customers, first and foremost.

Customers tell us, here's our plan, here's what we think we're going to do kind of customer by customer, fab by fab.

So that's the bottoms-up approach to it.

It doesn't mean it's exactly precisely correct, it changes.

Customers' plans can change, but that's one aspect of how we come at it.

The second aspect is actually a tops-down approach, which is we take a read on, okay, what is global GP going to look like this year?

What does that mean for electronics demand?

What does that mean for IC unit and semi revenue growth?

And what does that mean relative to wafers and technology?

And then we try to correlate the 2 approaches together.

We don't always get it perfect, Vivek, but it's a pretty robust process, and I think we do a decent job, but we don't always get it exactly right, but we do our best.

I don't know, Tim, anything?

Yes.

No, I think it stems, as Doug said, from pretty extensive conversations with customers.

One of the nice things about this business is that we have developed very deep relationships with the customers, and they're just as interested in having us deliver the technology they need on time.

And so I think it's pretty open sharing about when tools are needed, when the ramp is going to occur.

So that factors quite a bit into the calculation.

And anything on first half versus second half?

Yes.

I think, Vivek, it's going to be a little bit of a first half weighted year, really driven by Foundry and Logic.

I think Memory spending is going to be pretty steady through the year.

But I think Foundry is probably a little bit first half weighted.

Got it.

And for my follow-up, your installed base has grown at a 10%, 11% annual base in the last 5 years, which is pretty remarkable given the volatility in WFE.

My question is, how fast have your services kind of grown annually during those 5 years?

And how should we think about the installed base and services growth for the next 2 to 3 years?

Yes, I'll give you a little color.

I mean, the best indicator of how its growing is just chamber count because that really is what drives the business opportunity.

Having said that, we're doing a lot of innovative things around advanced service offerings in different ways to deliver value to customers that drives our objective, Vivek, at the end of the day to have dollars grow faster than just the growth in chamber count.

And if you come see us in March, we'll probably give you a little more color around what this business looks like.

Yes.

And I think the only thing I'd add, I made comments about the productivity upgrades and solutions part of that business.

It's -- obviously, every customer is driving to reduce their running costs in their fabs.

And so between upgrades for existing tools as well as new really data-enabled services that we offer to help recover tools more quickly with less labor and with a higher probability of first-time right.

Those -- that part of the business is actually growing very fast, and I think has the ability, just as it did this past year, growing 30% year-on-year, the ability to outgrow the installed base.

And so we're investing in that area.

Our next question will come from Krish Sankar with Cowen and Company.

And congrats guys on the great results.

Just 2 quick ones, either Tim or Doug.

When you look at the calendar year 2020, and like you mentioned, WFE is probably front-half loaded, mostly foundry logic biased going towards Memory in the back half.

Are there any nuances in gross margin we should worry about?

Or is there like one segment has a better gross margin profile than the other?

And then I had a follow-up.

No, there's nothing specific to gross margin by segment that you should be thinking about.

It varies with customer mix.

It varies with product mix and then overall business volumes is what, I think, I've said every single quarter that I've been at the company, it's the same language.

So there isn't anything specific by segment.

Got it.

Got it.

Doug, and then just a follow-up on the OpEx.

I understand March has the seasonality with all the merit increase, et cetera, but how do we think about OpEx on a go-forward basis?

It's like 18% of sales a good bogey to use?

Or is the $500 million plus run rate, quarterly run rate, the right metric?

If you look at what we've done over the longer term, we've let spending grow at a rate less than the top line has grown over the last, I don't know, since we brought the 2 companies together.

And I think you're going to see us thinking that same way as we see a stronger year in front of us.

And again, come see us in March, I'll give you an updated financial model as well.

All right.

Next, we have a question from Timothy Arcuri with UBS.

Doug, I had 2. First of all, China has been running about 30% of revenue for the past 3 quarters.

Should we expect it to be about 30% as a percent of revenue in March?

And I assume within that, that it would still be like the vast majority of it would be the domestic guys.

Is that correct?

Yes.

I'm not going to get into kind of telling you quarter-by-quarter what it is.

Let me remind you what I said earlier.

I think it grows $2 billion to $3 billion off a base of 6-ish -- a little above $6 billion for the year.

It will be lumpy.

Tim, you know, how this goes.

There's a project that takes tools, then it has to ramp it for a while and digest it and something else kicks in.

I'm not going to give you exact color on the March guide, but I think it will be a good year for local China this year.

Okay.

Great.

Doug, and then I guess, this as a kind of a bigger picture question, but given what's going on with export control, I mean, you're saying that domestic China is going to add maybe $2 billion to $3 billion to 2020 WFE.

Do you think that there's any of this incremental stuff going on within China, where customers are concerned about maybe someday down the road, they can't get access to tools?

And so maybe they're pulling in time lines of these fabs?

Yes.

I guess, I'll take that one, Tim.

But I mean, we've answered this a couple of times that we've seen no indication that the China buys or pull-ins.

There appear to be solid plans behind what they're trying to do.

And every tool we ship, they're trying to ramp as quickly as they can.

So I really don't -- I don't think that plays into it.

Having said that, we are watching what's going on with the talk about regulatory controls.

But at this point, no details that we could even really comment on.

Our next question comes from Joe Moore with Morgan Stanley.

I wonder, back to the topic of this China sovereign business, if you could give us some color on the number of customers, I think people tend to think about Memory, but just -- I'm not looking for specifics, but how much of it is Memory versus Other stuff?

And just an indication of the breadth that you're seeing?

Yes.

So Doug made a comment about the demand being pretty broad-based.

I mean, we are seeing -- and again, you sort of think about what's great about our market right now is the breadth of driver -- end market drivers.

There's a lot of IoT and specialty technologies, whether it'd be automotive, power devices, image sensors.

You kind of see all of that be -- people in China trying to address that end demand.

And so it's not isolated to any one thing.

Clearly, the uptick in Memory has outsized importance for Lam's revenue.

And obviously, we watch very closely what happens, especially in the NAND and DRAM areas.

But really, our outlook in the $2 billion to $3 billion increase that Doug talked about in 2020 is pretty broad-based across a -- quite a number of customers.

Okay.

Great.

And then you guys have talked about a kind of $14 billion -- or $70 billion 5-year WFE for NAND, kind of keeping us on a historic supply growth trajectory.

Is that still kind of the right ballpark?

And we see -- it would seem to be quite a bit lower than that right now.

Is that still kind of the right framework for thinking about what WFE ought to be if we stay in the historical demand context?

Yes.

That's still the way we look at it.

And our comment, you just said, we were quite a bit below that at this point.

That was part of our commentary as I said, through last year, we were actually under-investing for long-term demand.

That long-term demand is what establishes our $70 billion 5-year investment.

And even this year, as we see a strong uptick in NAND, we still, as we said, expect ourselves to exit the year, maybe slightly still below the long-term demand rate.

Our next question comes from Patrick Ho with Stifel.

And congrats also.

Maybe, Doug, first, in terms of the installed base business and the growth you're seeing there and the sustainability of that growth.

Can you just give a little bit of color of how you've seen maybe over the last few years and going forward, how the business has kind of transitioned from, I guess, the transactional break-and-fix type of business to more of the contract base, which gives you a better visibility into the revenue and cash flow stream of that business?

Yes.

Patrick, I'm actually going to redirect it to Tim because I think he'll give you a more comprehensive answer than I would.

Doug, can certainly add something there.

But I think we have seen that transition, and it's not just something that's happened to us.

It's been a strategy we've been driving to try to make this business less transactional, much more value-add and also much more long term.

And I talked in my script about one way in which we do that, you transition the transactional to longer term through multiyear contracts.

Even there, you're able to help the customer control their cost in a much more predictable way.

And that gives us better visibility going forward, makes the business a little less transactional.

But I think the biggest way that we've transitioned this business from transactional to strategic is really through our focus on new products in the portfolio of services and upgrades that allow the customer to change the cost profile of running their installed base.

And so in that way, you can make this much more strategic.

I talked about the Corvus R technology, the self-maintaining tool, where the system itself replaces some of its consumable parts rather than having to have technicians go and do that work.

When we engineer that type of new upgrade, you then create for yourself an entire new SAM within your installed base.

And now, as customers recognize the value of that automated replacement system, they go and they upgrade to all of those older tools that are in the installed base, and it creates another source of ongoing revenue for us.

And in the meantime, we're looking for that next upgrade to engineer and offer to the customers.

And so I think it's becoming much more of a product and strategic refocused business.

It just happens that its targeted market is the installed base that we've already shipped.

Great.

That's helpful.

And maybe as my follow-up question, and maybe for Tim, since you talked about in your prepared remarks about the ALD market and the market expansion opportunity there.

On top of the share gains you've made in that segment as well as the market growing, particularly in the foundry logic space, do you see additional applications and foundry logic driving this growth?

Or is there a opportunity for it to expand to other segments like DRAM?

Yes, it already is expanding into other markets.

I mean, the only difference being, again, with ALD, you're essentially creating a deposition technology, a platform, and that's one thing that Lam is very good at is we're creating high productivity platforms.

And then we expand the portfolio of films that we can deposit, and that creates more application opportunities for us.

So I talked about oxides and carbides.

There are other films that are also being developed at the same time.

And I think that, ultimately, every device type reaches a point where it needs to do some level of atomic layer processing.

And so long term, I think this is just where the technology is going.

That's why you see ALD starting to replace, as I said, some older process alternatives for certain steps.

Our next question comes from Joe Quatrochi with Wells Fargo.

Congrats on the results.

I was curious on the NAND flash side.

I was wondering if you could talk a little bit about the amount of clean room space that you see across the industry today and that's empty.

And then how do you think about just the industry's kind of views on filling it versus past cycles?

Joe, are you still there?

Sorry, you broke off at the end I thought.

Yes.

Okay.

There's clean room space out there in the industry, for sure, across all of the customer base.

And so what I always say is, in fact, I think, I don't know, a quarter or 2 ago, we talked about the fact that we were tracking double-digit number of new fabs in 2019.

I think, the number is probably about the same in 2020, a lot of those being memory fabs.

The thing to understand with our customers is building the shell is, it's a cheap call option on the future, right?

It doesn't mean they're going to equip it tomorrow, but it does indicate to you a long-term intention.

And then they buy equipment when they really need the output.

That's where the real money and capital gets deployed.

But there's plenty of clean room space out there.

Yes.

I guess the only thing I'd add, I thought perhaps you were going to ask about the importance of clean room space and I was going to be able to tell you.

That's the -- one of the things that we focus on, you might have understood in the past from past commentary that a benefit of some of our products like the Strata, and why it's popular within the NAND market is, the number of wafers that we output from that tool per square meter of fab space that's consumed is actually the best in the industry.

That's an important metric.

Even when Doug says, there's lots of fab space, which I certainly agree with, how they use it is always important to the customer.

And so again, as we think through platform design and launching new tools, what we call footprint density or throughput density is really one of the key metrics that we focus on.

That's helpful.

And then just back on the services side.

Given that we should start to see, I would think, an acceleration in the growth of your installed base this year with just a rebound in Memory demand, should -- is it fair to us to think about accelerating year-over-year revenue growth for that business in 2020?

Joe, the way I generally think about it is, the chambers ship and then they monetize over a period of time after they ship.

And so the fact that you saw chambers grow in 2019, that's a good indicator of what's going to drive growth in 2020.

That's how I generally think about.

There's a little bit of a lag to it after the chamber count, which is why, historically, we've given you a chamber count number one time a year at the end of the year.

Our next question comes from Mehdi Hosseini with SIG.

All the good ones have already been asked.

Just a quick follow-up for Doug.

How should we think about OpEx just roughly beyond the March quarter?

Yes.

I kind of answered that already.

And I kind of didn't answer it in how I answered it.

But what I said, I'll give you a new financial model in March.

Okay.

And what I said is, over the last several years, the same management team is here that's been around the company for a while.

We like to see revenue grow faster than that we want to let spending grow so that we deliver some leverage to the bottom line.

That's still very much how we think about it.

But your March quarter shows typical seasonal uptick.

And I know we have to wait until March 3. But as we model this kind of fine-tuning our estimate, should I think of your OpEx remaining at the minimum at the same level, it's not going to just be a onetime event?

Yes, that's probably not an unreasonable way to be thinking about it for now, Mehdi.

All right.

Our next question comes from Mitch Steves with RBC Capital Markets.

I think the majority have been answered, but I just want to clarify, maybe 2 small things.

So number one is the comments on DRAM.

So it sounds like that's going to be up year-over-year for 2020.

I just want to make sure that's right.

And there wasn't anything to read between there?

And then secondly, I don't expect numbers from this, but is the message that you guys believe you're going to gain more share in NAND relative to the Other markets in '20?

I'll take the first one and then I'll let Tim comment on the share.

I don't really see -- when I look at the WFE, the color that we just gave you a real recovery in NAND.

I wouldn't describe it that way -- sorry, in DRAM, in DRAM.

Clearly, in NAND, yes, Foundry and Logic continues to be strong.

DRAM is kind of a push, plus or minus.

And what we see happening is, there's still a little bit of inventory out in the channel.

It's got to get burned off.

And at some point, you can't continue to sustain supply growth below where demand growth is.

And when we look at DRAM bit growth, we think supply growth in 2020 is below demand growth.

Yes.

And I think a little bit longer term and also relative to clearing the inventory, we've been very encouraged by comments we've heard about what we understand are drivers for DRAM.

The growth in the server market, the introduction of 5G, where you're seeing pretty substantial step-up in terms of DRAM content per handset.

And so it's -- and I think Doug said this in the past, but it's even for DRAM, a little bit more matter of when, not if.

So I think it's just -- right now, we have a, as you said, maybe about a push for the outlook that we have just given you.

Your comment about NAND.

I guess, when we think about share or share of total spend, one of the things we've talked about is the strength for us in NAND, obviously, is that we are -- and our tools are key enablers of building taller and taller NAND stacks.

And as those stacks increase our share of total WFE continues to increase.

And so given that year-on-year, we continue to see a greater portion of spend being spent on taller NAND devices, our share of total spend will increase.

Okay.

That makes perfect sense.

Just really a small one if I could.

What was just like the dollar impact to the virus that you guys are taking out of the Q1, just to get an idea?

I'm not going to describe the specific dollars.

It was an amount that we think is consistent with what we're hearing from customers and suppliers.

All right.

Our next question comes from Quinn Bolton with Needham & Company.

One near-term question, one longer term question.

On the near-term question, given the travel restrictions in China around the coronavirus outbreak, can you guys say, are you guys having difficulty accessing some of the fabs, either in the installed base business or delivering new tools?

Or is it still somewhat business as usual and you have access to service the tools in the fab, you're able to deliver tools here in the near term?

well.

I think very specifically, I think, most people are aware, the travel restrictions have been placed on travel in and out of Wuhan itself.

And so obviously, we can't travel into that area, where there are some customers located.

But in general, I think that, as Doug said, we're moving through the Lunar New Year, which now has been extended and got of abundance of caution Lam has also implemented travel restrictions for our people to kind of business-critical situations only.

And I think from the standpoint of the impact and the way we've looked at it is, there could be delays of a few weeks in our -- in supply chain or shipments to customers.

And we don't see that as any real change though to the demand picture that we've outlined.

And therefore, we're really talking about just hopefully seeing this normalize over the next few weeks.

Great.

That's helpful color.

And the longer term question is your 2020 WFE forecast, I think, is higher than probably what many of us were thinking.

I guess, as I look out to 2021, it sounds like DRAM is not part of the 2020 forecast.

Do you think 2021, WFE can continue to trend higher if that DRAM spending comes back?

You're kidding.

You know we're not going to talk about (inaudible) '21 is.

Listen when I think about it though the long-term growth drivers are totally intact.

We've been talking about this for years, right?

You got data exploding in society.

You got cloud and hyperscale, the big seven, what have you, investing in more in equipment this year, that's good.

You've got 5G on the com line, that's good.

You've got density growing there in every phone.

So the long term is really very positive for the industry and our position in the industry, enabling it, we feel really good about where we're sitting.

Okay, operator, we have time for one more question, please.

Okay.

My last question comes from Sidney Ho with Deutsche Bank.

Great.

My first question is on the Memory spending this year.

If things play out the way you think, are you thinking that NAND CapEx will still be below that $70 billion run rate?

I know you said that you weighted below the 40% bit supply growth.

I guess, in a way I'm trying to parse out that $8 billion to $10 billion increase in WFE you guys forecasting?

Yes.

Sidney, we never give you segment by segment WFE numbers, and I'm not going to now.

It is going to grow this year.

And as Tim described, we think bit growth in NAND is in the first half anyway, below where long-term demand is, and probably exiting the year is getting closer to being in balance.

But we still think it's below where demand growth is.

I'm going to leave it at that.

Okay.

How about DRAM bit supply growth exiting the year, if there is, to your point, if there's DRAM CapEx being pushed?

DRAM bit growth also, I think, as we said, would under grow long-term demand this year.

Okay.

Maybe a longer term question.

With the share gains you had in Foundry and Logic this year and I think it was as high as 48% of the total revenue last quarter.

How should we think about your exposure between Foundry and Logic and Memory through the next cycle, I think, versus maybe last 5 years, remember it was averaging like close to 70% sales?

We've got nice trajectory relative to what's going on in Foundry and Logic, and we had a really strong December quarter, record level on Foundry.

I expect our share of WFE in both Foundry and Logic to continue to be very strong.

I'm not going to pinpoint exactly what it is versus Memory because we're very strong in Memory as well.

I feel really good about where we're positioned.

I'll just put it that way.

I mean, I think the simplest way to think about it as a company, our objective is to improve our share, revenue share of WFE across all segments.

All right.

I'd like to turn it back over to our speakers for any closing remarks.

We just want to thank everyone for joining our call today.

We appreciate it.

Thank you, ladies and gentlemen.

This concludes today's teleconference, and you may now disconnect.

Please enjoy the rest of your day.