科林研發 (LRCX) 2018 Q1 法說會逐字稿

Good day, and welcome to the Lam Research September 2017 conference call.

At this time, I would like to turn the conference over to Satya Kumar, Vice President of Investor Relations.

Please go ahead.

Yes.

Thank you, and good afternoon, everyone.

Welcome to the Lam Research quarterly earnings conference call.

With me today are Martin Anstice, President and Chief Executive Officer; and Doug Bettinger, Executive Vice President and Chief Financial Officer.

During today's call, we will share our outlook on the business environment, review our financial results for the September 2017 quarter and our outlook for the December 2017 quarter.

The press release detailing our financial results was distributed a little after 1:00 p.m.

Pacific Time this afternoon.

It 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 factor disclosures of our SEC public filings.

Please see the accompanying slide 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 the 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.

And as always, we ask that you limit yourself to one question per firm and a brief follow-up so that we can accommodate as many questions as possible.

As a reminder, the replay of this call will be available later this afternoon on our website.

With that, let me hand the call over to Martin.

Thank you, Satya, and thank you all for joining us this afternoon.

Lam delivered another outstanding quarter, with September shipments, revenue and gross margins above the midpoint of our guidance range and operating margins and EPS above the high end of our guidance range.

I'd like to thank our employees, our suppliers and our customers for their active support and confidence in Lam.

With enhanced opportunity, we have enhanced responsibility to contribute to the success of our customers and, in that context, we are inspired to achieve our full potential.

Calendar 2017 appears destined to be another terrific year of outperformance for Lam, a year that materially exceeded the expectations we had communicated at the beginning of the year for levels of demand-driven customer CapEx and our business as a direct result.

The midpoint of our December quarter shipment guidance today reflects growth of approximately 50% year-over-year, nearly double the consensus growth expectations for WFE in 2017 and putting us on track to gain over 2.5 percentage points of shipped share of WFE in the calendar year.

Based on our internal analytics, we continue to expect a strong industry spending environment once again in 2018 and essentially align to stated industry consensus at this time.

We believe there are enduring drivers behind the sustainability of industry spending levels and our outperformance.

Central to the argument on spending sustainability is the growing importance of data and the fundamental role that the road map of silicon is playing.

We are entering a new innovation era.

Central to the position of Lam outperformance is the increasing scope of industry relevance and competitiveness in the Lam product portfolio today.

The generation, transmission, storage and analysis of increasing quantities of data to create artificial intelligence has the potential to transform every segment of the economy, and we believe Lam Research is positioned comprehensively to capitalize on that trend.

The last decade saw the transition from a PC-centric to a mobile-centric world.

We expect the next decade is defined by a transition to an AI-centric world.

Mobile devices have enabled an incredible platform of almost 4 billion global smartphone users, which allows for the rapid delivery and scaling of new AI-enabled services on a growing array of IoT devices and systems.

Together, these trends are resulting in semiconductor industry revenue growth rates accelerating from an average of just over 2% in the last 6 years to over 15% in 2017.

Even more important than revenue and profit growth, a dramatic diversification in end-use demand bodes well for our future.

The critical role of data in this market transition is particularly positive for memory and storage semiconductor demand in the cloud and also IoT.

This is strength and opportunity both for Lam.

Cloud computing is a foundation for extremely data-intensive applications, which is driving strong demand for bit growth in server DRAM and NAND of over 30% and 50%, respectively.

For IoT devices, we are in early stages of adoption of exciting new technologies like 3D cameras, AR and VR, and broad automation trends.

This is driving over 40% demand bit growth in these NAND markets.

Combined, these demand trends and disciplined investment by our customers are driving record revenues in the memory semiconductor segment, which is on track to grow approximately 60% to about $130 billion in 2017, at strong levels of profitability.

We see our customers increase their spending at sustainable and rational levels to respond to these long-term trends, with memory capital intensity slightly lower in 2017 when compared to the last 2 years.

As we have stated previously, the products and services portfolio of Lam Research is unmatched, we believe, in its broad fit to the primary technology inflections of the industry.

Our focus is to further increase this position of leadership and strategic relevance through a commitment to disruptive and customer-enabling technologies.

During our presentation at the recent Flash Memory Summit, we offered perspective that lithographic shrinks alone are no longer able to provide the device performance and cost improvements that are required to drive these megatrends in end demand.

Logic devices require new solutions to deal with dark silicon and Dennard scaling limitations.

DRAM devices require new solutions to deal with signal-to-noise scaling issues, and NAND devices require new solutions to deal with cell-to-cell interference and reliability.

Technology inflections such as vertical scaling, multi-patterning, advanced transistor and interconnect technologies, and advanced packaging are all critical in addressing these challenges to deliver necessary device performance improvements.

These inflections are fundamentally enabled by etch, deposition and clean technologies, resulting in our served markets growing significantly faster than WFE again this year.

We believe this trend is sustainable.

Anticipating these and future inflections, we grew our R&D investments by nearly 60% in the last 4 years, reporting during the same period 140% growth in revenues.

We continue to release significant capability enhancements across our portfolio, and we have the strongest commitment in our history to new concepts, technologies, products and services from R&D.

These investments continue to pay off with significantly more wins and successful defenses than not this year.

Calendar '17 is a stronger market share performance year for Lam than calendar '16.

Although with SAM expansion, this tells less than half of our outperformance story.

In addition to maintaining our leadership position in vertical scaling in 3D NAND, we had several new wins during the quarter across the spectrum of DRAM, NAND and logic devices, and across the process flow in front-end transistor, middle of line as well as advanced interconnects.

In dielectric etch, we extended our momentum in critical transistor contact applications by delivering differentiated atomic layer etching solutions that resulted in a new PTOR position at a second leading foundry.

We continued to maintain strong conductor etch positions with increasing adoption of advanced technology and productivity options on our Kiyo conductor etch products at leading-edge foundry and memory makers.

In deposition, we won PTOR position for advanced cobalt interconnect at a major foundry.

We had multiple wins for new applications in atomic layer deposition for both dielectric and metal film applications for non-volatile memory devices.

To further expand our value proposition, we recently acquired Coventor, a market leader in 3D modeling and simulation in our industry.

The addition of Coventor supports Lam's vision that advanced process and equipment control capabilities further enhance our competitiveness, our time-to-market, and enable us to deliver significant incremental value to our customers.

Already, the potential value proposition of harnessing Coventor's modeling expertise with Lam's process and physical characterization capabilities to deliver more simulation and virtual fabrication for the development of next-generation devices has been reinforced by the excitement and substance of discussions within our engineering and collaboration partner communities.

The growing strength of our systems business is further enhanced by our strong customer support business group, which continues to deliver advanced productivity and technology improvements to the installed base and actively promotes new product offerings in support of MEMS, automotive, power management and IoT device innovation.

This business, which represents approximately 1/4 of our company revenues currently, is on track to deliver another record year of growth at a rate significantly faster than the rate of installed base growth this year, a byproduct of an enhanced portfolio and the strategic fit of the business with our customers' needs.

Combined, these strategies have extended our differentiation and delivered consistent outperformance for Lam.

We are on track to grow significantly faster than WFE and our served markets this year.

Over the last 4 years, we have grown our shipments by 25% CAGR to approximately $10 billion this year at approximately twice the growth rates of both WFE and our largest peers and competitors combined in the equipment industry.

As we look ahead to 2018, we expect to build on this momentum with over 90% of customer decisions on PTOR selection decisions already made and on the back of what already looks to be another strong year for industry CapEx spending.

In conclusion, we are encouraged by the scale and sustainability of Lam opportunity, and we continue to invest in extending our differentiation and strategic relevance to our customers.

We believe these dynamics position us well to continue to outperform.

With that, let me turn the call over to Doug.

Okay, great.

Thank you, Martin.

Good afternoon, everyone, and thank you for joining us today.

We posted another solid quarter, delivering results above the midpoint of guidance for all financial metrics and extending our positive momentum heading into the end of the calendar year.

Revenue, operating income, cash from operations and earnings per share were again at record levels in the September quarter and, overall, the company continues to perform well against our financial and operational objectives.

Shipments for the quarter came in at $2,382,000,000, down 6% from the record-high level we delivered last quarter and just above the midpoint of our guided range.

The combined Memory segment made up 66% of system shipments, which was down from 73% in the prior quarter.

Non-volatile memory shipments represented 49% of the system shipments, which was down from 59% in the June quarter.

DRAM shipments ticked up in the quarter, coming in at 17% compared to 14% last quarter.

Our customers continued to invest in technology migrations at a rational and sustainable level to support demand growth in the mobile and server markets.

DRAM pricing continues to be strong, supporting the increased investments.

System shipments into the Foundry segment made up 21% of the total, down slightly from 22% in the June quarter.

Foundry spend was biased toward investments to execute the 10-nanometer ramp as well as 7-nanometer pilot projects.

Logic and Other shipments increased nicely in the September quarter, accounting for 13% of system shipments compared to 5% in the previous quarter.

This record level of Logic and Other shipments is the highest percentage for us since the June 2015 quarter and is the highest in absolute dollars in the history of Lam.

This increase in shipments was driven by the ramp of our application design wins at 10-nanometer as well as broadly higher demand for applications in image sensors as well as power management devices.

September quarter revenue came in at $2,478,000,000, which was up 6% from the June quarter.

Gross margin for the period came in at 47.2%, which was an improvement of 70 basis points and again above the midpoint of our guidance.

Our gross margin performance, as always, is determined by several factors such as overall business volumes, product mix and customer concentration, and you should expect to see variability quarter-to-quarter.

Operating expenses were essentially flat at $438 million compared to $440 million in the June quarter.

Expenses came in a bit lower than the implied guidance as we adjusted our outlook for variable compensation for the year.

We continue to focus our spending toward innovative R&D programs that are enabling our customers' roadmaps.

Operating expenses declined as a percentage of revenue to 17.7%, which was down from 18.8% in the prior quarter.

Operating income in the September quarter was $733 million, up 13% from $650 million last quarter.

Operating margin increased to 29.6%, which was up from 27.7% in the June quarter and again above the guidance range.

Operating profitability was very strong in the quarter, driven by higher revenue as well as the improvement in gross margin.

The tax rate for the quarter was 14% compared to 13% last quarter.

For the December quarter, I'd be modeling a rate in the low to middle teens.

Based on a non-GAAP share count of approximately 181 million shares, earnings per share for the September quarter came in at $3.46, which was above the guided range.

The share count includes dilution from both the 2018 and the 2041 convertible notes.

The net dilutive impact from the notes is approximately 17 million shares on a non-GAAP basis.

As we noted during last quarter's earnings call, with the increase in our stock price, we have received requests for early conversions of our 2018 and 2041 convertible notes.

Conversions that settled in the September quarter totaled $302 million, of which $209 million related to the 2018 convertible bond and $93 million to the 2041 bond.

Dilution schedules, including updated notional amounts for the 2018 and 2041 convertible notes, are available on our Investor Relations website to help you with your modeling.

We continue to execute on the capital return program we announced in November last year, spending approximately $230 million during the quarter in share repurchases as well as dividends and over $1 billion in the first 3 quarters of the calendar year.

As of the end of the September quarter, we have completed approximately 88% of our current $1 billion share repurchase authorization, buying back a cumulative total of 6.3 million shares at an average share price of $139.17.

We paid out $0.45 per share in dividends or $73 million during the quarter.

Let me now move to the balance sheet.

We ended the quarter with cash and short-term investments, including restricted cash, of approximately $6.4 billion.

This was up from $6.3 billion in the June quarter.

Cash from operations was strong at $858 million, up from $729 million we generated in the June quarter.

Cash generation in the quarter was at a record level, enabling us to grow cash while funding the convertible note redemptions as well as our capital return programs.

DSO came in at 56 days, down from 65 days in June.

Inventory turns came in at 4 compared to 4.1 in the prior quarter.

Deferred revenues were $938 million.

This number excludes $344 million in shipments to customers in Japan, which will revenue in future quarters.

Company noncash expenses include $42 million for equity comp, $39 million for amortization and $40 million for depreciation.

We incurred $60 million for capital expenditures in the quarter.

We exited the quarter with 9,800 regular full-time employees.

Headcount additions came primarily in the field and factory, with further additions in the technology areas.

Let me now turn to our non-GAAP guidance for the December quarter.

We expect record shipments of $2,600,000,000 plus or minus $100 million.

I'd just mention that 30% of the system shipments in December are to new greenfield fabs.

We expect record revenue of $2,550,000,000, again plus or minus $100 million.

We expect gross margin of 47.5%, plus or minus 1 percentage point.

We're forecasting operating margins of 30%, plus or minus 1 percentage point.

And finally, we're forecasting earnings per share of $3.65, plus or minus $0.12 based on a share count of approximately 182 million shares.

So in summary, we're very pleased with our performance, delivering another quarter of solid operational execution, and we're on track for record financial results for calendar year 2017.

And while things could always change, as we look into '18, we continue to see what appears to be a strong year.

Our current bias is that shipments in the first half of 2018 will be stronger than the second half of 2017.

Shipment momentum heading into the March quarter seems meaningfully stronger to us than it is in December.

Before transitioning to the Q&A part of the call, I wanted to share some information about the timing of our Investor Day.

We're now planning to host this event in the March quarter timeframe on an ongoing basis.

This timing aligns well for us with executive availability, with the timing of our planning cycle and with what we think will be good investor attendance.

We're excited about the sustainable outperformance opportunity we have ahead of us, and we look forward to sharing with you a comprehensive update on the company's plans and objectives at that time.

Operator, that concludes my prepared remarks.

Please open up the call to questions.

(Operator Instructions) And we'll take our first question from Farhan Ahmad from Credit Suisse.

My first question is on operating margin.

You're guiding December quarter to 30%, which is meaningfully higher than the long-term target model that you shared last time, which was at 28%.

Can you just give us a sense of what's a sustainable level of margin for the business?

And assuming the business can grow from here, do you -- should we think that there's upward bias to the operating margin?

Yes, Farhan.

I mean, I'll give you a new -- formally give you a new model when we get to the Analyst Day.

I mean, the way I'd be thinking about it, when I look at gross margin right now, it's probably not going to get much higher than it is right now.

In fact, there's really a bias to be a little bit lower than it is.

And then it all comes down to what do you think the revenue levels of the company are.

We clearly have delivered, in the last several years, leverage to spending.

And as business volumes increase, we would continue to deliver that leverage.

I'm not going to quantify it for you, but that would be the way I'd think about it.

And then stay tuned.

We'll give you a formal update when we get to the Analyst Day.

And just to supplement with just maybe some color relative to our thinking and that kind of guides how we run the company, I mean, we run the company with a focus on absolute dollars of profitability, and our aspiration is to increase the value of our company and outperform relative to increasing the value of our company.

And so the profitability that we care about most is a dollar, not a percentage, and the execution of our business on a day to day basis requires a very careful balance.

And certainly, the guidance that I provide to the company is intended to claim our fair share of opportunity, our fair share of profitability without compromising the success of our customers because without the success of our customers, our industry is going to struggle.

So it is as much a choice as anything else, and the choice is defined by how we optimize performance.

And we optimize performance according to dollars of profit and dollars of profit growth.

And a quick follow-up.

In terms of the segment outlook for DRAM and NAND, for the December quarter and first half of next year, do you -- how are you looking at the segment?

Just quantitative guidance on which segments are stronger or weaker.

Yes.

So I would say -- I mean, Doug already provided a little bit of color relative to the first half of next year.

Just to remind everybody, he said we bias first half of '18 shipments stronger than the second half of '17.

As he said, shipments momentum meaningfully stronger in March over December as best we see it today.

What I'll add is at a WFE level, our expectation is probably in the mid- to high-single digit range year-over-year, and we would expect more than 85% of the growth in WFE year-over-year to come from memory.

And we'll take our next question from C.J. Muse with Evercore ISI.

I guess, first question, Martin, a longer-term question on the memory side.

Hyperscale guys are asking for longer-term supply agreements from chip makers, which is clearly a strategic change and speaks to how much more important memory is and how much more strategic it's becoming.

And so my question to you is, is that changing your relationship at all, your visibility, how you partner with chip makers, given this kind of change?

The simple answer to that question is yes.

I do think we have more visibility.

We are clearly more relevant to the success of our customers today by virtue of the scope of the portfolio and the relevance of the portfolio to the technology inflections that are -- that we've been talking about for some years.

We still have to work very hard to do what we do, but I would say relatively speaking, our visibility and our engagement is stronger, consistent with the strengthening of strategic relevance of Lam to the industry.

Okay, I guess, as a follow-up, Doug, you talked about OpEx coming in a little lighter in the September quarter related to, I guess, bonus plans.

And so curious, does that suggest perhaps a delay in terms of timing of clean room coming online?

Or other factors like that have pushed shipments into the first half of calendar '18?

No, not really, C.J. I mean, the 2 are somewhat independent.

You always -- I mean, when you get to the second half of the year, you start truing things up and your perspective could be a little bit different.

It's more got to do with what's going on internally than clean room coming online.

There's a lot of clean room space coming online as we speak, and I think for the most part, it's consistent with what we expected to see, what the industry has communicated it's doing.

There really isn't any difference there.

And just to add, I would say it's almost completely independent because our bonus plans accrue on revenues and profits on revenues, not on shipments.

And we'll take our next question from Krish Sankar with Bank of America Merrill Lynch.

I have 2 of them.

One is, Martin or Doug, if I look at your revenue as a percentage of WFE, it's been running around 22%, 23%.

I'm just trying -- I'm curious, like to get to like a 24% or a higher percentage, what needs to happen?

Is it just a SAM expansion or do you think share gains are going to be more critical to get a higher percentage of WFE?

And I have a follow up.

I think they're both relevant.

We've articulated for a number of years the opportunity for SAM expansion, beyond the boundaries that we're currently reporting.

So the technology inflections are far from done and they're relevant for many, many years yet to come.

And as you also know from prior disclosure and our reported financial results, the market share momentum in the company is targeted to be positive and is positive.

This is a stronger market share year for the company than was true last year, and that sits in the context of the long-term models which you're pretty familiar with.

So whether it's a 50-50 conversation or biased slightly to the SAM expansion conversation, there's certainly some exciting business opportunity and upside for the company in the years to come.

Krish, this is a Doug.

What I get excited about is just looking at the growth in the installed base and what's going on with our installed base business, which, as you know, isn't part of WFE necessarily.

It just grows along with the installed base, which is, as Martin said in his scripted remarks, it's doing really, really nicely this year.

Got it, got it.

That kind of leads into my second question or the follow up, which is, can you talk a little bit about your services business?

How much is it as a percentage of revenue?

And what is the margin structure there like?

And what do you think the CAGR for that is going to be?

Yes, I mean, Krish, what we've talked about is we've got objectives to grow it faster than the installed base, and it absolutely is doing that over the last several years.

I get really excited about what's going on there.

Profitability isn't all that different than the rest of the company.

When we look at it in aggregate, the gross margins may be a little bit less than selling new equipment, but the level of investment isn't anywhere near what we need to invest in the new equipment side, so the operating income and the cash generation is very attractive in this part of the business.

We'll take our next question from Joe Moore with Morgan Stanley.

Wondering if you can talk a little bit about the NAND market in the next few quarters and how do you see the spending shifting between sort of conversion of planar NAND to 3D.

How much do you think now is shifting over towards 3D scaling and what happens to your -- I assume your share of the market potentially goes up as you move to more of a 3D scaling environment.

Just some more color on that will be good.

Yes, I mean, the construct of this year's spending by the customer in the NAND space is reasonably evenly balanced across the kind of 3 implementation scenarios, right?

So you've got kind of the addition of new 3D NAND capacity, you've got a conversion from 2D to 3D, and then you've got this kind of vertical scaling of some 3D capacity from, let's say, generation 2 to generation 3. And they're all represented.

The smaller of the 3 is the 3D NAND scaling segment.

As you've heard us characterize a couple of times now, we are actually almost agnostic to the path of the customer between the addition of new capacity and the conversion from 2D to 3D.

It's more or less the same-sized opportunity for the company, and I think that's unique in the industry as a byproduct of the process flow and the position of the company's products to support that inflection.

As the 3D capacity becomes a bigger proportion, and still even by the end of this year, I think it may only be about half of the installed base capacity, and that half is in various forms.

Not all of it is latest generation, highest layer count kind of 3D.

The opportunity for 3D NAND vertical scaling investment is going to grow over time, and that's the most efficient path for our customers.

It's also the path which is most biased to etch and deposition, so the segment concentration of a vertical scaling is even stronger for us as a relative opportunity to the rest of the industry.

So at the end of the day, the customers will make the right choices.

They'll optimize fabs and optimize economics, and that's a pretty dynamic space.

So it's one that's pretty difficult to predict, and we actually don't need to for the reasons that I've just summarized.

And we'll take our next question from Harlan Sur with JPMorgan.

My question is kind of similar to the last one, except it's more DRAM.

The industry in DRAM has done extremely well, tight supply, very strong demand environment.

All eyes are starting to turn to next year when the capital spending trajectory in DRAM is going higher.

And the question from investors again is, are the DRAM competitors going to remain disciplined on a capacity expansion perspective?

And you guys had talked previously about 2017 maybe kind of 1% to 3% increase in total DRAM industry capacity, pretty disciplined.

So as you guys look into your forward pipeline, talk with your customers, how do you view total DRAM capacity expansion next year?

Are you guys continuing to see sort of focus on profitability?

Yes, I would say intensely so.

It's a consolidated industry and everybody fights for their share of economics.

And I think the evidence of the last several years is reinforcing of everything that was embedded in your question.

To the best of our ability to model, the capacity for DRAM at the end of '17 is similar to the capacity at the end of '15, and that's kind of wafer starts out statement.

And that's a statement of slight growth after the contraction year in calendar '16.

What's maybe a little bit different today than was true 3 months ago is the proportion of the installed base that is converting to the 20-nanometer technology node is actually progressing a little bit faster than the assumptions we had 3 months ago.

So if you look at the construct of DRAM capacity at the end of this year, between 1x capacity, 20-series product, and then more than 20-nanometer, we would guess probably 25%, 55% and 20%.

It's for sure incredibly disciplined.

And the balance of conversion is obviously very high, so it's a very efficient investment for our customers.

And to the extent that any capacity gets added, it will get added as best we can tell from legitimate demand drivers.

And we have the same bit growth assumptions as everybody else, in the kind of low 20% range for DRAM.

And it will come from the relative transitions and density per wafer out.

And as I think everybody appreciates, the density-per-wafer-out improvement at the latter stages of the DRAM road map are a little less than was true, let's say, 5 years ago in a transition.

So that's part of the story as well.

But I mean, as a basic premise, we absolutely expect discipline to continue, and we can't guarantee that.

We don't have a crystal ball, but that's the planning assumption and that's the behavior as best we can tell from our customers.

Thanks for the insights there.

You guys are almost completed with the $1 billion share repurchase program.

I guess you'll be done with that by the year end.

Free cash flow generation, sitting kind of in the high 20% range.

So does the team wait for potential tax reform?

Or can you take on some debt to fund sort of the next tranche of your repurchase activity?

And then just roughly, what is the additional debt capacity that the team can take on while maintaining their investment-grade status?

Yes, Harlan, I'm not ready to give you a definitive update on that right now.

Obviously, we're paying close attention to what's going on relative to the tax discussion in Washington.

Your guess is as good as mine in terms of what might happen as well as the timing.

We certainly have an ability to take on more debt should we choose to do that, and that's something I'll be talking to the board -- Martin and I will be talking to the board about.

When I've got something to the update you on, I'll share it with you, but I don't have that right now.

And we'll take our next question from Toshi Hari with Goldman Sachs.

The uptick you saw in Logic shipments in the quarter was a nice surprise.

Doug, you talked about the ramp of 10-nanometer, I think, image sensors, and if I caught you correctly, power management.

I guess the question is, is this level of shipments on a quarterly basis the new normal given that 10-nanometer is ramping?

Or was this quarter sort of a one-off in terms of shipments?

Things will always ebb and flow, as you know.

I mean, we've been talking about the applications wins in 10-nanometer for a while now.

Obviously, as that ramps into production, that will continue.

Image sensors, it comes and goes, and likewise, with the power management stuff.

But I expect the trajectory to continue here.

We're doing real well.

We've been waiting for this for a while, and the ramp is beginning to happen.

Yes, 2 things to add.

One of them is it is more the new norm than not, and I hope it really isn't a surprise.

We've been talking about it for long enough, so it's really nice to actually finally answer the question for everybody and demonstrate the success of our engagements in Logic and Foundry in etch and deposition.

Both have taken a lot of hard work from a lot of people, and it's really nice to see a further increase in the balanced engagements of the company in the industry.

We're fortunate to be in a very strong position around technology inflections.

We're fortunate to have the strength of the company in memory at a time when memory is clearly an increasingly relevant and critical part of the road map of success for all of our customers.

And this is icing on the cake at some level, so a really nice result.

Okay, great.

And then I had a follow-up on the Coventor acquisition in the quarter.

Martin, you went through the acquisition in your prepared remarks, but if you can kind of reiterate what the rationale was, what you'd gain from this from a technology perspective and how it could potentially impact your incumbent businesses, that would be great.

Yes, thank you for the question.

In many respects, it's a repeat of much of the rationale we talked about in the context of prior M&A focus for process control in the company.

And in some respects, this is consistent with the messaging on data and AI having the potential to change the world, and that includes being relevant in our industry and the industry of our customers as well.

So we believe that we have value to deliver to our customers and value to deliver to our shareholders through an enhancement of process control capabilities, and we are generally collaborative in execution today as strategies.

But we have an opportunity to acquire and fully harness the capability of Coventor in the mix.

It brings expertise and competency into the company.

That, for sure, supplements and strengthens our pre-existing computer science algorithm analytics capability, modeling capability.

And it's a really nice partnership, I would say, for a virtual modeling business and a true process and a hardware characterization business.

So the combination of those 2 things, we believe, has an opportunity to grow faster than was true for a standalone business, the pre-existing products and services of the Coventor business.

And it has an opportunity to provide us vehicles to introduce new products and services in that space.

And last but not least, it's an opportunity to increase the competitive differentiation of the etch and the deposition and the clean portfolio through enhanced control and faster cycle time in development and better predictability and opening up process windows and so on and so forth.

So it's part of what we've been talking about for some time.

It was a nice opportunity to take custody over a valuable asset, and we're super excited about the opportunity in front of us.

And I think that optimism is shared by our customers.

And we'll take our next question from Patrick Ho with Stifel.

This Brian Chin calling in for Patrick.

Looking at your shipment distribution for the September quarter, China is only your fourth-largest geography, but shipments were down sequentially by a decent amount in September.

Just curious what it's attributed to and whether you can comment on how the pipeline or trajectory in China looks moving through next year.

Yes, I'd say that what happened in the quarter relative to China is completely unrepresentative of any broader trend.

So certainly, don't read anything into that other than it's the same for any customer in any region.

Investment is a byproduct of discrete projects in any one quarter, and that's always going to have an ebb and a flow.

If we look at kind of new fab projects, maybe I'd kind of give you an update from the disclosure of prior periods.

From now through the end of calendar '18, we're tracking, I think, 18 fab -- greenfield fabs today, and a greenfield fab can be a second floor on a double-stacked fab, just to be fully disclosed on that.

12 of the 18 are in China.

And about 8 of them, I think, are classified as domestic China, and 4 of them are kind of global China.

And when you look at the distribution of investment that we're expecting next year between global and domestic players in China, it's actually close to 50-50.

So equipment purchases, wafer starts, capacity additions in China, domestic community and global community, looks 50-50-ish at this point in time.

That could change, but that's the headline today.

So we still say the same thing today that we said before.

China, for global companies and domestic community, is a relevant part of our future.

We believe that execution is as disciplined an execution in China as it is anywhere else in the world and hope that, that continues.

And we're very focused.

We're investing heavily, and our position in memory and logic both in China is very strong.

So hopefully, that is helpful to you.

Very helpful.

And just one quick follow-up question on your comment about shipments being very healthy through the March quarter.

Just wondering, are there any discernible shifts you're seeing in spending within memory between NAND and DRAM over that horizon?

Too early for us to comment at this point.

I think both segments are disciplined.

Both segments have overlapping demand drivers, but they have unique demand drivers, and we'll kind of hold off until January to get specific.

What I should say, by the way, going back to the China question is when I gave my reference to 8 domestic, I was referring to memory and logic both, and both segments are represented in the plans of our customers.

We'll take our next question from Sidney Ho with Deutsche Bank.

A couple of months ago, you guys suggested that the WFE opportunity for NAND would be $70 billion over the next 5 years, and that's assuming 40% bit growth.

Now that one of the suppliers actually suggested that the market will grow 50% next year, and it seems like you agree with that, does that make your forecast too conservative now?

And if it does become 50% per year going forward instead of 40%, how do you think about the NAND WFE?

How will that change?

There's always announcements, positive and negative, there always will be.

We have no basis to update a $70 billion reference at this time.

Same message today as was true at Flash Memory Summit.

Okay, maybe switching over to DRAM, as a follow-up to an earlier question.

There's quite a bit of increase in DRAM CapEx by your customers, and it seems like there's a consensus view that maybe CapEx per wafer is going higher.

Obviously, cost per bit is still going down.

If it is getting more expensive per wafer, how do you think customers justify the high cost?

And are they either eating into the cost, granted they are making a lot of money these days?

Or do you think that's being passed along to the end users?

I think you might have answered your own question.

They're making more money than they ever have done.

They're making more money than I do.

So I would say as a basic headline, it looks like they've rationalized an ability to make an investment and get paid for it.

And I think going back to the questions of discipline, the consolidated industry and a disciplined capacity addition is kind of part of that story.

I think your question is very good because it is appropriate that everybody recognizes that the cost of density is higher, and that's one of the reasons why the next-generation memory conversation shows up in the industry today when it didn't do several years ago.

But what I want to remind everybody is that there is tons of life left in the DRAM device.

There are multiple technology nodes ahead of us in the road map of our customers today, and there are many years of investments still left in that product and technology.

And I'm sure a year from now, we'll see something even better than that because the industry has a tendency to innovate its way to some pretty outstanding solutions.

And your next question is from Weston Twigg from KeyBanc Capital Markets.

I wanted to follow-up on one earlier about the record shipments in IDM and Logic.

Just wondering, when you mentioned your growth in 2018, I think you mentioned that 85% of the growth looks like it's coming from memory.

Why not a little bit more growth coming from the Logic and/or maybe the Foundry side given the 10- and 7-nanometer ramps?

Okay, I'm glad you asked that because I maybe wasn't quite as clear as I should have been.

When I referenced 85%, I was trying to describe our outlook for calendar '18 WFE compared to calendar '17 WFE.

So our expectation is calendar '18 is higher than '17.

I introduced a mid- to high single-digit reference for that, and I provided a perspective that said we expect 85% of the growth in WFE year-over-year to be memory based.

I apologize for the confusion.

I think I understood it properly.

I'm just wondering with, say, your 10-nanometer ramp at one of the big Logic customers and the 7-nanometer activity at the foundries why you wouldn't see more of the growth coming from those segments.

And I don't know if your particular exposure to memory is just higher or it's a broader reference.

Well, we do the best we can to take every piece of data that's available to us in terms of demand -- device demand, the implications of that device demand on various kind of manufacturing segments to the best of our abilities.

We look at kind of reuse capabilities, which is relevant for every segment of the industry.

We take into account public and private disclosure of the companies -- of our customers, and we tell you to the best of our abilities what we think our outlook is.

So the 85% headline is the best communication we can give you at this point.

Okay.

And then as my follow-up, just on the guidance, you've had several quarters of beats and raises, and I know the market is very strong.

Are you guiding too conservatively?

Do you think that -- I guess, what are you missing when you're guiding over the last few quarters?

Why is demand out-running your expectations a little bit?

No, it's not out-running it by an enormous amount, Weston.

And we've done a little better in the last couple of quarters in terms of gross margin.

And I just guided you to the highest gross margin, I think, in the last decade of the company, so it's not conservative by nature.

We've done a little better over the last several quarters for sure.

But I'd just tell you, we tell you what we see as we look into the quarter.

And maybe again just to supplement, it's actually not easy to give guidance, just to be clear.

I mean, this is a consolidated industry.

The investments of our customers are big.

Any one application is big.

Any one investment is big.

And so the potential to predict this at a quarterly level can actually be quite challenging.

And I would say our guidance is really good because consistently, we are guiding and performing within the range of our guidance.

So with due respect to your question, maybe we're not as precise as you'd like us to be, but I feel really good about the integrity of the guidance conversation and the performance of the company inside of the guidance range consistently.

We'll take our next question from Edwin Mok with Needham & Company.

So I guess, just a quick follow-up to Wes' question.

I think the general consensus is that China's spending will grow in '18.

We've lost you, Edwin.

The contribution from China.

Edwin, ask your question again.

Sorry, you broke up midway through there.

Sorry about that.

I'm trying to understand how much China is contributing to your growth on 2018 within your expectation of mid- to high single-digit growth.

Too early for us to disclose at this point.

We'll give you better color on next year.

I mean, it's a little bit early.

Normally at this juncture, we wouldn't be talking about '18.

We'll give you the normal color when we get to next quarter's earnings.

Okay.

All right.

I think that's fine.

So I guess, the follow-up question I have on EUV.

There's a lot of talks about EUV adopting 7-nanometer.

And I remember you guys have talked about developing some etch and deposition.

(technical difficulty)

to help with challenges like line edge roughness.

Is it possible you can give us some updates around that?

Where do you stand in that?

And do you think,,, that can potentially be a SAM expansion opportunity as EUVs are adopted?

I would say the basic message -- and I apologize again, we lost you mid-sentence, but the basic message of the company today around our opportunity in the context of an EUV implementation is almost identical to the scenarios we characterized in the investor and analyst meeting about a year ago.

So the insertion plans of our customers appear to be consistent, and the balance of spacer-based patterning implementations in the industry appears reasonably consistent with the assumptions we made before.

So I don't think we can give you anything better than we already did.

I apologize.

And we'll take our next question from Craig Ellis with B. Riley.

I just wanted to follow-up with some of the China commentary with a clarification question.

I think you mentioned that of the 18 greenfield fabs you're tracking, 12 are in China, and there's a split between domestic and global.

But even looking at the number of domestic fabs the company has cited, that could mean that either that exposure is with some of the smaller fabs that are trailing-edge logic or they may be with some of the larger fabs.

Can you clarify if you're participating with that number that you provided in some of the larger China fabs next year?

Do you see those coming on thereafter?

Yes is the answer to your question.

It includes -- the reference of 8 includes the bigger domestic guys and the smaller domestic guys.

It's memory and logic both.

It's DRAM and flash.

The flash investment is obviously materially greater than the DRAM investment in the context of market opportunity, primarily.

But we are seeing the emergence, as we've talked about for some time, of investments by domestic customers in calendar '18, big and small, memory and logic.

And it's still a fairly gradual implementation, so there's no pendulum swing, there's no paradigm shift.

But there's a commencement of pilot-line investments and then transition out of pilot line into first phase, high-volume manufacturing.

But the discipline that we all hope we are seeing to the best of our abilities today.

And Craig, when we think about WFE and pilot investments that Martin was referencing, I mean, it's -- for the local guys, it's an incremental $1 billion or $2 billion in terms of total WFE.

A lot of the activity continues to be, as it is this year, the global multinationals doing projects in China.

Got it.

That's very helpful, guys.

The follow-up is just regarding the comments for first half '18 shipments to be up meaningfully versus the second half of '17.

And acknowledging, Martin, your point that the business is a hard-to-forecast business.

I'm just wondering if there's any color that you can give us on linearity as we think about the first half of 2018 for the business.

Yes, let me just clarify, Craig.

What we said is first half '18, higher than second half '17.

The meaningfully comment was relative to the March quarter, not necessarily the entire half.

And it's really too soon for us to call linearity into next year.

Again, that's something we typically would do at the end of the December quarter, and you can expect to hear from us then.

And we'll take our next question from Tom Diffely with D. A. Davidson.

Another question on the DRAM side of the market.

You talked about how wafer starts are roughly flat from 2015.

Do you expect them to go up over the next couple of years or mainly just going down to the sub 2 -- the 20-nanometer node?

The planning assumption we have today is up a little, but it really is a little.

And it will only go up a little if there's legitimate demand for devices that require that investment.

So there really is no headline in terms of capacity expansion in DRAM unless there's a dramatic change to the demand statement for the segment.

So it might go up and down a little bit, but it's basically a headline of technology conversion.

And the technology conversions, to the point I made before, are less effective gradually in terms of delivering density per wafer out than was true 5 years ago.

But they're still incrementally valuable, and our customers appear to be making lots of money manufacturing and selling those devices.

Okay.

So when you look at 80% to 85% of the growth from memory next year, is that mainly NAND then?

No, it's not, but I'm not going to give you specifics at this point in time.

I felt like I was super extravagant with my 85% memory disclosure, so apparently you have a never-ending appetite for more.

Unfortunately, I don't have a never-ending appetite to provide more, so if you'll just wait on that question until we get a little further into next year, we'd appreciate it.

And that does conclude our question-and-answer session for today.

I'd like to turn the conference back over to Satya for any additional or closing remarks.

Yes, thank you once again for joining us, and have a wonderful rest of the day.

Thank you.

Thanks, guys.

And once again, that does conclude today's presentation.

We thank you all for your participation, and you may now disconnect.