科林研發 (LRCX) 2018 Q3 法說會逐字稿

Good day, and welcome to the Lam Research Corporation March 2018 Conference Call.

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

Please go ahead, sir.

Yes, thank you, and good afternoon, everyone.

Welcome to Lam Research Quarterly Earnings Conference Call.

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

During today's call, we'll share our outlook on the business environment, review our financial results for the March 2018 quarter and our outlook for the June 2018 quarter.

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

Pacific Time this afternoon.

It can also be found in 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 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 the GAAP and non-GAAP results can be found in today's earnings press release.

This call is scheduled to last until 3 p.m.

Pacific Time.

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

Due to some technical difficulties, our presentation today will be audio only, and the slides accompanying today's presentation have been posted on our website separately from the webcast.

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

Again, thank you all for joining us today.

As reported, Lam delivered strong financial results for the March 2018 quarter with revenue, gross margin dollars and EPS all at record levels and above the midpoints of guidance.

In addition, we completed our first ever quarter with over $3 billion in shipments.

Including the June quarter outlook, Lam will deliver revenue growth on a year-over-year basis in 23 out of the last 24 quarters, which is a testament to both the successful scaling of our company and a multiyear environment of demand-led and more sustainable customer investments.

I would like to take this moment to thank our customers, our employees and our partners for the opportunity they provide us, for their trust and contributions made, without which the performance of Lam would not be possible.

I will keep my comments brief today given that we provided a comprehensive update on our vision, strategy and objectives at our recent Investor Day.

Cost-effective scaling of semiconductor device performance remains critical to addressing opportunities provided by the transition to the data economy and now is substantially more diverse in scope than traditional shrink.

In this context, our focus is on systematically strengthening the competitiveness and relevance of Lam's product portfolio to the success of our customers.

Broadly recognized, etch and deposition technologies are now foundational and permanent features in the current road map of vertical scaling, multipatterning, advanced packaging and advanced transistor architectures.

These inflections have resulted in substantial growth in our serviceable addressable market, our SAM, and reiterating a headline of the recent Investor Day, we are on track to grow our SAM to over 40% of wafer fabrication equipment spending by 2021.

As an acknowledged leader in etch and deposition markets, we have an opportunity to capture a greater proportion of our customer spending by delivering unit process excellence, leveraging our multiproduct capabilities and engaging in closer collaboration within the semi ecosystem.

We are making substantial, comprehensive and disciplined investments in R&D to fund innovation that is intended to extend the differentiation of our products and services portfolio, and we continue to prudently scale the infrastructure of the company to support our 2021 objectives.

Customer equipment selection decisions in the March 2018 quarter were consistent with expectations in the context of our long-term objectives.

Illustratively, we secured additional wins in dielectric etch, one for a high-aspect ratio DRAM application and another for NAND flash, both at leading memory manufacturers.

We also won a new nonvolatile memory application with an advanced etch capability, harnessing value from our unit process excellence and unmatched collaboration between Lam etch and deposition business units.

We see solid momentum with our deposition portfolio, and we're pleased recently to celebrate our 500th VECTOR Strata PECVD product shipments.

This platform continues to offer industry-leading productivity and film property control, that together creates customer enablements and competitive differentiation for Lam, creating value and enhancing codependency in mold stack deposition applications in 3D NAND, for example.

Our systems business is augmented by our installed base business, where we have been focused on enhancing the breadth and competitiveness of our products and services as a means of further enabling our customers' success.

Customer support business unit revenue growth continued to outperform that of our installed base growth in the March quarter, and we achieved key customer penetrations within our reliance and advanced services business.

At our recent Investor Day, we provided several metrics on long-term growth, not only to illustrate the value of this annuity-like business but more importantly, to emphasize the integral role it plays in value creation, value capture and increasing codependency with our customers.

Now turning to the macro and wafer fabrication equipment, WFE, spending environment.

Expectations for global economic growth remain strong and are healthy.

We believe that content growth remains a powerful multiyear driver of demand in the data economy.

Context for multiyear customer investments is not simply a byproduct of market efficiency from consolidation and disciplined operational execution through the semiconductor supply chain.

It is more fundamentally endorsed by the evergreen verticals of climate change, education, food and water, health care, security and transportation that together, define the opportunity for silicon-based artificial intelligence, AI technologies and applications and services innovation globally.

In the smartphone end markets, innovative data-intensive services will increasingly be deployed, leveraging AI and using a range of enabling artificial and virtual reality technologies.

The deployments of 5G networks will improve the quality of these services.

But to deliver the best user experience, the smartphones themselves will require a combination of higher screen resolution, faster refresh cycles and lower power.

This will drive the need for nearly 2 times the DRAM contents in smartphones relative to the current global average.

Additionally, density increases associated with higher layer counts in 3D NAND create the opportunity for terabyte-level storage in smartphones within the next few years.

For 2018, WFE investments by our customers remain prudent and demand-led.

Since our update in January, WFE continues to track up low double digits in 2018 compared to 2017, with slightly higher DRAM mix and slightly lower logic investments compared to our earlier baseline.

For NAND and nonvolatile memory broadly, the totality of our perspective would be that long-term demand drivers are compelling.

Customer investments in 2018 appear essentially equivalent to those of 2017, and considering the impact of customer-specific tool moving plans this year, spending, we expect, will be first half-biased.

For Lam, aggregating all customer segments, we currently anticipate a first half-second half shipments weighting at the low 50s, high 40s level in 2018.

We expect relatively balanced revenues through the year.

As always, this prospective can and will likely change as our customers seek ways to exploit their competitive advantage and market timing later in the year.

It seems that the perpetual bull-bear debate surrounds the sustainability of memory-based investments.

In that context, we think it is worthy to note that associated with the emergence of cloud and connectivity, the 7-year cumulative annual growth rate for memory WFE is 15%, 10x greater than the comparable foundry logic WFE CAGR.

Combined with the concentration to etch and deposition technologies, a strong memory business in the new data economy is a sustainable asset and not a liability from our perspective.

In addition, and perhaps still somewhat underappreciated is the strengthening of Lam business, the diversification of Lam business into foundry and logic.

In only the 24-month period ending December 2018, we expect Lam shipments in these segments to grow 2x faster than the pace of the underlying foundry logic WFE.

In short, we feel that the historic strength of Lam has never been more valuable.

Further, the historic relative weakness of Lam is becoming much stronger.

In closing, we target outperforming overall industry growth again in calendar year 2018, and we are focused on successfully executing to our near and long-term objectives, growing SAM and increasing market share with an enhanced product and services portfolio.

Strong March quarter performance is consistent with our expectations, and we remain pleased with the company's execution in financials in an overall healthy industry environment.

With that, I'll turn the call over to Doug.

Okay, great.

Thank you, Martin.

Good afternoon, everyone.

Thank you for joining us today.

We're very pleased with our results for the March quarter, delivering record levels in shipments, revenue and operating income dollars.

Each of these items grew double-digit percentages sequentially from the December quarter.

Execution from the company continues to remain very strong.

Shipments for the quarter came in at $3.135 billion, which is up 19% sequentially and within the guided range.

Shipments were up 30% year-over-year.

Timing of shipments to certain new customer projects was a primary factor in our shipment profile coming in slightly below the midpoint.

As we expected, memory shipments continued to grow in the quarter, with the combined memory segment making up 84% of total system shipments, and that compares to 77% in the previous quarter.

Our overall nonvolatile memory shipments remain very strong, representing approximately 57% of system shipments compared to 53% last quarter.

DRAM shipments represented 27% of system shipments, which is up from 24% in prior quarter.

The NAND and DRAM markets continue to benefit from density growth as we transition to the new data-enabled economy.

The foundry segment was down, accounting for 10% of system shipments relative to 15% of system shipments in December.

The logic and other segment contributed 6% of those system shipments compared to 8% in the prior quarter.

And we would point out, we expect that both foundry and logic shipments will be stronger as we go through the remainder of this year.

We delivered record revenue of $2.892 billion in the March quarter, which was an increase of 12% from December and above the midpoint of our guidance.

Gross margin for the period came in at 46.8%, which was down 80 basis points sequentially but towards the higher end of our guided range.

And as we shared before, our actual gross margins are a function of several factors such as business volumes, product mix and customer concentration.

And we expect to see variability on a quarter-to-quarter basis.

Operating expenses in the quarter grew to $486 million but decreased 60 basis points on a percentage basis to 16.8% of revenues.

On a dollar basis, R&D spending and SG&A both increased sequentially, and we continue to have approximately 63% of our spending allocated to R&D.

Investments in R&D are fundamental to driving long-term value creation for all of our stakeholders.

Within the S&P 500 sectors, the semiconductor and equipment group ranked the highest in the percentage of R&D spend on average for the last several years.

A strong R&D investment strategy is central to maintaining our technology leadership as we position the company to benefit from the multiple technology inflections and to deliver the growth that we highlighted during our Investor Day in March.

Operating income in the March quarter came in at a record level of $867 million, which was up over 11% from the prior quarter.

Operating margin came in at the high end of our guidance at 30% due to the stronger gross margin performance and slightly better operating expenses than we expected.

Excluded from non-GAAP earnings are approximately $47 million in losses from the sale of investments in anticipation of cash repatriation under the recent U.S. tax reform.

We are essentially liquidating our fixed income portfolio to enable the cash to come to the United States.

Our tax rate for the quarter was 1%.

We expect the tax rate to be in the mid single-digit percentage range for the first half of calendar 2018, with high single-digit percentage for the June quarter.

Now I'll just remind you, in the longer run, a tax rate in the middle teens remains the right level for you to include in your models.

Based on a share count of approximately 178 million shares, earnings per share for the March quarter were $4.79, above the high end of our guidance.

The primary drivers of this upside versus our guidance were higher revenue, higher profitability, lower taxes and a lower share count.

The share count includes dilution from the 2018 and 2041 convertible notes, with the total dilutive impact being approximately 13 million shares on a non-GAAP basis.

Total conversions that settled in the March quarter were $228 million, with $193 million related to the 2041 bond and the remainder related to the 2018 convertible note.

The dilution schedules for the 2018 and 2041 convertible notes are available on our Investor Relations website for your reference.

And just a reminder, at our Analyst Day in March, we announced the plan to return at least 50% of our free cash flow to stockholders over the next 5 years.

This included a plan to increase quarterly dividends by 120% to $1.10 and an additional $2 billion share repurchase authorization for a total of $4 billion authorized since November of last year.

At the end of the quarter, we had completed approximately 25% of the current $4 billion share repurchase authorization.

This was primarily executed through an accelerated share repurchase program that is still ongoing.

We're planning to complete our authorization over the next 12 to 18 months in tandem with the expected timing of our cash repatriation.

During the quarter, we paid roughly $80 million in dividends.

So now, let me switch to the balance sheet.

Cash and short-term investments, including restricted cash, increased in the quarter to $6.7 billion compared to $6 billion at the end of the December quarter.

Approximately 87% of this total cash is still offshore.

We do expect to be able to move some portion of the offshore cash to the United States during the June quarter.

Cash from operations was slightly over $1 billion, up from $29 million in December.

Cash generation was partially offset by our capital return programs and capital expenditures.

Days sales outstanding decreased by 14 days to 66 days.

You may recall me talking last quarter about the timing of certain collections falling just outside of the December quarter.

All of those were collected in the March quarter.

Inventory turns remain roughly consistent with the prior quarter at 3.7x.

And as we mentioned in our December quarter earnings call, adoption of the ASC 606 new revenue recognition standard will start in the September quarter for Lam.

We will provide additional clarity on the impact of the standard in our next earnings call, but I wanted to highlight that adoption of ASC 606 will make our revenue more closely aligned with the timing of shipments.

Company noncash expenses included approximately $41 million each for equity comp, amortization and depreciation.

Capital expenditures were $49 million, which was down from $85 million in the December quarter.

And as a reminder, we expect CapEx in 2018 will be higher versus 2017 levels to support manufacturing network expansion and growth in strategic R&D investments.

We exited the quarter with approximately 10,600 regular full time employees.

The headcount additions were primarily in the factory and field with other additions in R&D.

Now looking ahead, we'd like to provide our non-GAAP guidance for the June quarter.

We are expecting shipments of $3 billion, plus or minus $150 million.

We expect continued strength in memory and slight growth in both foundry and logic.

We're forecasting revenue of $3.1 billion, plus or minus $150 million; gross margin of 47.5%, plus or minus 1 percentage point; operating margins of 31%, plus or minus 1 percentage point; and finally, earnings per share of $5, plus or minus $0.20, based on a share count of approximately 178 million shares.

We're pleased with our performance in the March quarter and with the guidance we just shared for the June quarter.

We anticipate our shipments to be biased to the first half primarily due to heavier first half NAND spending by our customers.

We expect comparatively more balanced to our revenues half-on-half.

Overall, we continue to execute well by having the right products and making the right investments at the right time to take advantage of the technology inflections driving the transition to the new data economy.

That concludes my prepared remarks.

Operator, please open up the call for questions.

(Operator Instructions) And we'll take our first question from C.J. Muse with Evercore.

I guess first question, clearly, a lot of debate around sustainability of memory and was hoping maybe you could help out in the DRAM side.

As you think about shrinking at the next node, what kind of bit growth are you seeing overall for the industry?

And what kind of new greenfield capacity do you think is required each and every year as we proceed from here?

That's a very precise question, C.J., each and every year.

So I mean, I think the headline that you've just highlighted with your question is that we're at an inflection relative to the economics of DRAM in 2 respects.

One of them is the cost of the investments.

It costs more for bit density today than it did 5 years ago.

But on the other side of that coin, the value proposition associated with what our customers are selling and what they're ultimately getting paid for those devices is dramatically different as well.

And so the point that we try to make in our investor meeting, while it's true to say that our customers have never spent more money, it's also true to say that in many respects, they've never spent a smaller proportion of their profits.

And so when we look at bit density per wafer out, it is obviously lower today than it was 5 years ago.

But pricing stability in DRAM appears there, profits, for sure, are healthy in the memory community.

Last time I looked, most of our customers were reporting 40% to 50% operating income levels.

And I think when we look at the spending, to best of our abilities to figure out, C.J., it looks disciplined, it looks balanced, and it looks healthy in the context of how customers are executing with a prudent investment in additions and significant investments in conversions.

Very helpful.

And as my follow-up, I guess, Doug, can you kind of walk through how you're expecting mix and gross margin to be impacted by mix as we move into the second half of the calendar year?

C.J., we only really give guidance 1 quarter at a time.

But what I'd suggest to you is, as I always do in my scripted remarks, gross margin will move around quarter-to-quarter depending on business volumes, product mix, customer concentration.

And if I'd look at the last, I don't know, year or 2, we've bounced around between 46% and 47%, and that's kind of what we've got baked into the financial models.

So that's probably a good signpost to use relative to how to think about it, C.J.

Our next question comes from Vivek Arya with Bank of America Merrill Lynch.

First one, Doug, I think you mentioned on shipments some timing of shipments came out of March went into June, if you could help quantify that.

And then as we look into June, you're guiding to some small decline in shipments quarter-to-quarter, which I believe is somewhat different than the seasonal norm.

If you could provide any color around that, it would be also useful.

Yes, Vivek, I mean, quite honestly, we missed the midpoint of guidance by $40 million, which honestly is 4 or 5 tools, when you think about it in the grand text of what's happening with our numbers.

So as always, you've got things that shipped, then they move around every quarter.

And I'm not really going to comment on any specific customer, but it was a pretty modest variance in terms of how to think about it.

And I think on the seasonality component of the question, one of the themes of the investor meeting was the irony of a consolidating kind of customer base but a more diversified demand portfolio.

And in transition away from kind of simple units to a density and the role that enterprise and cloud and data center has in the construct of this marketplace, I'm not sure any of us would agree what the headlines of seasonality are.

I'm not sure I could articulate it.

As a guy that's only been in this industry close to 20 years, I wouldn't be able to describe to you what seasonality is today because it's a dramatically different profile of demand for IC.

And the discipline of our customers is very different today than what's true 10 years ago.

And as a follow-up, Martin, foundry contribution, only about 10% in March.

I think probably one of the lowest levels you've seen.

How do you think about that contribution on a qualitative basis as we move throughout the rest of the year and especially the pipeline ahead of some of the node migrations that your foundry customers are going to go through?

Yes, I mean the foundry headline for the company is awesome.

I mean it could be better.

It can be more awesome, but we've made a significant amount of progress in the last couple of years.

And I talked to that in my prepared comments that not only do I think we have a foundational strength from the memory presence in the company, but the diversification of foundry and logic is a strength that perhaps is a little underappreciated.

The objective we have running the company is to ship products to customers when they ask for them, not early, not late.

And so when you see these quarterly kind of movements, it really is nothing more than a customer's request.

And frankly, I mean we would love to live in a world of perfectly calm and predictable and nonvariable world, but we don't.

So we do the best we can to ship to customers when they ask for it.

So there really is no headline associated with the 10% number that you referred to for the foundry, other than that just happened to be the schedule of the customer's request compared and contrasted with December and compared and contrasted with the June quarter.

So the strength that we characterized is real, we believe sustainable, and we're working hard to make it even more valuable going forward.

And for the group, we expect foundry will grow as we go through the year as well, just to reiterate a comment I had in my script.

Our next question comes from Toshiya Hari with Goldman Sachs.

Martin, I was hoping you could provide an update on China.

I think there have been a couple of media reports recently about some of your customers winning real business for the first time and some of your customers planning on ramping capacity later this year and to 2019.

So it does feel like we are getting closer to a decently sized ramp, if you will.

So I guess the question is, what are you seeing in terms of bookings today?

And what are your expectations over the next 12 to 18 months?

Yes, actually, I think our expectation today is almost exactly the same to the expectation that we started the year with, maybe even had almost a year ago.

I mean we believe in the vision that's being articulated, and we see our customers investing to build competency, capability and know-how to go execute that vision.

And indeed, we're aware that customers have reported real end customer business and output that presumably legitimizes this.

So going back to where we started this conversation in -- earlier in the year, we expect a couple billion dollars of incremental WFE this year in China.

And that represents 10%, 12% or so of the overall WFE kind of marketplace.

So the bulk of the industry is still outside of China.

And obviously, included in China is the global company investment, as well as the domestic community.

So I would say it continues to appear like there is steady progress.

Okay, great.

And then as my follow-up, I was hoping you guys can talk a little bit about the ALD opportunity longer term.

I think you touched on the subject briefly at your Investor Day.

But if you could remind us how you size the opportunity set longer term and if you can talk a little bit about the competitive landscape there, that would be helpful as well.

Yes, I mean I think we've obviously articulated, not just for the deposition business unit but also for the etch business unit, the importance of control and repeatability and opening up process windows for our customers, and atomic level processing is an important part of delivering that.

There are other significant challenges for our customers, not least aspect ratios, but atomic level processing is increasingly relevant to the success of the company and the success of the customers.

So it's a big investment.

It's one of the biggest areas of competitive momentum.

We don't actually characterize in any detail the proportion of WFE that we think is directly assignable because we obviously have competitive sensitivity to that.

But I think we've got fundamental differentiation on technology and productivity, and the engagement with the customers is comprehensive, and I'm pretty pleased with the momentum we have.

Our next question comes from Romit Shah with Nomura Instinet.

Martin, I definitely appreciate your comment around just DRAM spending looking balanced and disciplined, but seeing that shipments here are up over 30% sequentially and 50% year-on-year, I mean obviously realize that higher cost per bit is a factor.

But what level do you become concerned that your DRAM customers are adding too much capacity?

I would say I would be concerned that our customers are adding too much capacity if the analytics in our company in terms of end use demand concluded that there was a dramatic imbalance, and we have no evidence of that.

I mean we're trying to articulate, and our perspective may not be the best.

That's a choice obviously that you have to make, but we do our best to analyze kind of end markets and when we look at bit demand and bit consensus and then the construct of the market and isolate, for example, a 45% server bit growth assumption in DRAM, and that's the impact on investment levels and to the best of our abilities, understand bit per wafer and the investment choices for customers between conversions and additions, we end up in the same place that our customers are articulating plans.

And so it's less to do with how big the number is, in my opinion, and much more to do with whether it correlates to the statement of demand.

As well as affordability, Romit.

As you know, I mean DRAM revenue and profits are at all-time record levels.

Yes, got it, helpful.

And then could you also just talk about what's happening in logic?

You've -- or at least, my understanding is you have a very strong position at 10-nanometer and certainly a better position than you did at the previous node.

And we've seen revenues in this category run up last year.

You had a very strong Q3 but shipments down the last couple of quarters.

And I think you mentioned that logic was a little weaker than you anticipated in the first quarter.

Yes, I don't know that it's particularly material in the long term.

Again, I go back to my answer to the earlier question in foundries.

I mean all we try to do, if I'm very direct and honest, is ship product to customers when they ask for it, not earlier and not late.

And that's our focus and that's a priority.

So you're going to see up and down as customers adjust schedules, and most of the time, we get that right fairly well to our midpoint guidance, and we almost always get it correct in our range, in fact, I think always in our range.

So I mean I don't think we're doing a bad job signaling here.

But it's just a matter of the timing of investments by customers, and that goes left and right and up and down, and we just roll on through it.

The focus of the company to where you started is making sure that we have a product portfolio that's more relevant to our logic and foundry customers and to making sure that we're gaining more market share.

And in microprocessors and foundry both, hopefully, we've been effective at communicating pretty positive market share momentum in the last couple of years.

And Romit, the only other thing I'd add on that is I do think probably when you look at the logic shipments through the course of 2018, it will go up from where it is in March.

Our next question comes from Harlan Sur with JPMorgan.

As a follow-up to the prior questions on DRAM, maybe a different way of asking the question is if the team can help us quantify some of the reviews.

Overall DRAM industry capacity has been trending about 1.1 million wafer starts per month and relatively flattish for the past several years.

The market has been getting nervous in the amount of announced DRAM CapEx spend, announcement of new fab expansion products -- projects, but a lot of it, I think, is just to offset capacity declines and technology conversion and just to maintain a stable bit supply output.

So wanted to get your views on how much you see total industry capacity growing this year.

So I 100% agree with what you articulated in terms of the rationale for investments, whether it's addition or conversion.

I mean it's for demand, and it's the consequence of capital intensity relative to the incremental bits of various kind of nodes transition.

So I mean I'm in exactly the same place -- we're in exactly the same place that you articulated in your question.

Unfortunately, in a world of very few customers today, it's almost impossible for an equipment company to answer directly the question that you just asked.

It's too competitively sensitive for our customers.

So with due respect, I have to ask you to ask them.

But what I will say is we do see a modest increase in capacity in DRAM associated with the driver of demand to the consensus of the kind of low to mid-20 bit growth number that everybody seems to kind of talk about these days.

And it seems like it's perfectly relevant to the choices that our customers are making.

So unfortunately, quantification has to come from them.

Best I can tell you is it appears modest and balanced and rational.

Great.

Doug, a question for you.

Great job on the strong free cash flow.

Almost, I think, 35% free cash flow margin in the March quarter, although you did tell us that you were going to play some catch-up here versus the weaker December quarter.

But at a high level, going back to last earnings call and Analyst Day, let's assume that you can maintain operating margins in sort of the 30%, 31% plus range this year.

We would anticipate free cash flow margins for the full calendar year in the range of about 26% to 27% for the full year.

Is that kind of the right way to think about it?

Yes, that's the right way to think about it, Harlan I mean CapEx in the model in terms of how we think about it is 3% to 4% of revenue typically, so yes, that's what the company had in the model we gave in March.

Our next question comes from Farhan Ahmad with Crédit Suisse.

Doug, we saw a lot of volatility in the share prices in the March quarter so a bit surprised that we didn't see more significant buybacks.

Can you just talk about what are some of the variables that you think about in terms of buying back the shares?

Is there any constraints that stopped you from buying more?

Well, the primary thing relative to the $4 billion authorization is just availability of cash domestically.

87% of the cash that the company has still is domiciled outside the United States.

It takes time to move that around.

We've taken step one of that in terms of beginning to liquidate the portfolio, but there's administrative paperwork to dividend things up and down the tax structure to get the cash, and that's the primary thing that we need to have.

We need the cash before we can buy significant stock, and that will come over the next 12 months.

Look, almost all of the cash will be available by the end of the year.

And I think in the context of the question, it's also really important to emphasize kind of the long-term vision for the company that we talked about at the investor meeting, which is the redistribution of 50% of free cash flow generated over the long term, and nothing new to say today.

So it's a transactional moment in time, but the long-term commitment, I think, is the important message.

Got it.

And then one question on the NAND CapEx.

At the beginning of the year, you had talked about NAND CapEx being kind of modestly up this year and DRAM being up more significant.

If I look at your March quarter shipments, it's up a lot, and I was just wondering if your outlook in NAND has changed in any way in terms of the overall CapEx spending this year.

Not really.

It's more or less the same today as it was kind of 3 months ago.

Kind of what you're seeing is the first half bias, which Doug talked about in his prepared comments.

More or less the investment that we expected is the one that seems to be playing out.

And any 1 month or any 1 week or any 1 quarter, it can go left or right a little bit but pretty much where we expected it to be, and no fundamental messages in terms of kind of bit growth that are different from the consensus.

And maybe the only incremental thing I would say is our expectations of investments this year would kind of lead to approximately a similar level of shipped capacity not yet qualified at the end of '18 to the level that existed in '17, which said differently means the same level of discipline at the end of the year that existed at the beginning of the year.

Our next question comes from Joe Moore with Morgan Stanley.

To ask a slightly annoying question, I guess, that people are asking me as well.

You said in the last call that you thought that the shipment levels would be relatively balanced through the year.

They're down a little bit in Q2.

They're going to be down a little bit in the second half.

Is that just -- at the noise level, I realize it's a -- these are very small changes.

Has anything changed in terms of the way you see the pattern that you see through the year?

Not really.

I mean -- and I don't want to sound dismissive about the short term, but again, in context, whether it's in the context of the answer to the shipments kind of midpoint question or the difference between what we said last time and what we said this time, I think it's the same message in slightly different words, and none of it changes, in our opinion, the long-term opportunity and the importance of investing for long term and the potential that we described in the investor meeting.

So from the inside out, this is transactional noise more than anything fundamental.

And if that changes, we'll tell you, but that's about it right now.

Our next question comes from Timothy Arcuri with UBS.

I had 2 questions, guys.

First on China trade -- well, actually both of them on China trade.

First of all, is there the potential -- or is there any precedent for the government to restrict exports of equipment into China?

Well, we've lived in that world before, and it wasn't, I would say, a fundamental restriction.

It just made the process of doing business in China more bureaucratic.

We have to file for licenses and so on and so forth.

And so there is precedent for U.S. export restrictions of technology, not so much into semiconductor but into semiconductor when there was a dual use that the Defense Department or the Commerce Department were trying to manage.

Maybe if I take the essence of your question and go somewhere beyond it precisely to kind of a conversation about tariffs, we don't see an impact today on our business or on our customers' business or on our industry that would cause us to say that there is a new kind of long-term message.

But we're attentive to 2 very important things.

If tariffs start to be disruptive to consumer confidence, that has an impact in the global economy, and it will eventually impact semiconductors and in turn, equipment, and we're attentive to domestic equipment company agendas as well.

And I am not seeing that today, but if things got a little bit tit-for-tat, then there are obviously risks at a minimum that we need to be attentive to.

But at the end of the day, there's a fundamental value proposition that comes from every equipment company that's 30 years or 40 years old and that we win business, and we enable based on that expertise, competency and know-how.

So I'm not seeing kind of risks today, but we're attentive to the 2 things that we've just -- that I've just described: attentive to consumer confidence changes and attentive to domestic equipment company agendas.

Got it, Martin.

And then just there seems to be a school of thought that maybe this could accelerate the pace of the buildout of some of these indigenous projects.

I mean they seem pretty rational and pretty return-focused, but have you seen any signs that they may want to pull forward some of the time lines because of some potential down the road for this to evolve into something that's bigger than it is today?

No.

Our next question comes from Edwin Mok with Needham & Company.

First, my question is on foundry.

Just to clarify your comment, you said that the foundry CapEx you expect to be largely in line with 2017.

Is that correct?

I've seen more conversion from 10 to 7-nanometer.

I would expect maybe a slight decline on foundry.

And then kind of looking beyond this year, kind of based on how you guys look at demand and the foundry capacity, do you see a scenario where the foundry spending can potentially grow beyond this year?

I think we started the year with a kind of flat to slightly down kind of commentary for foundry year-over-year, and no change today.

And I think if I heard the second part of your question, it relates to whether there's upside, did I get it correct?

Yes.

Is there upside either late on this year or during 2018, there are some speculations that 7-nanometer being a big node.

I mean I'd say maybe.

I mean we're doing our best to articulate what we think is likely to happen, and the maybe is interwoven in this conversation around advanced computes as a component of this value proposition in AI and data economy.

And it serves no purpose to have a great cloud with great storage and great memory without great competition.

So if the value proposition accelerates, then I think it is an industrywide opportunity.

But in the context of the pilot line timings that we understand with our customers of 5-nanometer, the investments in 7 and 10 that are planned, the outlook that we've described is the best we have to offer right now.

Our next question comes from Mehdi Hosseini from SIG.

Two follow-ups.

You mentioned your revenue for the second half of calendar year is more balanced with the first half, and we have the revenue for the first half.

Does that suggest that we should assume a quarterly revenue that starts with a 3 handle for the September and December quarter?

No, Mehdi, not necessarily.

Relative balance could be plus or minus 51-49 something like that.

It's -- I'm not giving you guidance for every quarter through the rest of the year right now.

The comment was -- Martin described shipments low 50s, high 40s half-on-half.

My comment was revenue would be a little bit more balanced than that.

Sure, sure.

In terms of guidance, let me go back to some of the commentaries from the Analyst Day.

Obviously, we have been debating sustainability of memory spend since your shipment hit $1 billion target, and now we're at the $3 billion.

And sustainability of memory spend keeps coming up.

In the meantime, your revenues are actually diversifying.

You have recently started to highlight the services mix and especially with the new tool shipped to the NAND customers, there's a higher services content.

And if you really believe in a new paradigm shift, why can't you help us to better model and kind of shift away from these seasonality, where if I take your shipment for the March quarter, analyze it, that would suggest shipments up well over 20% for the year?

And perhaps maybe some additional color on the revenue mix could better help us move away from this seasonality that's been just lingering.

I'm not sure exactly what you're referring to seasonality, Mehdi.

I mean at the end of the day, shipments and revenue mix would be exactly the same.

One is timing, right?

Shipments sometimes happens sooner than revenues.

So at the end of the day, it all normalizes to the same thing, which is why we give you the shipment color the way we do.

I would say -- if I can just add one more thing.

I would say, I mean for anybody on the call at any point in time listening to Lam Research, if you have recommendations on specific disclosure that you think would be helpful, if we feel like we can do it in the context of respecting our customers and if we can do it in the context of preserving competitive advantage, we have every motivation to do that.

So we're doing the best we can to articulate an outlook and an opportunity and risk and so on and so forth.

And if you have specific recommendations, please make them offline, and we'll do the best we can.

Does your shipment represent services, spares and parts?

Or is it just a system?

When I give you the color on the percentages every quarter, Mehdi, I specifically say system shipments, so it's system.

Right.

So if you're going to realize $1 billion of incremental revenue from services, that won't be captured by shipment, and that's what could become a difference looking forward.

Yes, potentially.

Yes.

Right.

That's not a good suggestion with the guide?

We'll consider.

We'll think about how we can give you a better understanding on what's going on with services, which I think at the end of the day is what you're asking for, right?

Yes.

All right.

We'll think about it, and offline, if you want to share with us what would be helpful to you, that would help us think about it.

Our next question comes from Atif Malik with Citigroup.

I just want to go back to the shipment miss relative to Street and your midpoint of expectation and the timing of the project.

Can you share with us if there is one end market that's responsible for this $40 million shortfall Is it more foundry?

Or is it more NAND?

Atif, as you can appreciate, we have a very small set of customers, and giving you more color than just timing of certain projects, with certain customer projects, is probably the best we can do for you.

Okay.

And then as a follow-up, just broadly speaking, can you share with us your executions on the timing of the 92, 96 layer 3D NAND for the industry versus the 64 layer NAND migration?

When should we expect the majority of the NAND makers moving to 92, 96 layer NAND?

I think the best we can do is refer you back to the Flash Memory Summit presentation we made last year, which to the best of my recollection, details all of our assumptions in terms of phases of development for NAND flash.

And rather be selective to one and risking being inconsistent with that, I'd ask you to kind of go back to that disclosure.

And if there's a remaining question, then please follow-up with Satya.

Our next question comes from Patrick Ho with Stifel.

Martin, maybe first off, in terms of the foundry, logic commentary you detailed, given a lot of the noise out there surrounding EUV, can you detail from a Lam perspective how you -- what application and what new areas do see for both etch and deposition that are helping you grow that served available market in spite of, I guess, some of the bear concerns about EUV reducing the capital intensity of those 2 processes?

Well, I guess at a very basic level, we've attempted for the last several years to articulate a set of assumptions, which I think are generally consistent with the assumptions that our customers communicate and the assumptions that ASML will communicate around adoption of EUV.

And using those assumptions, we've attempted to communicate this through the 5-nanometer transition our SAM increases in etch and deposition.

So I would say the first part of answering the question is, we have articulated to the best of our ability increasing SAM through 5 nanometer, and we'll see what integration schemes show up with 3 before answering that question.

In addition to that, the company is making a very specific set of investments, and you've seen one of them quite publicly, or maybe 2 of them quite publicly, but I'll pick one of them now, and that's the world of advanced equipment process control.

So we've articulated incremental SAM that we're targeting for the purposes of delivering more control in process and more repeatability, more uniformity, so on and so forth.

And that is directly relevant to foundry and microprocessor opportunities for the company as well.

And I'm not going to go into details, but we have 3 or 4 pretty significant SAM expansion objectives above and beyond the inflections we've been talking about for the last several years.

And that's all obviously incorporated in the long-term financial models that we presented a few weeks ago.

Great.

And Doug, as a follow-up, I know you probably won't say that visibility has significantly increased, but just based on customer projects and the timing of them, it seems like the equipment industry as a whole is getting much better visibility than they've ever gotten.

Your balance sheet metrics and the procurement and everything you've done seems to be reacting very well.

What changes have you made that has allowed you to adjust to, I guess, maybe increased visibility as well as the higher demand levels?

I mean all the technology inflections you've seen happened over the last several years.

Patrick have required our customers to have very advanced conversations, so that's well in advance of when they need a certain capability because we need to develop the equipment, quite honesty.

So I don't think that's new.

It's been happening over the last several years, and it's enabled us to plan our company to be able to support where we're at today.

So the conversation is very deep and rich when you're enabling the customers' road map in the way that we are.

Our next question comes from Craig Ellis with B. Riley FBR.

I just wanted to ask a question to get some longer-term context around full year shipment commentary.

I think if I look back over the last 2 years, there have been years when we look for a half-on-half profile that would be down in the second half and yet the year played out and it wound up being up.

So as you look at the risks on both sides, can you just recap the upside risk to that shipment outlook as well as the downside risk?

I guess I am not sure what to say here.

I mean at the end of the day, if you believe the statements that we're making around discipline and demand-led investments, then the kind of known commodities at this point in time relates to the capital intensity of any one technology node or any one kind of device architecture.

So I don't think there's kind of too many risks or opportunities relative to people's understanding of kind of cost consequence of a DRAM transition or a 3D NAND transition.

Individual customer plans, in the short term, can create artificial kind of disruption, I would say, in terms of WFE.

In the long term, I don't think at all because if one customer decreases, another one likely increases.

If one increases, another one slightly decreases.

So you kind of see kind of customer risks and opportunities.

And if your window of focus is a quarter, then that will be disruptive to you.

If your window of focus is a year, it's irrelevant because it's a dynamic kind of marketplace.

I do think -- and I would extend the customer conversation into a regional conversation, something like China, right?

I mean I think our hypothesis still remains that with any region of the world, including China, that the investment will be demand-based and rational.

So I do think where there is still a significant amount of variability in plans is kind of how people choose to go execute.

And you can change the WFE investment profile quite materially through the choices you make around new capacity additions versus playing that to 3D transition versus 3D transitions scaling versus maybe modifying a DRAM line to a flash line or vice versa, so -- and maybe even reuse strategies.

And that all sounds maybe quite transactional, but it can have a fairly significant impact.

So we do our best to try and dialogue with customers to understand where they're going to head.

I would say the simplest headline to your answer is, if demand is stronger or weaker, it has the -- for ICs associated with these AI transitions and value propositions of a data economy, whether you'll see it in the calendar year is debatable, but you'll see it over a couple of year period in a positive or negative direction.

And the rest of it, I think, is more operational execution as customers try to optimize their fabs, their line layouts and they try to optimize their use of cash.

And that concludes today's question-and-answer session.

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

That's all the time we have for the call today.

Thank you for joining us.

That does conclude today's presentation.

We thank you for your participation.