Lam Research Corporation (LRCX) Q4 2011 Earnings Report, Transcript and Summary

Ladies and gentlemen, thank you for standing by, and welcome to the Lam Research Corporation June 2011 Quarterly Results Conference Call. [Operator Instructions] And I would now like to turn the conference over to Shanye Hudson, Director of Investor Relations. Please go ahead.

Thanks, Mikaela. Good afternoon, everyone, and welcome to Lam Research Corporation's quarterly conference call. With me today are Steve Newberry, Chief Executive Officer and Vice Chairman of the Board; and Ernie Maddock, Senior Vice President and Chief Financial Officer. Shortly, Ernie will discuss financial results for the June 2011 quarter. Steve will then share Lam's business outlook for the September 2011 quarter before opening up the call for Q&A. The press release detailing our financial results was distributed over wire services a short while ago and is also available on our website at lamresearch.com. Today's call contains certain forward-looking statements, including those related to our expectations of market size, technology trends and customer spending and behavior, our investment plans, the capabilities of our products and operations, our tax rate and our forecast of market share, shipment, revenues, expenses, margins, operating profit, earnings per share, cash generation and free cash flow, as well as other statements of the company's expectations, beliefs and plans. There are important factors that could cause actual results to differ materially from those described in these forward-looking statements, and a list of these factors can be found in the slide package accompanying this conference call and on our most recent Form 10-K filed with the Securities and Exchange Commission. All forward-looking statements are based on current information, and the company assumes no obligation to update any of them. This call is scheduled to last until 3 p.m. [Operator Instructions] And with that, I'll turn the call over to Ernie.

Thank you, Shanye. Our June quarter was largely in line or better than our midpoint of guidance and represented a strong close to our fiscal 2011. For the quarter, shipments were at the high end of our guidance range at $793 million, a sequential decrease of 2%. Application and market segment breakdown for the quarter were as follows. Applications of 65-nanometer and below represented 91% of overall system shipments and 82% of overall system shipments were for applications at 45-nanometer and below. System shipments for NAND were 44% of overall system shipments, followed by DRAM at 11% and other memory at 1%, making the total Memory segment 56% of overall system shipments. Foundry shipments represented 31% of overall system shipments and the remaining 13% were for Logic and Other. I want to take an opportunity to highlight the fact that starting this quarter, we are using non-GAAP rather than the phrase ongoing to describe the results of our operations that we feel are relevant for comparing the company's operating performance. The only difference between the ongoing and non-GAAP nomenclature is the exclusion of the non-cash interest expense accompanied with the convertible notes issued on May 2011. Otherwise, the presented results are directly comparable to those previously described as ongoing. Revenues for the quarter were approximately $752 million, slightly above our guidance midpoint and down 7% sequentially. Gross margin was at 45% at our guidance midpoint and was down from 46.2% in the March quarter due to lower factory and field utilization associated with the decline in business volumes. Total non-GAAP operating expenses were $180 million, up $3 million sequentially due to continued investments in strategic development activities at the customer interface. Although we expected slightly higher levels of OpEx, we benefited from a release in our accounts receivable reserve. As…

Thank you, Ernie, and thank you all for joining our call today. As many of you are aware, we held our Analyst and Investor Event at SEMICON West a few weeks ago, and I would like to take a few moments to reiterate the key messages we outlined during that meeting prior to providing our view of near-term market conditions and business outlook. We remain encouraged by the opportunities in the wafer fab equipment industry over the next couple of years, which is supported by a couple of things we discussed during SEMICON West. Those themes include trends in the mobile and computing space, where content-rich devices, such as smartphones, tablets and notebook PCs continue to drive demand for IC units. We expect that IC units will grow by 10% to 11% annually over the next 3 years to roughly 275 billion units by 2014. This bodes well for equipment suppliers as this strong IC unit growth will drive the need for additional wafer-processing equipment. The highly competitive mobile market is driving semiconductor device manufacturers to continue down the path of MOSFET to produce logic and memory chips capable of meeting the power, performance and form-factor requirements of these mobile products. The scale and challenges presented by these demands are significant and will require new and highly complex device architectures starting around the mid-to low 1x node in Memory and 20-nanometer or 14-nanometer node in Advanced Logic. As device complexities increase with each subsequent technology node, capital intensity is rising. As a result, IC manufacturers will need even more highly differentiated technology and productivity solutions, and Lam is well positioned to deliver on both of those requirements. These industry trends form the basis for our view of an annual WFE spend environment of approximately $30 billion plus or minus 10%…

[Operator Instructions] And our first question comes from the line of C.J. Muse with Barclays Capital.

This is Olga calling in for C.J. I guess, my first question is on the commentary you made around shipments into the second half of the year. You talked about how you see the potential for that to be flat to down 19%. Is that for the industry or for Lam, specifically?

Well, it's for the industry. I think, that if you look at what's occurring right now that if we were at a $32 billion first half annualized run rate, if we're going to end up at the lower end of our $29 billion to $32 billion range, then the potential is that the second half shipments for the industry would be down 19%. For us, given that we don't participate in the microprocessor market, it really isn't any different from a percentage standpoint, because spending is pretty balanced between first half and second half relative to microprocessor. So if we're on the low end of the range, then our shipments would likely be mirroring a decline of the industry, with the exception of offset improvement for us as a function of market share gains in Etch and Clean.

Got it. And then as a follow-up of that, if you sort of do the math around it, it would imply a decent uptick in the December quarter shipments. Are there any specific customers or lines that you're tracking that would give you confidence that you would see that level of shipment uptick in the fourth quarter?

I would have to say that I have no confidence or very little confidence based on kind of the environment that we're in right now. I think, it wouldn't be appropriate for me to really talk about specific customers. But clearly, what you're suggesting is that, given where we think the shipments are going to be for September that it implies that December is better. That's certainly true if the customers execute the plans that they are currently communicating. And as I've said many times on these conference calls, it is with the consolidated industry, with short lead times, it's essentially impossible to predict with accuracy who's going to do exactly what, when are they going to exactly take deliveries and with consolidated customers spending large amounts of money, all it takes is one customer to decide to move a shipment out of or a set of shipments out of the September quarter or out of the December quarter and things can change, significantly. So as Ernie and I have been talking about, I think, the industry is going to be volatile on a quarter-to-quarter basis, and therefore, I tend to look at what happens on a rolling 6-month basis and what happens on a rolling 12-month basis, and that's why you hear me commenting on, we think that on average, over the next couple of years, we're likely to see $30 billion on a rolling 12-month type basis. But what's going to happen on a quarter-by-quarter basis, I think is going to be volatile.

Got it. And then just a follow-up question. On the market share comments, any sense, given we're now halfway through the year, where you think you're Etch market share will be for the year and on the Clean side as well?

Well, as we said at SEMICON West, we think, we'll pick up 1 to 2 market share points in Etch. We think we'll pick up 3 or 4, possibly even 5 market share points in Clean. Clearly, market share is a function of total spending and the revenues and shipments that we generate, but also customer mix. And so with the volatility around customer mix, I think, we'll be in that ballpark. But it's impossible to tell for sure with the volatility.

Our next question comes from the line of Satya Kumar from Crédit Suisse. Satya Kumar - Crédit Suisse AG: Steve, I was thinking that you would guide shipments down 25% just based on what you said at the Analyst Day and what the other semicaps have said. But you mentioned that you saw some additional push-outs just 2 days ago. I was wondering if that was the case, why shipment guidance were not even weaker. I was wondering if you could add a little bit of color on what segment...

Satya, can you speak up, it's very difficult to hear you. Satya Kumar - Crédit Suisse AG: I apologize. I think at SEMICON, based on what you guys said and based on what others semicaps are saying, a down 25% shipment guidance seems exactly par for the course. But on the other hand, you also you said you saw some additional push-outs 2 days ago. I was wondering if that were the case, why shipments weren't even weaker. Perhaps you can give a little bit of color on what was pushing out just most recently and what's happening in the very near term?

Yes. I think, I mean, you would know better than I, because, I think, some companies were spending time talking about their bookings were declining. Some companies didn't talk about shipments. Up till a number of days ago, I would have expected that our shipment environment would not be down as much as it is. We're essentially down 27%, June quarter to September quarter. The reality is with this volatility, I mean, if I had another conference call in 2 weeks, I might be telling you that it's gotten better. I mean, it really is one of the most kind of unpredictable environment that I've been a part of in many, many years. And so I think that we're in the vicinity of where we thought, but I do think that as everyday goes by, we just never know if somebody's going to call us and say, "Hey, pull something in or push something out." So this is kind of where we're sitting right now. And unfortunately, really, from a customer confidentiality standpoint, all I would tell you is that if I were to look at what has changed over the last quarter from what customers were asking us to deliver, I would say 2/3 of the change is coming out of the Foundry space and about 1/3 of the change is coming out of the Memory space. And most of that being DRAM-related, which I don't think is a surprise to anybody, given most of the comments around consumer PC really not being as robust as people thought, and that's caused DRAM bit growth to really be more around 45% than what was earlier thought to be about 55%. So we've seen a pullback in the DRAM segment, and I think, the utilization issues in the Foundry and a change in ramp plans and expectations on the 28-, 32-nanometer node in the Foundry has been kind of well-documented over the last month or so. Satya Kumar - Crédit Suisse AG: Okay. And a quick follow-up, in terms of your gross and operating margin guidance for September. It perhaps alluded to your Semicon model, the gross margins to me, looked like they may be 50 to 100 bps lighter and OpEx was up about $8 million or so, which I would have thought perhaps a little bit less than that. I know you're investing in R&D. I was wondering if you could add a bit of color to that.

Sure. If you actually, look -- Satya, this is Ernie. If you actually look at the specific wafer fab environment that we're facing in the September quarter, it's significantly below that low end of the $28 billion model. So depending on how the perspective, it's somewhere in the $20 billion to $22 billion range. And so you -- if you took the $28 billion model and scaled it down, that kind of puts you within a relevant range of the sort of gross margins that you would expect. So we do believe it is very directly attributable to the specific September quarter environment that we're facing.

Our next question comes from the line of Patrick Ho from Stifel, Nicolaus. Patrick Ho - Stifel, Nicolaus & Co., Inc.: In terms of the shipments outlook for the September quarter, can you just give a little bit of color in terms of the customer breakdown, how it's shaping up right now in terms of Foundry, Memory and Logic?

I can talk in generalities. I mean, as Ernie said, in the June quarter, we had about 56% of our total system shipments were memory, with most of that being NAND. In the September quarter, we're going to see probably a little over 60% be memory with again a lot of that being NAND-related. Foundry and Advanced Logic would be about high 30s. And I think, while the percentages are what I just said relative to mix, both segments quarter-over-quarter in terms of the amount of shipments they're taking, all the segments are down. It's just that the Foundry and Logic segment in the September quarter is down even more than Memory. Patrick Ho - Stifel, Nicolaus & Co., Inc.: Great. That's helpful, Steve. Maybe a little bit of color in terms of the DRAM market, given your exposure to that segment. Obviously, the consumer PC market is weak, as you stated. But when do you see the industry making that transition from a 4x to 3x and eventually 2x, where some of those technology transitions start to benefit Lam, particularly? What's going to be the catalyst there for them to get the, I guess, the cost benefits from those transitions?

Well, I think, the industry, it's clearly right now, in a significant transition, where 4x is the bulk of production. But there's a lot of 3x production as well that's beginning to come online. And I think, that a lot of companies are expecting to be converting to the 3x activity throughout the course of this year. But a significant ramp to 3x in calendar year '12, we think that there is a greater challenge to do technology conversions at the 3x node than there was to go from 6x to 5x to 4x. And so therefore, if we get a reasonable amount of bit growth in the 2012 timeframe, we may see that the DRAM spending in 2012 is going to have to come up even if you had the same bit growth. And then if we get a stronger consumer segment, because, I think, most of the companies that we talked to, the commercial, industrial segment and the server segment is strong and expected to stay strong in 2012. So the consumer's kind of the wildcard in PC, which really drives a lot of DRAM bit growth. And so if the consumer comes back in 2012 stronger and we have stronger bit growth in 2012 relative to '11, then, I think, we could see $1.5 billion to maybe even $2 billion of increased DRAM spending if we get that kind of environment.

The next question comes from the line of Stephen Chin from UBS.

Stephen, Ernie, just a question on the shipment push-outs, again. Are they mostly capacity-related, Steve? Or are you also getting an equal amount of technology upgrade type push-outs from the customer base?

Well, when you're talking about memory, memory's a pretty dominant conversion-related activity this year, particularly in DRAM. We think there's probably 450,000 wafer starts per month of conversion activity, with only 60,000 wafer starts of new DRAM capacity. NAND, it's a little better story in the sense that while there's 500,000 wafer starts or so of conversion, there is 240,000 to maybe even 250,000 new wafer starts. So when you're talking memory push-outs, you're always talking about some of the aspects of conversions are present. But if they are pulling back on capacity expansion, which is certainly the case in DRAM and not the case in NAND, you're going to have some of that be full system that are a bigger part of the push-out than just conversions.

Our next question comes from the line of Jim Covello from Goldman Sachs.

Specific to the foundries, in particular, the foundries that are pushing out in part, because they have some yield issues, do you think them fixing the yield issues alone could drive a resumption of order activity? Or has it gotten to the point where they need to fix the yield issues, and then also see an improvement in utilization?

Well, they're probably a heck of a lot more qualified to really answer what they're going to do. I think, that in some respects, the speed at which they're getting up the yield curve at the 28- and 32-nanometer node is actually, at least, on schedule, maybe even ahead of schedule. And so I don't think any of us should be thinking that there is like huge yield problems that are not really kind of the norm in what you would expect. But when you talk to the customers who are going to be buying those 28-nanometer ICs, most of them really have it in their plans that they're going to start ramping up their demand in the first 2 quarters of 2012. Certainly, there is some output that's going into products at the second half of 2011, but that's relatively low. And so I think that when we talk about what's going on at 28-, we're talking about 6 months ago, what did the Foundry and Advanced Logic companies think the speed was going to be that they could potentially ramp? What did they think the demand was going to be based on what their customers were either asking them to do in terms of where we are now, is the reality is it's both they've got to get up the yield ramp curve especially when you think about how expensive it is. You don't want to just throw more equipment and throw more wafer starts at it and have to amortize $1 billion for 10,000 wafer starts over relatively lower yielding. They could afford to do that when it was half the cost for 10,000 wafers, and sometimes they did that. So I think it's a combination of demand. I think, it's a combination of that with yield, but I do expect that it's, likely, we won't see the 28-nanometer ramp in volume until kind of the first half of 2012.

Our next question comes from the line of Timothy Arcuri from Citigroup.

Steve, one of your largest memory customers, in fact, your largest memory customer, one of the senior executives was on the tape on Monday talking about raising memory CapEx. And I think, a lot of the people close to that company are believing that tomorrow they will raise memory CapEx. So I'm -- speaking to the volatility of the environment, which is what you mentioned. How much of a potential raise from that company is in your guidance, because it seems like it's not in the guidance. So I'm just sort of wondering what you make of those comments?

Yes. We've heard those comments, too, and none of any potential increase in CapEx spending is in our guidance. And clearly, if more demand for deliveries comes from that customer, then that would be upside from the kinds of numbers that you've heard me talking about.

Got it. Great. And then just a follow-up, Ernie, I'm just sort of -- back to a question that was previously asked, I'm looking at the OpEx necessary to get to the guidance. And it seems like OpEx is going to increase for both R&D and SG&A. I guess, I can understand why it's going to increase for R&D, given maybe some early 450 development. But I'm curious as to why it's increasing for SG&A. And I'm also wondering maybe how much of the increase in R&D is due to early 450 development, if any?

Tim, as we've talked about before, the vast majority of the increase in our OpEx is truly R&D related based on our internal accounting for certain things that we are providing as a part of that R&D activity, specifically, since it's customer-facing, and for example, may involve providing a tool to a customer, we might classify that as SG&A expense. So in terms of what you would possibly think of as SG&A, there is relatively little increase associated with that. Relative to your question about 450, we are spending at about the level we feel we need to spend to be prepared when our customers are demanding of that diameter. And so I would say, that our level right now is relatively modest compared to what it will need to be at full scale. But it is -- it's certainly present in the spend.

So Tim, I would amplify one thing relative to 300-millimeter spending. Normally, when you're focused on, let's say, ramping up what we'll call the 28-nanometer node in Logic and the 3x node in DRAM and the load 2x node in NAND, you're really out there with JDps and [indiscernible] for one technology, no doubt. We're out there now both for that one node and then the second node, because the second node is where the new architectures are coming into play and customers are doing far more work and running far more wafer starts on the development tools to work through all the issues associated with the new 3D NAND, the new FinFET structures, the new DRAM capacitor structures, even MRAM activity. So that's making R&D more intensive. And as you well know, whether shipments go up or down or revenues go up or down, the customers' activity from the standpoint of next-generation technology never ends, and so we're going to invest through this adjustment period, because we believe it's going to be short-term in nature and that's why you hear me talking about what I think kind of a rolling 12-month environment should be, $30 million plus or minus 10%. And we believe that the investments we're making now, while they put some pressure on our P&L, they are going to position us for the kind of market share gains that we want to see, and it's going to benefit us significantly in the 3- to 5-year period, when those next-generation architectures ramp.

The next question comes from the line of Ben Pang from Caris & Company. Benedict Pang - Caris & Company: One the test [indiscernible] [Technical Difficulty]

The next question comes from the line from Krish Sankar from Bank of America Merrill Lynch.

Two questions. Steve, if you look at your shipment split in both June and September, how much of the Etch portion is coming from dielectric versus conductor etch?

Well, let me think about that. And reference the documents, since I don't have that off the top of my head. We have that data, but we don't really put it together for this kind of call. We kind of tend to look at what's happening in the etch market relative to what's going into Logic, Flash and DRAM. So I really can't answer that. We'll have to get back to you with kind of how that's breaking out. But the thing I would tell you is that fundamentally, conductor has grown. It used to be somewhere around 40%. But over the last 2 years, it's now about 50% of the total etch market. And it really doesn't matter, whether you're talking about Logic or you're talking about Memory, conductor applications have grown. And so therefore, I don't think you're going to see that there is much of a difference in terms of where the activity is going between dielectric and conductor.

Got it. Now if I could just ask a follow-up. You said you're targeting 3 to 5 points share gain in Clean market this year. I'm just kind of curious. Where do you think that opportunities are going to come from. Is it going to come from your Japanese competitor? Or is it going to come from the local Korean suppliers?

Well, we're making a lot of progress in memory, which is important to us, because if you look at the historical market share that we've had in single wafer spin clean, it's been largely, Logic, back end of the line. And so what we've talked about at SEMICON was that we expect to pick up 12 to 14 new penetrations, the majority of those will be in memory. And the interesting thing is they come from a whole variety of competitors, from some Japanese clean companies to Korean clean companies to even non-Japanese and non-Korean. So it's actually a pretty wide distribution and at the end of the day, we're not so concerned who it comes from. We just want the win and we want the market share.

The next question comes from the line of Edwin Mok from Needham & Company. Edwin Mok - Needham & Company, LLC: So Steve, I have a question regarding the NAND space. It sounds like you're commentary suggests that space is still progressing this year. And that they're adding capacity throughout the year. I was wondering are you worried that the NAND space will be coming like DRAM in the coming year after all these capacity expansion, they will just switch their focus on conversion, and therefore, cause a slow NAND spending environment in the coming year?

Well, conversions are a function of whether or not in fact the next technology node from an architecture standpoint can be converted from a technology standpoint. The other aspect of conversions are the more wafer or the more bit growth, and therefore, silicon output you need, there's only so much that you can take and convert before you have to add new wafer start capacity. And so the reason that you see NAND having to add a couple of hundred thousand of new wafer starts is that bit growth. Now if the bit growth were to slow down, then I think the ability at least until we get to 3D NAND, will be to have a greater percentage of conversions. So the real question is what happens to NAND bit growth in 2012 and '13, and I think a lot of that's going to be a function of what happens to SSD penetration, what happens to tablet growth, which is obviously SSD-related, and how quickly can they get down to really low-cost, high-yield at the low 2x node, what's the ramp profile for the high 1x node. And I think, that we'll have to say that play out. But I would expect that we'll see pretty healthy NAND spending similar to what we saw in 2011. But it's going to depend on what the bit growth profile is and if it stays up around 80%, it will be about the same. If it drops, then we would see some less spending. But the potential for bit growth to expand in NAND exists. I think for next year, it's all about whether or not you get more significant SSD penetration. And right now, I think the jury's still out on that. Edwin Mok - Needham & Company, LLC: I see. It's all come to demand. That's great. Just a quick follow-up on the buyback, just curious how do you guys think about that? I mean, obviously, your business is sharing a lot of cash even in a moderated demand environment. Are you guys going to use a lot of that cash generated buyback stock, given that you avoid so much cash in the balance sheet?

Edwin, it is completely consistent with what we talked about at SEMICON. So we have a purchase plan in place that is on autopilot at this point. And based on the trading price of the stock, we will be buying back various volumes over the course of any given quarter.

And we have a question from the line of Ben Pang from Caris & Company. Benedict Pang - Caris & Company: But on the $10 billion to $13 billion that you commented on -- for the logic Foundry, what was your outlook prior to the last couple of months? And then a one follow-up.

Well, at one point, if you go back 3 or 4 months ago, its probably more likely to be $12 billion or $13 million. I think, where we are now, it's more likely that if more spending doesn't occur, it would more likely be $11 billion. It could go as low as $10 million, but I think it's right now looking like it's going to be $11 billion. But you just really don't know. All it takes is 1 or 2 of these companies to come in, decide to add 5,000 wafer starts at 28-nanometer and the next thing you know, you've got another $1 billion, even $2 billion that could be added to WFE spending. So when you're talking about 28-nanometer now, and you're talking $1 billion for 10,000 wafer starts, it can move the needle pretty quick. Benedict Pang - Caris & Company: So that was kind of my follow-up that is the growth right now for Foundries solely kind of at the 28-, 32-nanometer that even 40-, 45-nanometer, there's kind of no boost here at -- even with like a 10%, 11% unit volume for semiconductors?

No. I don't think that's true. I think, we tend to focus a lot on 28- and 32-, because it's so capital intensive. But if I look at -- in 2010 for the 40-, 45-nanometer node, we think they probably added about 30,000 wafer starts. We think they're adding somewhere around 40 to 40,000 wafer starts at that the node this year. So clearly, at about 650 for 10,000 wafer starts, they're spending a fair chunk of change at the 40, 45 node, and they're adding more capacity there clearly than they're adding at the 28 and 32-nanometer node for now. But as we go forward into 2012, I'm expecting that we'll see the 28-, 32- node will add more new wafer starts than the 40 to 45-nanometer node, and therefore, we'll see more Foundry, more Advanced Logic spending in 2012 as the capital intensity of the node rises.

And we have a question from the line from Atif Malik from Morgan Stanley.

I apologize if this question had been asked before. Steve, can you comment on the utilization levels in the leading edge, 40-nanometer and below and maybe what utilizations are at foundries for 65-nanometer and above?

Well, I think, it varies by Foundry for sure. But in general, when you're looking at 65 and above, you're looking at utilizations anywhere from 65% to 80%. I mean, you can find a foundry here or there where maybe that utilization is lower. I think, that when you look at the 40, 45 node, the utilizations are rising. They're pretty strong at the biggest foundries. That's where there is a lot of demand moving. And there isn't as many total wafer starts at that technology node. So if the demand is there, it's easier to have higher utilization at that node. If you then look at the 28, 32-nanometer node, you've got some capacity and relatively low. And depending upon which fab, the utilizations are anywhere from 50% to 75% or even 25% to 75%, depending upon which customer and who they're trying to ramp for. It really is all over the place. But I do think, the encouraging thing is if you look at there's more wafer starts on the leading edge at 40, 45 and it's growing, clearly, 65 has the most wafer starts in terms of the technology node. But customers are moving to that 40, 45 node and utilizations are in pretty good shape at that node.

Great. And then just a follow-up. Lam does not have high exposure to Intel in the U.S. If I look at Intel spending this year was quite high, about $10 billion in CapEx and next year, if I assume it's around $8 billion, how can Lam close the gap between its growth rate year-over-year versus the peers, assuming all else is equal, like in the share gains and the capital intensity increase and the NAND side offset your lack of exposure at Intel?

Well, I think, that in a year like this, where Intel really is 20% of wafer fab equipment, even with market share growth, we can't offset that in any 1 year. And so we focus on winning application market share at those customers that we do business with. We can't change how much money Intel's spending right now. But we can certainly, in a normalized year, we can take those market share gains that we're getting at the other customers and have that manifest itself in faster growth than the WFE environment which we've clearly done over the last 10 years. Certainly, we are involved with ongoing evaluations, investigations, discussions with all of our customers. And certainly, we recognize that if there is an opportunity to be able to be a supplier for Intel, we would certainly find that to be a situation that would ultimately make it easier for us to accelerate our growth rate faster than WFE. And so we'll just have to see how that plays out. But I think with the market share gains that we're targeting in Clean and in Etch, that against a normalized Intel spending, we'll grow faster than WFE, if we execute to our plans.

Our next question comes from the line of Mehdi Hosseini from Susquehanna International.

Steve, I'm looking at the mix of your NAND and it seems like NAND shipment is up almost 50% year-over-year, if I were to take the June quarter and the March. And if I were to look at the Foundry, that one is also up 20%, first half of this year versus the first half of last year. So doesn't that imply that maybe the NAND guys are done spending and unlike previous cycles, they're more cautious and industry's consolidated so that they're not as aggressive with the spending. And then on the Foundry side, most of the large established chip companies have reported, they've actually talked about decent yields at leading edge 28-nanometer. And given the fact that Foundry shipment is up 20% year-over-year, doesn't that imply that maybe Foundry guys are also done and front-loaded. And put together, it's just that the CapEx upside that the industry -- equipment industry was expecting is not materializing?

Well, to the extent that I can understand your question, I think, at the end of the day, all spending has to be looked at in terms of what's the demand environment that is present for NAND, present for DRAM, present for Advanced Logic, what customer is ramping on what technology node and what's the capital intensity for that. And when you look at supply and demand numbers relative to NAND, they're very much in balance. When you look at where the NAND companies are relative to cost and the pricing, they're making money. When you look at DRAM, it's probably running a little bit below, the normal cost curve. There's probably a little bit of excess demand in the pipeline for DRAM, which is why I think you see that there is some slowing down of how much more conversions and capital is going into DRAM. When you look at Foundry, as I commented, I think, they're doing fine relative to the targeted rate of yield improvement on the leading edge. But let's not confuse that with the yields at 65 and the yields at 40, 45 are higher than the yields today at 28. And so, a company has to be careful not to take too much demand relative to what it costs to output that demand. And so while yields are still coming up as a function of the learning curve, you want to keep your investments in capacity expansion in sync with a reasonable demand that you can meet, because if you take on too much demand and you make too many commitments and your yields are running low, you have to offset that by putting a lot more equipment in, and the cost doesn't change even though your output is reduced. So I think, when we look at the demand environment, when we look at capital intensity, I think we're in a period where the customers are kind of consolidating their gains. They're adjusting very rapidly, which they can to kind of align how quickly they bring capacity on relative to demand, and therefore, accounts for demand. And like I said, I think, we've got volatility in the near term, but I don't think that it fundamentally is changing the trend of what the investment levels are that are going to have to be made to deal with 10% to 11% average IC unit growth over the next 2, 3, 4 years.

Our last question comes from the line of Christopher Blansett from JPMorgan. Christopher Blansett - JP Morgan Chase & Co: Quick one, Ernie. You have a restructuring charge in the statement. I don't think you mentioned that. So first, can we figure out what that is?

Sure. We have a restructured facility that was restructured, actually, probably, the better part of the year ago that's covered under synthetic lease. And as a result of that, we're obliged to do an assessment against our residual value guarantees for the lease. And as we did that assessment, given market conditions, we took the $17 million charge. It was non-cash. And we think that, that writes the building down to the market values that we could reasonably anticipate over the remainder of the lease.

Thank you. And I would now like to turn the conference back over to Shanye Hudson.

Thank you. Thank you all for joining us today. The audio replay of today's call will be available on our website later this afternoon. And that concludes our call.