KLA Corporation (KLAC) Q2 2013 Earnings Report, Transcript and Summary

Good afternoon. My name is Candice, and I will be your conference operator today. At this time, I would like to welcome everyone to the second quarter fiscal year 2013 conference call. [Operator Instructions] I would now like to turn the call over to Ed Lockwood with KLA-Tencor Investor Relations. Please begin your call.

Thank you, Candice. Good afternoon, everyone, and welcome to our conference call. Joining me on our call today are Rick Wallace, our President and Chief Executive Officer; and Mark Dentinger, our Chief Financial Officer. We're here to discuss first quarter results for the period ended December 31, 2012. We released these results this afternoon at 1:15 p.m. Pacific Daylight Time. If you haven't seen the release, you can find it on our website at www.klatencor.com or call (408) 875-3600 to request a copy. A simulcast of this call will be accessible on demand following its completion on the Investor Relations section of our website. There, you'll also find a calendar of future investor events, presentations and conferences, as well as links to KLA-Tencor's SEC filings, including our annual report on Form 10-K for the year ended June 30, 2012, and our subsequently filed 10-Q reports. In those filings, you'll find descriptions of risk factors that could impact our future results. As you know, our future results are subject to risks. Any forward-looking statements, including those we make on this call today, are subject to those risks, and KLA-Tencor cannot guarantee those forward-looking statements will come true. Our actual results may differ significantly from those projected in our forward-looking results. More information regarding factors that could cause those differences is contained in the filings we make with the SEC from time to time, including our fiscal year 2012 Form 10-K and our current reports on Form 8-K. We assume no obligation and do not intend to update those forward-looking statements. However, any updates we do provide will be broadly disseminated and available over the web. With that, I'll turn the call over to Rick.

Thanks, Ed. Thank you all for joining us on our call today. We're pleased with the company's performance in the December quarter, with bookings, revenue and EPS results for Q2 coming in at the upper end or above the range of guidance and marking 2012 as another year of market leadership, differentiated financial performance and high stockholder returns for KLA-Tencor. New orders in the December quarter were $760 million, an increase of 50% compared with the September quarter, with strong foundry and logic demand. Second quarter revenue was $673 million, and non-GAAP EPS was $0.63 per share. For calendar year 2012, KLA-Tencor delivered revenue of $3.1 billion and non-GAAP earnings per share of $4.23. In calendar 2012, we also returned $521 million to stockholders in the form of stock repurchases or dividends. Our dividend payments of $250 million in calendar year 2012 reflect the July 2012 increase in the level of our quarterly dividend to $0.40 per share. Turning now to our view of the current business environment. Consumer demand for mobility continues to be the primary force behind semiconductor industry demand, with strong unit growth in the mobility markets, competition and increased cost and complexity fueling our markets today. For KLA-Tencor, the Q2 bookings profile featured another period of strong foundry demand, with foundry customers accounting for 2/3 of the new orders in the period. The foundry markets are being driven by mobility. And though current industry projections are for foundry spendings to be flat to slightly down in 2013, the market leader in foundry is forecasting an increase in CapEx for the year, and overall foundry investment remains at a high level. The sustained high level of foundry investment we have experienced in this cycle is favorable to KLA-Tencor and a key contributor to our strong relative performance, as foundries tend to be the highest adopters of process control. Logic was 16% of bookings in December. The logic markets are in a period of transition as growth in mobility markets crowds out the traditional PC markets. We believe the demand for logic customers should strengthen in the first half of the year, reflecting the important role process control plays in improving yields and driving device innovation at the leading edge. Finally, orders from memory customers grew to 17% of total orders in December. We anticipate higher memory bookings in March, but we do not expect memory demand to improve meaningfully until the second half of calendar year 2013. The driving force behind KLA-Tencor's market leadership continues to be our focus on the customer. As our customers execute their investment strategies at the leading edge, each successive generation of new technology is characterized by increased process complexity, shorter market windows and more challenging yield requirements. Process control is on the critical path to successful management of these complex, technical challenges. And as the market leader, KLA-Tencor is positioned to continue to benefit from strong relative adoption of inspection and measurement technologies as our customers move ahead with their advanced technology road maps. Now for some perspective on our broader outlook for industry CapEx for calendar 2013. As we look ahead to 2013, even though global economic and political uncertainties and cyclical factors present a headwind for growth, we believe the industry demand environment has stabilized, and we see 2013 setting up to be another period of high capital intensity and featuring relatively high foundry and logic investments as a percentage of the total. Our current yield is for the overall industry CapEx to be down in the range of 5% to 10% in calendar 2013, with industry order levels stabilizing in the first half of the year and demand momentum building in the second half of the year. Against that backdrop, we believe KLA-Tencor is well-positioned to continue to benefit from demand trends that have been favorable to the adoption in our core process control markets and help to strengthen our market leadership. Turning now to guidance for the March quarter. Gross bookings are projected to be in the range of $700 million to $850 million. March quarter revenue is expected to be in the range of $690 million to $750 million and non-GAAP earnings per share in the range of $0.76 to $0.96, assuming an effective tax rate of 18% in the March quarter to account for the reinstatement of the R&D tax credit. With that, I'll turn the call over to Mark for his review of the numbers. Mark?

Good afternoon, everyone. As most of you know, we present our income statement in 2 formats, one under U.S. GAAP and the other in a non-GAAP format, which excludes amortization and write-down of intangible assets associated with acquisitions, restructuring-related charges in credits and any cost of credits which are outside of our core operations, including unusual tax items. There is no difference between this quarter's GAAP and non-GAAP EPS because excluded pretax expenses, which increased EPS by $0.02 after taxes, were offset by a $0.02 per share discrete tax item. Balance sheet and cash flow statements are presented in GAAP format only. Most of my prepared remarks on operations will refer to non-GAAP information, but where I mention GAAP numbers, I'll make the distinction. A reconciliation of our GAAP to non-GAAP income statement is attached to our press release and available on our website. Q2 new orders were $750 million, up sharply from $506 million in Q1, and Q2 net orders were $740 million. The regional distribution of new systems orders in the quarter-to-quarter change in distribution were as follows: the U.S. was 33% of new systems orders in Q2, down from 43% in September quarter; Europe was 1% of new systems orders, down from 13% in Q1; Japan was 7%, down from 13% last quarter; Korea was 13%, down from 20% last quarter; Taiwan was 39%, up from 7% last quarter; and the rest of Asia was 7%, up from 4% in Q1. The distribution of new orders by product group and the quarter-to-quarter change in distribution were as follows: wafer inspection was 48% compared with 51% last quarter; reticle inspection was 10%, up from 1% last quarter; metrology was 19%, up from 15% in the prior quarter; our non-semi businesses were 3%, down from 4% last quarter; and service was 20% in new orders in Q2, down from 29% last quarter. Finally, for semiconductor systems, the distribution of new orders by segment and the quarter-to-quarter change in distribution were as follows: 67% of new systems orders in Q2 were from foundry customers compared with 54% in Q1; logic customers were 16% of new orders in Q2, down from 30% in Q1; and memory orders were 17% in Q2 versus 16% last quarter. Looking forward, we expect new orders for fiscal Q3 will be within a range of $700 million to $850 million. In Q2, we shipped $673 million versus $686 million last quarter. The shipment numbers include both system shipments and services revenue, and we expect shipments between $690 million and $750 million in Q3. Total backlog at the end of Q2 increased by $65 million from September 30, and we ended the quarter with almost $1.1 billion systems backlog. The backlog at December 31, 2012 included $252 million of revenue backlog or products that have been shipped and invoiced but have not yet been recognized as revenue and $837 million of systems orders that have not yet shipped. Total revenue for Q2 was $673 million, down 7% from $721 million last quarter. Systems revenue in Q2 was $523 million, and services revenue was $150 million, which is a record for KT. Our expectation for total revenue in Q3 is a range between $690 million and $750 million. Non-GAAP gross margin was 55.1% this quarter, down from 56.6% last quarter. The largest contributors to the quarter-over-quarter decline in gross margin percentage were an increase in inventory reserves associated with several product transitions and the continuing modification of our product -- production and service inventory management processes to keep pace with our customers' rapidly changing requirements. For Q3, we are expecting gross margins between 55.5% and 56.5%. Operating expenses were $214 million in Q2, slightly higher than the $211 million in Q1. Most of the increase in operating expenses was due to higher research and development expenses. Selling, general and administrative expenses were roughly flat with Q1. We expect operating expenses in Q3 to be up between $2 million and $6 million from Q2. OIE was a net $8.4 million expense in Q2, down about $1.6 million from Q1. Most of the quarter-over-quarter decline in net OIE expense was due to a gain on the sale of a nonoperating investment. For modeling purposes, we expect OIE to be a net expense between $9 million and $10 million in Q3. In Q2, our non-GAAP income tax expense was $43 million or 29% of pretax income versus a 24% rate in Q1. The Q2 rate increase was largely a function of a change in the estimated distribution of earnings between our U.S. and international locations. We anticipate our rate will decline significantly in Q3, following the reinstatement of the R&D tax credit in January. And our best estimate for the non-GAAP tax rate for the fiscal year is about 24%. Non-GAAP net income was $106 million or $0.63 per share in Q2. Using our standard 24% tax rate, our Q2 EPS would have been $0.67. At the revenue range I previously mentioned in defining our planning tax rate of 24%, we would expect our Q3 non-GAAP earnings to be somewhere between $0.70 and $0.90 per share. At the lower tax rate we are expecting for Q3, our EPS range will be somewhere between $0.76 and $0.96. Weighted average share count used to compute EPS in Q2 was 169.1 million versus 169.8 million in Q1. During Q2, we spent $68 million repurchasing about 1.5 million shares. In November, our board approved an additional 8 million shares for our repurchase program. And as of December 31, 2012, we had approximately 8.4 million shares available under our current authorization. For guidance purposes, we are modeling an average share count of about 168 million for Q3. We also paid $67 million in dividends in Q2. We anticipate continuing to repurchase shares, as well as paying a quarterly dividend of $0.40 per share in Q3. On our balance sheet, cash and investments ended the quarter at $2.6 billion, about even with September 30, 2012. Cash generated from operations was $77 million in Q2 compared with $245 million in Q1. Net accounts receivable ended the quarter at $606 million, up from $537 million at the end of September. DSOs were 82 days at December 31 versus 68 days at September 30. Both DSO figures are net of allowance for uncollectible accounts and factorings. Net inventories were down $27 million from Q1 and ended the quarter at $663 million. Inventory turnover based upon GAAP cost of revenues was 1.8 turns in Q2 versus 1.9 turns in Q1. Capital expenditures were $17 million in Q2, down from $20 million in Q1. Total headcount at December 31, 2012 was 5,816, essentially flat with September 30. We expect our headcount to remain about flat in Q3. In summary, our guidance for Q3 is new orders between $700 million and $850 million, total revenue between $690 million and $750 million and non-GAAP earnings between $0.70 and $0.90 per share applying a 24% tax rate and between $0.76 and $0.96 per share using our anticipated Q3 tax rate. This concludes our prepared remarks on the quarter. I will now turn the call back over to Ed to begin Q&A.

Okay. Thank you, Mark. At this point, we'd like to open the call up to Q&A. [Operator Instructions] So Candice, we're ready for our first question.

[Operator Instructions] And your first question comes from Satya Kumar with Crédit Suisse. Satya Kumar - Crédit Suisse AG, Research Division: A couple of them. First off, I think previously, Rick, you were saying that you thought CapEx will be down, closer to 10%, and now you're thinking around 5% to 10%. What segments of the 3, between logic, foundry and memory, has improved the most? And in this environment of down 5% to 10%, what do you think about the outlook for KLA in 2013?

Yes, we did see some strengthening recently. Primarily, it looks like foundry is a little bit better than we felt when we had talked about the down 10% to 15% based on some recent news that we've seen. What that means for KLA-Tencor is positive because we do all in foundry and the needs for process control are only increasing as the transition happens to 20-nanometer, which is what's going on in 2013. So we anticipate that we will outperform the industry this year, as we have in the last couple, based on the increasing intensity of process control, plus the mix of customers. Satya Kumar - Crédit Suisse AG, Research Division: All right. On gross margins, it was a little light for December related to your guidance and given your -- what came in at the higher end for revenues. I understand the quarter-to-quarter variation, but could you help us go through the effect of the product mix, customer distribution and competition related to your gross margins?

Yes, Satya, this is Mark. It has less to do with the customer distribution than it simply had to do -- we are going through a pretty active period of product transitions right now, and we built up the inventories somewhat purposely to accommodate that. And when we do, do that, you do get some excess and -- excess demand outcome as a result of that, and we had to take some reserves. Now historically, we tended to get some of those back, but that does take a little bit of time to come back through the system. And we do expect it will digest these and then return to a more normal level in the next quarter or 2.

Your next question comes from Terence Whalen with Citibank.

The first question is related to general foundry activity. In terms of the strengthening that you've seen, has that been in the form of pull-ins by customers? Or has it been just an inflation and overall forecast?

Terence, yes, it's Rick. I think that it's associated with 2 things. One is, I think there is more broad demand, so it's not simply with one customer. We're seeing some strength in others as well. And we do see 20-nanometer. I think the understanding of the challenges associated with 20-nanometer are playing it out. So we believe it's likely that they'll be a differential investment in process control as a result. So I think it's both factors at work.

And then the follow-up is to take that a little bit further. When you look at the 20-nanometer node, including 16 or 14 variations, more looking at it planar versus FinFET, is there any way to help us understand what perhaps the requirement increase would be moving to FinFET from planar?

It's a great question. Unfortunately, it's a rather complicated answer. I think it depends upon a couple of factors. One is, if the companies that are moving forward, in general, have an experience in 3D, then their process control perhaps is better understood and they might be a little less intensive than others that haven't. But we see, in general, increased demand in metrology as one area certainly where 3D drives demand, especially in product lines such as our optical CD and our films measurement tools. We also see a different kind of defectivity concern, which drives adoption, in our case, at the very high-end brightfield inspectors, which is necessary as people are trying to deal with the aspect ratios and the defectivity. We haven't really quantified exactly how much is the increase between 28 and 20. We already see some of the demand. We'd say that probably 1/3 of our orders today are associated for 20 or below. So we're already seeing that start to play out, but we think it will accelerate through the rest of the calendar year as people look at adding more capacity for the 20-nanometer node. Beyond 20, the question of 16 or 14, depending on how you define it, we think there's still investigation as to what's going to be needed, and we're working very closely with customers on that.

And your next question comes from C.J. Muse with Barclays.

I guess, first question, Rick, you talked about stabilization of orders in the first half and a pickup in the second half. And I'm just curious, you identified memory as a place that would pick up. But also, I would love to hear your thoughts on how you see foundry trending into the second half, as well as given that large budget we heard from a company down the road from you, how we should see that playing through the model as we exit the year.

Sure, C.J. I think that when we looked forward, we actually saw some increase -- slight increase. And memory, actually, as a percent of our whole, went up a little bit in the December quarter, and we're anticipating that to continue to grow as a percentage of our total. So we'll see some memory pick up even in March but more pronounced in the second half, and we think that's driven more by technology investment and some capacity in NAND as we get to the second half. Logic, pretty steady in the last couple of quarters. I shouldn't say steady, I should say lumpy but kind of a steady average run rate. We see it at about 25%, 26% for the March quarter and then continue investment through the year. Foundry, actually, we anticipate coming off as a percent, and more in the 50% to 52% range is what we're snapping right now when we look ahead, which is down from the 67% we did. As we go through the year, foundry, I think, is somewhat dependent on overall what happens with 20-nanometer and how many players are really in -- and expanding that. We know there are some activity and some customers, but how that broadens out and the timing of that will influence, I think, whether it's flat or slightly down for the year. We do see some memory strength in logic, I think, pretty steady, maybe some increase in logic, but again, we'll have to see how the year progresses.

Very helpful. And I guess as a follow-up, on the foundry side, Morris [ph] spoke to you, 20-nanometer being a bigger node than 28. And I think there's some confusion as to whether he was talking revenues or clearly wafer prices are much higher or wafer starts. We'd love to get your thoughts on what that kind of node could be like relative to 28 and whether you would include within that the 16- or 20-nanometer FinFET.

Right. I think it's a big question right now. I think from many of the established guys, they're trying to figure out if they are going to lean in to 20 heavily or if they're going to wait for 16 or 14, depending on what they see as end-demand. I think the other challenge that people have is what the manufacturability is of 20 and how the prices come down. There's certainly a lot of interest. I think that the initial skepticism about the number of customers has reduced. Frankly, I think there's more interest than there was. As far as whether it ends up being bigger than 28, I'm really not in a great position to answer that. I do know that from a process complexity standpoint, there's certainly a lot of concern about the number of processes that are changing, and that's what's driving a lot of our business. But I don't think we're in the best position to answer the actual magnitude of the capacity. I think it depends on a lot of factors. But we certainly see a lot of challenges associated with getting the process under control both from a defectivity standpoint and a process -- overall process control from metrology. So with all the changes with APL, 3D structures and the general scaling challenges, it should be a robust environment for process control and for KLA-Tencor.

Your next question comes from Jim Covello with Goldman Sachs.

First of all, a cycle question. You guys have been very clear on the cyclical outlook, but the one thing I'm just trying to figure out for modeling purposes. [indiscernible], obviously, you have a very robust CapEx budget. They said of their stated budget, 2/3 of it was first half-loaded and then 1/3 second half. But they certainly held out the possibility that things would pick up in the back half of the year if they did some more for Apple. Incorporated in your guidance for the year in terms of that pickup in the back half, does that assume TSM is coming down on the back half or foundry TSM in particular? Does that assume kind of the more robust budget that they have or maybe the more conservative budget? In other words, is there upside to your forecast if they do wind up spending a little more relative to their commentary around what they would do for Apple?

We're not anticipating anything beyond what they've said publicly, which is a slight increase. I think they've stated 6% up for their CapEx. So the target mostly, the anticipated, what I've heard them talk publicly about is that, that will be first half, more loaded toward the first half. But I think as we all know, the foundries actually review their CapEx investment strategy more frequently than annually for sure in this environment. So I think it's very hard for us to speculate because our customers don't actually know what they're going to end up doing. We do know that the change, the volatility, the upside of all these great foundry business has been that there's good business for KLA-Tencor and we continue to perform very well in this market. The challenge is that as a result of the volatility and the lack of ability to really forecast, we end up, as a result, having to be more responsive, which impacts some of the things like we've talked already around inventory and our flexibility in operations to support it. But I can't really tell what they're going to do because I think, frankly, our customers are nimble and respond to the market as it plays out. Our overall thesis though is we'll continue to outperform the market, and the general consensus we're getting is a better outlook than we had a few weeks ago even and probably in the range of flat to down 5 right now.

That's great. Very helpful. If I could for my follow-up ask, AMAT is once again kind of refocusing on process diagnostics, and this is something that's been around for several decades in terms of them trying to come in and compete with you in this segment. Obviously, they have some new actors that are doing it this time, but they've tried to do this before and it hasn't worked. Do you have any incremental concerns about their ability to penetrate even a little bit of share in this market when you so successfully defended that position so many times over the years?

I don't underestimate any competitor, and I think it's true that AMAT's got a renewed focus in general. It has caused us, I think -- we take note of all competitors. And certainly, we're in a very attractive space, and people are going to try to penetrate. So we're increasing our level of investment, and I'd say that we are very focused on satisfying our customers. And our strategy is to play our own game and play it -- take it to the next level and respond to customers and continue to invest in our road map. But I absolutely take seriously all competitors and have a lot of respect for the guys over there.

Your next question comes from Krish Sankar with Bank of America Merrill Lynch.

Rick, I had a question in terms of metrology. When you look at moving down the technology curve from 28 to 20, is that a trend of the metrology going more in line, especially for the critical steps like lithography where chipmakers are beginning to put it in foundry tools rather than ex situ or you don't see any trends like that?

Oh, I think there are certain cases where it makes sense to have metrology integrated. I think it ends up actually being a question of process control. And when there's a high-enough variation, then you want to reduce the lag time from the measurement to the correctables. And for a long time, for example, some basic measurements have been made in situ, in places like film thickness and CMP for many years. The question is, can you make the measurements that you need to make and get the feedback quick enough? And is it cost-effective? And do you need to make that frequent of a measurement, depending on how the variation is going? But I certainly think people are always looking for ways to improve their process control, and integrated has always been an idea that people have had. It's interesting because I don't think it's truly in situ. In situ would imply that it's actually in the environment. So the real -- there's only a couple of in-situ measurements that happened, and those tend to happen in chambers or in environments. These are tools that are actually linked to the process equipment themselves, whether it's to the -- often in the case of litho, it'd be linked to the tracks. And so I do think that there are opportunities there, and it's always a question of cost effectiveness against the perceived need, against the impact of reliability and uptime. But it's certainly an area where people have been exploring for a number of years. And I wouldn't be surprised if there's more interest, especially as the design rules create more need for more and more control.

Got it. That's very helpful. And then as a follow-up, your memory orders moved up in December, and you're saying it's going to move up further in March. Is it all -- the driver coming from NAND? And if so, what is the breakdown of NAND and DRAM?

A lot of drivers, NAND, but actually, in the December quarter, it was about 50-50. So we did see some investment associated and primarily in technology-related. And I think for a long time, the memory guys, especially DRAM, have been somewhat on hold regarding investments, and now we're seeing some catch-up. There is an outlook that has NAND increasing and maybe even adding capacity later in the year. Our forecast, for what it's worth, for the March quarter is that it bumps up a little bit higher to 60% of the total memory forecasts. So still not back to levels in the past but certainly off of what was a very low number just a few quarters ago.

Your next question comes from Chris Blansett with JPMorgan. Christopher Rand Blansett - JP Morgan Chase & Co, Research Division: Rick, I had a question about equipment reuse since this has been a theme as of late, and how you're seeing that play out in terms of how it affects process control maybe versus other equipment types?

By reuse, you mean extending technology nodes with equipment? Christopher Rand Blansett - JP Morgan Chase & Co, Research Division: Yes.

Yes. It varies. I think it depends on product capability, process change and the availability of tools. I would say that by and large, we do not see a lot of reuse happening, and we actually have set up and offer to sell, even refurbish some of our older equipment. But in the leading edge, the reuse tends to happen in process areas that are not as sensitive, in fact, often not even key inspection or measurement points. So therefore, it doesn't really lend itself because our tools tend to be utilized pretty -- people are very careful about where they implement them, and they tend to implement them on the processes that have the least amount of margin, which is not where reuse happens. So even though years ago we set up a business to support it, it's actually not been extremely active us -- either for us or anyone else because on the leading edge, the things people are pushing are things that need metrology and inspection. So it hasn't impacted KLA-Tencor, and I think it may be another reason why process control will continue to increase as a percent of the overall spend, as people try to leverage the assets that they have. Christopher Rand Blansett - JP Morgan Chase & Co, Research Division: And then my follow-up question was tied to your gross margin. It's come down a bit. You mentioned that some new products coming out, some inventory buildup affecting it. But has there been a meaningful shift in your equipment shipments lately that have affected that? And what would you need to kind of get those back to the high-50s range?

No, there's been no meaningful shift in the equipment mix. This is Mark. Basically, what we're experiencing right now is that we're getting clipped a little bit on excess demand issues, largely in conjunction with building up the inventories to be more responsive to the customer circumstances and to manage the supply lines. And we do believe that, that is, at least -- let's call it a 1-, 2-quarter phenomenon and then it should resettle back into more normal levels. We also are responding the same issue a little bit on our service side as the distribution in the multiyear migration towards the foundries has reallocating balances into the service inventory levels as well. So those 2 phenomena, I think, are underway. It's just a question of how long that -- if you will, that transition is going to be with us. But in terms of just the raw margin, in terms of the -- on the product sales, that does not appear to be a factor.

And your next question comes from Stephen Chin with UBS.

Just a follow-up question, Rick, on the 2013 industry CapEx you use. So have you accounted for the higher CapEx spend by this leading logic customer in your new view? Or is there still possible upside as you kind of sort things out with that customer? Then I have a follow-up.

So, again, we're not -- I don't think we're the best people to forecast this because we're going largely on what's being said publicly. But in order for us to get to a flat or down 5 kind of range, we are accounting for what's been publicly stated. Now we also think that, that's going to change based on history. So yes, it accounts for increases on the 2 major players and the decrease in the other player.

Okay. And just a follow-up on that same topic. Can you share maybe some of your updated thoughts on the timing of KLA's 450-millimeter R&D investment since that same logic customer says it is setting aside a significant amount of CapEx this year or next year for facilities?

Sure. Yes, we've already shipped 450 tools, and we continue to do that. We're shipping primarily to equipment manufacturers. And so some of the first tools that went out for that were -- are bare wafer particle inspection tools. We're actually seeing some demand for other equipment or services that we offer for 450. We have wafers that we make smart wafers that are used in that, in our sensory division that we're seeing some demand for as well. So we are already supporting the equipment world for their preparation for 450. And we will have a rollout over time of the additional equipment as we phase in the next-generation. We're not retrofitting any existing tools. What we're doing is, when we introduce new models in the road map going forward, they have 450 capability in them, by and large, is how we're going to satisfy it. So we believe we're well-aligned to the latest industry thinking on timing. I'll also point out that in the past, the initial entry of 450 or any wafer size change was well in advance of any production. So the challenge, I think, the industry has is satisfying pilot needs, followed often several years later by the time you get to full production.

Your next question comes from Vishal Shah with Deutsche Bank.

Rick, just wanted to understand how I should think about the 20-nanometer capital intensity for process control. Is it a 40% increase that some of the foundry customers have been talking about? Or is it more or less for you guys?

Well, I think it depends on how you quantify the overall investment. I wouldn't want to say that our business is going to go 40% up as a result of the same silicon capability, maybe the same die out but if you take into account shrinks. We are seeing additional points. I think what's important to understand is, the way our customers figure this out is they'll go through pilot, which they're doing, and they'll try to map out where they need, what kind of capability. And that's kind of an ongoing characterization process. And then they will -- as they're ramping, tune that, depending on what they discover in terms of how much process control they actually have over the process, both inspection and metrology, so we anticipate that the transition to 20 has similar characteristics to the transition from 45 or 40 down to 28 in terms of the challenges, the number of process steps, the number of critical layers and, therefore, we think in similar increase in intensity as we go forward. All that being said, I think in the end, that's how we derive the increased performance of process control in KLA-Tencor versus the rest of the capital budgets that people are -- the different sectors inside of WFE.

Appreciate that. And just one other question on the OpEx front. I mean, your OpEx is going to go up a little bit. How should we think about the trajectory for OpEx for the rest of the year?

This is Mark. I think you're going to see OpEx basically in the range of 215 to 220 for the rest of the year. We won't be aggressively adding headcount at this point, but you do get some lumpiness as a result of material bias and whatnot in the R&D lab. So I think that's not a bad range to think about, but again, we'll update you quarterly if things change.

Your next question comes from Mark Heller with CLSA.

I had a -- just a clarification. When you said flat to down 5%, was that for the industry or was that for KLA in 2013?

That was for KLA as a measure against a 5% to 10% down for the industry.

Okay. And then I was wondering if you have any updated view on e-Beam inspection and review on your competitor. And Taiwan has been very vocal about how optical inspection will reach limits or barriers at sort of the 10-nanometer node. So I was just wondering if you have any updated view on the e-Beam market.

Well, I think e-Beam is essential, I think that -- particularly for a review. I think that the ability to be able to see the defects that you're catching, you can catch defects without resolving them, but it's very hard to take corrective action without better insight into what the defects are, which is why the primary application for e-Beam is in review. Inspection offers some advanced capability but has their limitations. So we offer inspection as well and have for quite a number of years, trying to meet the needs in some niche places. The challenge for e-Beam inspection, much like the challenge, frankly, for e-Beam and lithography, is throughput and the ability to target the e-Beams to get them effectively, to get enough down onto a wafer and return so that you can image them and determine the defectivity leaves you at least 100x gap versus optical. And we don't see anything that's going to limit capability of optical inspection at 10-nanometer below. It will continue to get more challenging and expensive as it has. But inspection -- optical inspection tends to lag lithography by -- I think of it as 1 to 2 generations. So you can envision, if you're doing EUV-like lithography, you can optically inspect that with what is essentially 200-nanometer, sub 200-nanometer optical inspection, so you don't have to be at the wavelength. e-Beam has, as I said, some attributes that are highly valuable, but it has many challenges, not just throughput but also associated with the kind of nuisance effects that the e-Beams create and the charging. So we do have active programs. We believe there is a role for it. We've been investing in for a long time. I just don't see it replacing optical as the core technology for inspection.

Your next question comes from Mahesh Sanganeria with RBC Capital Markets.

Rick, you're sounding a little bit more positive on memory, definitely, compared to the last call. And -- but your CapEx number has not gone up significantly. Could you give us a little bit more color on what your thoughts are on NAND and DRAM? Are they -- year-over-year, are they up? And what's the driving -- what's driving the spending?

Yes. I don't see them being year-over-year up. Actually, when we're modeling it, we have them flat to slightly down but second half-loaded, primarily associated with DRAM capacity that comes on, depending on what you anticipate for results. We actually have seen some positive momentum in our space. But overall, for CapEx, we're not modeling memory up, in fact, slightly down or flattish, if you will. And I think that the benefit from some of the recent integration of some of the M&A out there was we have customers that are out now making investments in advanced inspection capabilities in order to enable them to do technology migration. So that's where we're seeing a benefit.

And the follow-up is on reticle inspection. So we saw the bottom in the order in the last quarter and we see some pickup. But where are we in that particular cycle? And what will drive the acceleration to the last peak on the reticle inspection?

I think reticle is going to be lumpy going forward. I don't think there's -- that there is not a huge demand being created at the leading edge based upon the work that's going on in double patterning. And there is a lot of high-end capability that's been served already. So we have basically done a great job of penetrating the leading edge with our leading technology. So most of the activity tends to be more in the fab lines for recall, which is not a big -- as big a market and is something that, of course, we're focused on. But I don't think I'll provide the same size that we saw when we were launching and introducing the 6xx. Compensating for that is the increased opportunity in process control around multi-patterning. So the corresponding decline overall in the reticle is made up by increases and opportunities in overlay. So I think that on balance, it ends up actually looking pretty comparable but a shift of products based on the technology. When it will come back is when we see or if we see advanced EUV really take off, then there'll be a lot of opportunity in reticle again. But right now, that seems to be further in the future than it was just a few months ago.

Your next question comes from Jagadish Iyer with Piper Jaffray.

Rick, two questions. First, how should we think about your products -- product mix as spending pattern changes given the foundry and logic spend much more compared to last year?

Not really very different from last year in terms of overall products mix for us. I mean, we are seeing that, basically, if you look at overall CapEx, you could have different models. But let's say you say down 5%, foundry is flattish to down 10%, logic is probably flattish. So the impact to us is that not a lot changes in the customer mix or the kind of process controls being utilized, except for the more advanced process control because we're a further along in the technology road map a year later.

Just on a follow-up. How would you characterize at least on a qualitative basis the 20-nanometer ramp versus the prior 2 nodes in terms of achieving acceptable yields? I'm just trying to understand the duration of the yield ramp versus the prior 2 nodes.

I think our customers are focused on accelerating faster. They certainly get a lot of competitive pressure in the market. They recognized -- certainly, the fabless companies have recognized that first to market wins. So when I talk to our customers about it, they're expecting and planning to ramp faster than they did on 28. And I think most of the customers recognize the challenges that they faced in 28. So that's why I think we're seeing increased interest in process control on KLA-Tencor equipment to facilitate that. But every technology node, our customers say that they want to ramp faster, but this one is coming on the heels of what was pretty challenging for many ramp. So I think that there's increased heightened sensitivity to that.

Your next question comes from Patrick Ho with Stifel, Nicolaus. Patrick J. Ho - Stifel, Nicolaus & Co., Inc., Research Division: Rick, given the limited capacity buys on the memory side today, where's the focus been, I guess, on the technology buys in terms of memory? Are you seeing more orders in terms of 3D NAND flash? What's driving it, I guess, in the near term?

There's certainly development going on around 3D, as you know, and I think there's -- that creates opportunity for us. There's definitely interest, in general, in pushing patterning further, but I think the NAND guys are struggling a bit with that. So we're seeing overall increased focus on supporting technology migration, and that is certainly driven, I'd say, right now in NAND by the 3D work, although we don't expect to see production for that until probably late in this year. [Audio Gap]

So, Rick, I have a question regarding your logic versus your foundry customer regarding what kind of option you get from these guys. So the foundry customer typically have a lot more different type of chips going for the lines. Do you see that as a result of that? Do you think that the foundry customer have a greater opportunity? And anyway you can kind of quantify it, for example, on a node-to-node basis or maybe the logic guy is the only one who know ahead. Is the opportunity similar? Is it 10% more or 20% more? Any way you can help us with that?

Pretty similar. I mean, I think that it depends. There are different logic manufacturers. Obviously, they are market leaders in that. But by and large, the question of high mix versus low mix is -- does have an impact, but often, it's offset by the people who have lower mix tend to be pushing technology perhaps a little bit harder. So that kind of counterbalances it. And I think in the end, there's probably more variation across different companies than there is across -- for example, foundry to logic is probably -- if you looked at all the data and didn't know if it was foundry or logic in terms of how people adopt and what their strategy is, you'd see actually more variation in the foundry space per se than you would just by going -- jumping between foundry and logic. Does that make sense?

Okay. All right. I think I get what you were saying. A follow-up question also on the NAND side. I think yesterday, SanDisk talked about it, one, why node started to see increased capital intensity and, therefore, they have to spend more. Do you think you guys benefit from that, 1y versus 1x, or the kind of very leading edge on the NAND side?

I think that there is definitely challenges that create different opportunities for us. And so we often talk internally. If we can solve our critical customer problems that creates opportunity for us, return for them, so any time there's a shift in technology, it creates opportunities. And often, it obsoletes existing capability, although we'll try to reapply whatever we can. But certainly, as people go in the y-dimension, one of the biggest challenges they have is, especially when there are many, many process steps that are in a closed environment, it pushes things like qualifying the tools very hard, which interestingly plays to our bare wafer capability because you wouldn't necessarily think it. But it's a way to qualify tools so that making sure that they are not contaminating at all as they're going through these multiple steps. It also creates opportunities for people looking for defects with optical tools and then trying to figure out what they are with review tools. So it helps us in that regard. So we do see opportunities, and it does cause us to have to modify some of the technology, which means new tool sets in order to do that often, or pushing, as I said in the bare wafer case, pushing existing capability even harder.

Your next question comes from Ben Pang with Caris & Company. Benedict Pang - Caris & Company, Inc., Research Division: On the December orders that you came in above your guidance, and you had a pretty wide range of guidance, was that entirely due to foundry? And if so, was that more 28- or 20-nanometer-focused?

Well, given that foundry was 2/3 of our orders, then obviously, foundry was a big player in terms of how the bookings came out. We've widened the range of guidance often because there is a customer concentration. So you never quite know if deals are going to happen, when they're going to happen. But the foundry space, as we said, about 1/3 of our business in foundry is around the 20-nanometer work and the rest is 28. So I'd say that foundry was the larger influence. But there was stuff that came in at the end of the quarter, which we were not expecting, which is how we ended up above the range of guidance for bookings. And that was actually across the board. We saw some memory business come in that we didn't anticipate, we saw some logic come in and we saw some foundry. So it was generally strengthening at the end of the year, which I take to be a very good sign. Benedict Pang - Caris & Company, Inc., Research Division: And I have one follow-up. At SEMICON, you had a graph that showed a die-size impact for mobility products at the different node changes, pretty big increasing die size. Are your customers expecting the same thing between 28- and 20-nanometer for the same number of cores?

Well, there's certainly -- I mean, part of the reason to go to 20 is to not have the die size go up. So you have your choice of either increasing the functionality on the same size or actually reducing it. But the only point was that what's been fascinating about mobility is the notion that people would be -- have good enough performance and stop innovating, and we've definitely seen that, that's not the case. So as people go to 20, they're definitely going to push very hard on shrinking. The question is, will they then add multiple cores or will they expand to get more performance? And I think certainly, in some cases, that's going to be the case. And then you also have graphic dies, for example, that try to use these larger die as they can to provide the most functionality. So we're seeing both, but I'd say that it's a similar phenomenon to what we saw in microprocessors in the '90s, where die not only got more complex but got larger over time.

And our last question comes from Weston Twigg with Pacific Crest.

I just had a quick one on ASML and metrology. So they even talked about this on their last call, but they're beginning to use closed-loop metrology to help improve their litho process window, especially as the double patterning becomes more prevalent. And it sounds adjacent or maybe even complementary to stand-alone KLA tools, but I'm wondering if there's anything here we need to watch relative to metrology and how it's used for litho and if there are any emerging opportunities or risks for KLA.

Well, I think that there are definitely reasons for people to try to increase the capability in overlay, and certainly, the idea of having more functionality closer to the process can make sense. What we're seeing right now is that the overlay market has expanded quite a bit. I do think customers are going to try to pursue alternatives to get their process under control, first and foremost, and then to manage their costs. So we always anticipate any change like that to be one that could potentially displace us unless we continue to add more value and create more capability for our customers. And so we take it seriously, we take any alternative approach seriously and it drives us to create more capability in our products and continue to innovate and continue to invest in them. So we're the leaders in overlay. We think we've got a great position to build upon. But we certainly look at the opportunity associated with the challenges of process control to also be a potential disruptor, and we've got to continue to make investments and support our customers to optimize our position.

Okay. Would there be an opportunity to actually integrate into a litho tool? Or is that something that's more or less off-limits?

Well, it wouldn't be a litho tool, it'd be a track. The integration doesn't really happen on litho because you have to have developed -- people have tried to do latent image measurement. It's just very hard to do and has been unsuccessful in the past. So the opportunity actually happens in the track. And yes, there is an opportunity to do that. But they're integrated and we've been looking at integrated for many, many years. And there are some advantages and then there are some challenges. The real question is, is the rate of change fast enough to warrant dedicating one tool per one track, which is what it would take? And there are customers that are experimenting with that, but there are those that have concluded that, that's not the best way to go. But we're certainly in close conversations with the people that are interested in that capability to provide that if that's the path they choose to go down.

And there are no further questions at this time. Let me turn the call back to our presenters.

Thank you, Candice. I'd like to thank everyone on behalf of the management team for joining us here today. An audio replay of today's call will be available on our website later this afternoon. And once again, we appreciate your interest in KLA-Tencor. Thank you.

And this concludes today's conference call. You may now disconnect.