Lam Research Corporation (LRCX) Q1 2013 Earnings Report, Transcript and Summary

Good day, ladies and gentlemen. Thank you for standing by. Welcome to the Lam Research Corporation September 2012 Quarterly Results Conference Call. [Operator Instructions] I would now like to turn the conference over to Shanye Hudson, Director of Investor Relations. Please go ahead.

Thank you, Britney. Good afternoon, everyone, and welcome to Lam Research Corporation's quarterly conference call. Here with me today are Martin Anstice, President and Chief Executive Officer; and Ernie Maddock, Senior Vice President and Chief Financial Officer. Shortly, Ernie will discuss financial results for the September 2012 quarter, and Martin will then share Lam's business outlook for the December 2012 quarter before we open the call for Q&A. The press release detailing our financial results was distributed over wire services shortly after 1 p.m. this afternoon and is also available on our website at lamresearch.com. Today's call contains certain forward-looking statements including those related to our expectations for the global macroeconomic environment; market size; wafer fabrication equipment spending; the revenue expectations of shipments to our Japanese customers; technology trends and transitions that may affect our business; market share changes; consumer demand; customer spending and behavior and the factors that will influence those expectations, as well as our spending projections; our investment plans; our business strategies; our expectations of the benefits resulting from our Novellus acquisition; our intentions for research and development activities; our contemplated tax rate and our forecast of market shares, shipments, revenues, expenses, margins, operating profits, share repurchase activities, earnings per share and cash generation, both on a GAAP and on a non-GAAP basis, 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 as well as on our most recently filed Form 10-K 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 3:00 p.m., and we ask that you please limit questions to one per firm with a brief follow-up. With that, I'll turn the call over to you, Ernie.

Thank you, Shanye, and good afternoon, everyone. I'm pleased to report good results for the quarter ended September 23. All metrics mentioned on today's call are for the combined company. And as a reminder, the September quarter results include a full quarter of merger-related activity, while any June quarter comparisons reflect only 20 days work. Total shipments were $935 million for the September quarter, up 15% from the June quarter and within our previously guided range. As expected, we did see a slight broadening of customers across multiple segments. And to that point, for the September quarter, approximately 20% of our system shipments were for technology nodes greater than 4X, while the remainder of approximately 80% were for applications at the sub-4X technology node. Segment data is as follows: Foundry shipments accounted for 48% of total system shipments, while the memory segment accounted for 42% of total system shipments, which include NAND at 22%; DRAM at 19%; and other memory at 1%. Constituting the balance, logic and other were 10% of the total system shipments. Revenue in the September quarter was $907 million, up 22% sequentially and in line with our expectations for the quarter. Non-GAAP gross margin was 44.4%, up 2.3 percentage points from the prior quarter and exceeding the high end of our guidance range, primarily due to solid operational execution and a more favorable product and customer mix. Non-GAAP operating expenses for the quarter were approximately $284 million, up sequentially from $215 million in June. A more instructional comparison would be the performance of the 2 companies on a stand-alone basis during the March quarter where combined expenses totaled $293 million. Between $5 million and $6 million of the reduction relates to the achievement of OpEx related synergies, while the remainder relates to lower stock compensation expense associated with merger-related acceleration of equity awards. The combination of higher gross margins and lower operating expenses resulted in non-GAAP operating income of $118 million and non-GAAP operating margin of 13%, exceeding the high end of our guidance range. Relative to our expectations for the quarter, non-GAAP other expenses of $2.2 million was favorably impacted by foreign currency adjustments and interest income. Our non-GAAP tax rate for the quarter was 16.3% compared to 16.9% in the June quarter. And presently, we're still expecting a fiscal year non-GAAP tax rate in the mid-teens. Should the federal R&D tax credit be extended prior to our fiscal year end, this rate would be reduced to the lower teens. Based on a share count of approximately 182 million shares, September quarter non-GAAP earnings per share were $0.53, exceeding the high end of this quarter's guidance by $0.06. Moving now to the balance sheet, we ended the quarter with cash and short-term investments, including restricted cash and investments of $2.9 billion versus $3 billion in the prior quarter, with our cash generation for the quarter being offset by our ongoing share repurchase program. Day sales outstanding for the September quarter were 64, and inventory turns were 4.0. As you know, we did not report these metrics for the June quarter as the 20 days stub period related to the Novellus transaction caused these metrics to be unhelpful. During the September quarter, these metrics for the combined company are aligned with normalized levels. And for comparison, DSO for the March quarter for Lam stand-alone were 65 days, and the inventory turns were 4.2. We ended the quarter with deferred revenue of $364 million, excluding approximately $21 million in the shipments to Japanese customers that will revenue in future quarters. Cash from operations was $249 million or 27% of revenue in the September quarter. And we currently expect that as a percentage of revenue, CY '13 cash from operations will approximate CY '13 non-GAAP operating income. Combined company noncash expense includes, among other items, $24 million for equity compensation, $45 million for amortization and $30 million for depreciation. Capital expenditures were $44 million, and we exited the quarter with approximately 6,600 regular full-time employees. Lastly, I'd like to provide an update on our ongoing share repurchase activity. During the September quarter, we spent approximately $344 million repurchasing shares and took delivery of about 12 million shares of common stock, which includes just over 2 million shares received in conjunction with our June quarter expenditures under structured repurchase programs. As indicated previously, our activity during the quarter was more systematic in nature, and we now expect to complete a significant majority of our $1.6 billion share repurchase during 2012. With that, I'll turn it over to Martin.

Thank you, Ernie, and good afternoon, everyone. Despite well-publicized adverse market conditions recently and the associated ongoing cyclical pressures, Lam delivered on all key customer and financial commitments for the September quarter. These achievements reflect incredible levels of commitment from our employees worldwide, continued focus on operational execution and the unwavering focus on our customers and our strategic growth objectives. We're very pleased with our progress. First today, I'll briefly update our views on the industry. Since our last call, a number of customers lowered their investment plans for the balance of this calendar year. Accordingly, we have further lowered our 2012 wafer fabrication equipment spending outlook to a $28 billion to $29 billion range. Having essentially met September guidance, this adjustment impacts the December 2012 quarter significantly, particularly in the area of shipments. Again, the NAND segment accounted for the vast majority of this latest series of spend reductions as manufacturers continue to react to the pricing environments and their view of supply and demand. Given the decrease in new capacity additions planned for the balance of calendar '12, we think it is likely they exit this year with very tight NAND device supply, which bodes well for a resurgence of spending in calendar '13, probably in the first half. Clearly, there is some evidence we are already seeing the impact of this supply situation in NAND ASPs, which has started to increase over this past quarter. This is an important reference point for confidence in future capacity expansion. While it is still too early to accurately forecast 2013 WFE spend to any significant detail, our initial analysis still supports the prospect for spending, once again, to be within a range of $30 billion. Smartphones, tablets and SSDs support the need for new NAND capacity and additional leading-edge foundry and logic investments. These drivers are also positive for mobile DRAM demand. However, projections for PC unit growth and DRAM content growth continue to be modest, resulting in another year of limited DRAM equipment spending, most likely. Considering the fundamentals of WFE capacity and our understanding of demand for electronics products, we would characterize the 2013 WFE spend to have more potential for upside than risk. Needless to say, the often unrelated macro influences independent of the semiconductor cycle are significant to the risk and upside both. As you might expect in this environment, Lam is even more focused on executing well, managing short-term volatility but staying the course on the primary elements of our longer-term market share growth objectives. We'll spend more time in our upcoming Analyst Meeting in November, focused on communicating progress to date on these long-term items, characterizing their magnitude and the probability of our excess -- of our success exploiting market opportunities and technology inflections. In calendar year 2013, we expect customers to begin ramping capacity additions for the n+1 technology node, which includes the transition to 40 [ph] -nanometer for logic and 20-nanometer for foundry. Potential 3D NAND pilot line investments could also begin late in the year. But primarily, we are assuming a planar-based memory WFE for now. For Lam, these transitions represent key opportunities and support achieving our stated market share growth objectives. We're pleased with our progress. It goes without saying that successfully defending existing positions is the most critical foundation of market share sustainability. Building off that foundation with targeted application penetrations to gain share is our proven growth model. In etch, clean and deposition, we continue to defend positions well. We believe that approximately 70% of the critical n+1 technology node tool selection decisions have been made at this time. And on that basis, we would estimate our application-based market share for etch, clean and deposition in the low 50s, low to mid-20s and mid-20s percentage range, respectively. Especially in etch and deposition, recent selection decisions by our customers have demonstrated good progress towards our goals. In clean, we are now more focused on n+2 transitions with the planned release of our next-generation product next year. In the remaining 30% of n+1 decisions available and n+2 decisions that will be made in earnest throughout next year, we are targeting additional growth hinged around key technology inflections consistent with achieving our 3- to 5-year targeted market share growth objectives of plus 3 to 5 percentage points for etch, 5 to 10 percentage points for clean and 4 to 8 percentage points for deposition. In etch, as market share leader for the conductor segments and particularly for critical patterning processes, Lam stands to benefit from the growing number of patterning steps in next-generation foundry and memory devices. In single wafer clean, we have a position of leadership in back-end-of-line applications and are now focused on delivering enhanced productivity and differentiated process results to better compete for a broader range of applications, including the front-end cleans. Since the last earnings call, our design solutions have generated considerable interest from leading device manufacturers, with the promise for delivering a flexible architecture and on-wafer performance capabilities aligned with their roadmaps. Decisions for clean process steps often come later in the selection process as we talked about previously, and we believe there are still opportunities to capture new applications at the n+1 technology node. In deposition, Lam holds the leading position for critical back-end-of-line interconnect processes, including diffusion barriers, ultra-low-k film gearing and copper fill. The complexity, in some cases the number of interconnect layers, is increasing, which requires more deposition systems to process the same number of wafers. As a newly formed combined company, we, Lam and Novellus, are more competitive together, and the timing of our transaction fits well with industry transitions. Illustratively, the 3D NAND devices roadmap represents a significant growth opportunity for both etch and depositions, sometimes together through adjacent process leverage. Repeatable and precise process capabilities are critical to 3D NAND device performance, our ability to etch high aspect ratio structures and deposit atomically smooth film stacks both uniformly and repeatedly position us well for opportunities in dielectric etch and PECVD both. Overall, we're delighted by the interaction and excitement shared among our combined engineering and field organizations, more by the results already evidenced to identify and develop strategies for accelerating Lam's growth. Worthy of note, we are already targeting incremental revenues in calendar '13 in excess of our stated synergy cost reductions. We continue to execute plans to combine our key business processes, management systems and infrastructure. The magnitude of these projects is sizable and includes, among other areas, the integration of both company's IT infrastructures, HR platforms and the integration of our ERP systems and processes. We completed a thorough assessment of these projects early on and have developed plans which address potential risks. I'm pleased with the pace and progress of our efforts so far, which we are largely targeting to complete by mid-2013, when a number of step function cost synergy actions become possible. Since closing the Novellus transaction in early June, we've made significant progress against our cost synergy targets. We are rationalizing our infrastructure at the executive level and across our regional management team and support organizations. We are nearly compete -- complete in our efforts to consolidate field offices and spare depots worldwide. We have started to eliminate duplicate corporate fees relating to outside services, and we have successfully concluded initial cost savings negotiations with a number of key material suppliers. To date, these actions have resulted in annualized cost savings of approximately $25 million. Our strategy is to run the company at a lower combined cost than the 2 stand-alone companies would have done separately, remaining committed to long-term growth and, where necessary, sustained long-term investments. Subsequent to our prior earnings call, we had received a number of questions about CY '13 operating expenses. We plan to deliver a comprehensive commentary on this at our November Analyst Meeting, but for now want to convey some important headlines. The combined company cost baseline for the first half of calendar 2012 was approximately $295 million per quarter in OpEx. Stand-alone Lam had essentially capped operating expenses for calendar '12 to $200 million per quarter, and Novellus, $100 million. The combined company targeted cost synergies are $100 million annualized effective at the end of calendar '13. Approximately 50% of those synergies are planned in operating expenses. The new quarterly operating expense cap is less than $305 million per quarter for the combined company in calendar '13, which would apply to the $1 billion quarterly revenue level. This amount would include our anticipated cost synergies. It would fund above market growth and would accommodate ramping 450-millimeter equipment development. Initial modeling should assume a graduated progression through the year. Customary quarterly guidance will, of course, be available ongoing. We anticipate, at least currently, only modest OpEx expansion in subsequent years. This balanced approach to managing costs while continuing to invest for future growth is crucial in today's rapidly changing business climates and with high risk and reward industry inflections. We feel that this target optimizes short- and long-term performance appropriately. Turning now to our outlook for the December 2012 quarter, our non-GAAP guidance is as follows: Shipments of $800 million, plus or minus $30 million; revenues of $850 million, plus or minus $30 million; gross margin at 44%, plus or minus 1%; operating profit at 11%, plus or minus 1.5%; and earnings per share of $0.45, plus or minus $0.07 based on a share count of approximately 172 million shares. And with that, Ernie and I will be happy to take your questions.

[Operator Instructions] Our first question comes from the line of Patrick Ho with the Stifel, Nicolaus. Patrick J. Ho - Stifel, Nicolaus & Co., Inc., Research Division: First, in terms of the outlook for the December quarter, can you just give a little color, in terms of percentages, what you're expecting in terms of the different customer groups?

Yes, for sure. As you might expect, relative to the shipments breakdown, there's a fair amount of intensity around foundry and kind of logic/other. So I would assume that about a 50% profile for foundries, 25% level for logic/other and about 25% with memory. And within that, memory sets a pretty heavy weighting towards DRAM upgrades is the profile that we see right now. Patrick J. Ho - Stifel, Nicolaus & Co., Inc., Research Division: Okay, great. The foundries moved to the -- both, I guess the 28- and the 20-nanometer nodes. You're seeing the increased talk about going to FinFET technology for the foundries. Can you give a little bit of color of where Lam is well-positioned in terms of that process technology or where you think you can gain share?

Yes. I guess, the first, I'll break that into kind of 2 components, 28 to 20. As I think we've articulated previously, that's a pretty expensive transition for the foundries. I think the estimates are between 1.3x and 1.4x in terms of cost. And our best estimates in terms of impacts on the etch fab [ph] are between a 25% and a 35% increase in spending. Naturally, the types of challenges around metal hard mask transitions in that node are definitely appealing to the natural strength of the company in terms of conductor etch. And certainly, that's a very positive headline for the company. As you move towards the FinFET transition, obviously, there's some announcements around adoption of the 16-nanometer or sometimes referred to 18-nanometer node. And again, the types of profile challenges that exist with the vertical structure kind of naturally exploits the strengths of the company conductor etch in the logic space, but also demonstrated performance in high-aspect ratio features in some of the memory applications as well. So we're feeling pretty good about the impact on our SAM in those transitions, which we think is kind of pretty close to the average WFE transitions for the industry, for the 28 to 20, maybe a little less in the case of etch, as I just spoke to. And the FinFET transition for us is probably a 15% increase in SAM. So good position and good impact in terms of SAM opportunity.

Our next question comes from the line of Vishal Shah with Deutsche Bank.

Martin, I just wanted to follow-up on your comment on the 2013 CapEx outlook. Can you talk about your assumptions for both memory and logic foundry spending and that $30 billion number? And then I will follow-up.

Yes. I guess first thing to say is we assume both memory and logic and foundry spend money next year. As I think we've articulated probably in equal measure through the year, the breakdown that we're assuming for 2012, WFE is about $8 billion of memory, $5 billion of microprocessor, $12 billion of foundry and $3 billion logic/other. The principal change we believe that exists in terms of segments if you buy into the $30 billion that we've just described, is that memory is a little stronger. It goes from the $8 billion level to the $10 billion level. That's the placeholder we're assuming. And there are a couple of reasons for that. One of them is that for the assumptions we're making on DRAM bit growth, and I think they're consistent with consensus, is slightly stronger bit growth next year at the high 30s. And I'm seeing some memory companies perhaps even speak to a little bit better in some instances. So we're assuming high 30s on the back of a 30% 2012 year. And despite the current environments that we're in with NAND, when we look at the construct of today's installed base and think through the 50% bit growth assumption for 2013, we would estimate that there will be about a 150,000 wafer starts of capacity addition next year and about 450,000 wafer starts of capacity conversion. And that is a slightly richer ratio biased to the additions that existed in calendar '12. Calendar '12 was the most efficient year in NAND history. That's certainly in the last 3 years relative to the mix of addition and conversion. And the 150,000 wafer start number that I just said is our assumption for next year NAND additions. It's still meaningfully less than the additions that were made in 2011 and 2010.

That's very helpful. And just one other follow-up. If let's say, industry WFE is up 10% next year, in that scenario, how do you think about Lam's revenue growth given the rising etch intensity at 20-nanometer?

Yes. I mean, I think I've framed the first part of the answer to the question for you in terms of the SAM expansion. So you're going to have to kind of build the construct from that because we're not about to kind of launch into giving guidance for the calendar year revenues. But I think the message in terms of SAM expansion through the 28 and the 20 transition for foundry and the opportunity through the memory roadmap is positive. And the n+1 market share numbers that I presented to you in my prepared comments were all positive to the current baseline, so you know what we're targeting to achieve. And certainly, in the second half of next year, n+1 will begin to feature in the spending of WFE.

Our next question comes from the line of Chris Blansett with JPMorgan. Christopher Blansett - JP Morgan Chase & Co, Research Division: A couple of quick ones. Martin, you seem to be more positive on near-term etch and deposition share gain than you do on clean. What types of customers or what types of applications do you think you'll gain share at in those products over the next year or 2?

Well, the good news is I seem to be more positive because I am a little bit more positive, at least in the near term. As I've described before, I feel really good about the competitive strength of the company in clean in the back-end-of-line. But until we have our next-generation product, we're not going to have the competitive differentiation that I think will deliver the traction in market share gain in clean. And that's something that will be with us targeted in the middle of next year. For etch, well-positioned in terms of 28 to 20 transitions. And as I said, 70% of the decisions, the n+1 decisions for critical applications, we believe, are pretty complete at this point in time. So we have reasonably good data on the etch position in foundry. And the patterning layers and hard mask steps in that transition are very beneficial to us given the natural strength of the company in conductor etch deposition. As I mentioned, it's really all about back-end-of-line process, diffusion barrier, copper fill, very strong market share positions for the company as a baseline. I mean, the DCV [ph] position of the company is about a 70% market share, so we're coming from a position of strength. And that's obviously a very important foundation. Christopher Blansett - JP Morgan Chase & Co, Research Division: And then last question was tied to residual Novellus revenue prior to the acquisition. I want to know how that plays into your guidance for the fourth quarter if it does play at all into that?

Yes, Chris, this is Ernie. As we mentioned, by the December quarter, there is virtually no quarterly impact resulting from that change in rev-rec policy because there'll be as much flowing into the revenue stream in the December quarter as there will be deferred. So that impact is largely behind us.

Our next question comes from the line of Stephen Chin with UBS.

Martin, just a question as you plan ahead into early 2013. How are you initially planning to see WFE spend trending in the March quarter? It sounds like NAND is a first half opportunity, in the full year it should be strong. Are you planning for the December quarter to likely be this near-term trough quarter?

So I'll give you kind of couple of data points. Obviously, to really put a stake in the ground on the proportion of WFE first half and second half next year is pretty difficult. But we would probably assume a slightly weaker first half than second half. And whether it's 14 or 16 or something slightly different, that's the baseline I would assume for now. In terms of answering the question specifically about where kind of a NAND investment kind of plays out, today, our customers have done an incredible job basically locking down qualified capacity. And so the amount of capacity that's unqualified exiting this year, I think it's at an all-time low. So if there's any demand, it's going to, by its very nature, require capacity addition. So I think the switch is going to be really immediate when demand exists. And when you look at some of the signals of demand, I mean, we're not lost without demand signals. So although we don't disclose bookings, perhaps to be a little helpful, if -- when we look at our forecast, our bookings forecast in December is slightly greater than our shippings guidance -- shipments guidance today. So the trajectory that appears to exist would tend to support March being better in terms of business activities than December. But as we all learned in this quarter, things can change very quickly.

Our next question comes from the line of Satya Kumar with Credit Suisse. Satya Kumar - Crédit Suisse AG, Research Division: Yes, a couple of questions. Just a clarification on this December bookings comment. Is December bookings up from September directionally?

Yes, I'm not going to comment because -- I had a feeling if I said something about bookings, someone's going to ask a follow-up. So I'm going to tell you exactly what I told you. And we've got some bookings-to-shipment ratio that's positive in our outlook in December, but that's the color you get on bookings unfortunately. Satya Kumar - Crédit Suisse AG, Research Division: Okay, I just thought I would try. Another -- a question on your logic shipments. Embedded guidance that you're talking about for logic shipments in your December, if I just do the math, logic is up almost 2x. Your September numbers, I don't if I got that right or not, that seems to be in contrast with what we're hearing as weakness in the PC space and generally investor concerns of equipment reuse. Is this share gain on your part or less than a perfect understanding, by perhaps investors, on what the equipment needs are for the logic segment longer term?

Yes, I mean, at some level, it may be a -- it's certainly assumed that it's in the category of smaller numbers on an absolute basis. So the percentages tend to kind of swing around a little bit. I don't think we've got a fundamental market share message we're communicating to you. I think it's a fairly distributed set of people spending money. You're absolutely right to make the point around reuse, at least in the microprocessor space, as being relevant to the purchasing activities today and maybe even in the December quarter as well. So we don't really have a kind of fundamental share gain message to be communicating. But the math that you did is certainly in line with the assumptions that we're making today.

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

Mark, on the opportunity that you articulated about improvement in NAND CapEx for next year, what's the risk that the first thing that happens is that we move some of the excess DRAM capacity over to NAND before we see new NAND capacity expansion? Recognizing that I think your logic on NAND price is improving on the heels of a lower CapEx this year makes perfect sense, but might we have to eat through some excess DRAM capacity that gets transitioned over before we see new NAND orders?

Yes, I mean, I think if you kind of back up 3 to 4 years, that was a much more kind of prevalent reality than it's been for some time. In the last 12 months, we've seen a couple of customers talk about it and/or do it. And in fact, the transitions have actually gone the other direction to the one that you're hypothesizing, at least in Korea, or I think that's the reality. It's not an easy thing to do. It can be a little costly and complicated and challenging. So we're not actually sitting here today with a very misbalanced supply and demand in DRAM. I mean clearly, the guys that sell DRAM devices can make the choices that you're describing and potentially the economics of DRAM, unless you're in the mobile DRAM space, are going to kind of motivate people to at least consider it, but it's not easy. So we're not seeing a big supply demand imbalance despite the fact that clearly the economics of DRAM, unless you're at 3X mode or headed to 2X, are very difficult. But it is at least a theoretical consideration.

Okay. And I guess if I could just use this as my follow-up. So you're articulating a scenario of flat-top wafer fab equipment spending next year. Is it fair to say that if the biggest logic customer and/or the biggest memory/foundry customer is going to be down 30% that, that's not baked into that $30 billion wafer fab equipment forecast? In other words, in that $30 billion, you're not assuming any of your large customers decline significantly, is that right? So if that were to happen, that's incremental downside to your forecast?

Yes. I mean, it's pretty hard to argue a global WFE that is disconnected from the realities of spending by the top 3 guys. They represent 60% to 70% of spending these days. So you're generally kind of right, I would say. Again, there's not a huge amount of science at this point to -- at least too precision here, and I do think we're biased to see a little bit more upside. The only thing that we've kind of assumed in terms of growth year-over-year from a segment point of view is in the NAND space for the reasons I've described. At least our assumption on the PC, which is still pretty impactful in terms of spending, we're only assuming 3% PC unit growth, and the range is like wild. I mean, I've seen minus 1% or minus 2% percentage point estimates all the way up to 11%. So we're in a fairly conservative range at the 3%. So I don't know if there's upside there or not. Well, time will tell. And there's the potential for some upside, I think, if the 3D NAND transition and/or the 20-nanometer foundry roadmaps kind of pop a little bit faster than people are expecting. But at least for us, in the $30 billion, we're pretty conservative on all of those.

Our next question comes from the line of Edwin Mok with Needham & Company. Edwin Mok - Needham & Company, LLC, Research Division: I was wondering on the gross margin improvement this past quarter, how much of that was contributing from this broadening on customer? And as we move to the new year, as some of these bigger customers come back and spend, are you worried that those bigger contract with those big customer might impact gross margin?

Well, the customer concentration kind of reality, I think, for every equipment company is the more the big guys spend, the tougher it is to make money. So certainly, the evidence of the last 12 months has shown that to us. And frankly speaking, despite our best efforts, it's probably a legitimate part of our future. Actually, in the September quarter, the mix that's in there relative to the top 3 is actually pretty reasonable. It's not a kind of wildly concentrated profile and it's not wildly distributed profile. The top 3 customers represent a fairly reasonable assumption for their share on an ongoing basis. So it's not a bad stake in the ground. Edwin Mok - Needham & Company, LLC, Research Division: I see. Okay, great, that was helpful. And then I guess going back to NAND, you mentioned that you guys are well-positioned for 3D NAND transition which could potentially start some part in the back half. Do you see a scenario where some of these NAND customers want to delay the spending because they know that they're going to transition to the 3D NAND in back half of next year or maybe 2014, and therefore result in lower levels of NAND spending in the coming year?

Yes. I mean, we're -- as I said in my prepared comments, we're assuming that in the $30 billion, it's all about planar NAND. So we got like next to nothing assumed for 3D. Relative to the timing of the transition, it's not an easy technical transition. There's still a decent set of unanswered questions on device flow and structure. And evidence of that is we are continuing to learn about our SAM opportunity. So you will remember in kind of prior quarters, I have said that if you look at kind of etch as a -- etch and deposition as kind of baselines for our company today, and you start to model what happens to our SAM or what happens to the cost of 10,000 wafer starts in a planar to 3D transition in the FinFET space and the NAND space, I had said I believe that our SAM goes up by about 15 percentage points by virtue of process time and process flow. And in the last kind of 6 months, the customers made some pretty significant choices around NAND process flow. And I would say that, that 15% that we have stated previously is probably underestimating the opportunity in etch and clean, both by up to a 2 to 3x factor. So the 15-percentage point number that we have said previously has the potential to be 2 or 3x greater than, that in this transition. So it's a very big deal. And that's a big deal in the context of overall WFE because unlike in the logic transition, the requirements for lithography actually get eased in that transition to a 3D structure memory. So not only is it a big deal in terms of the amount of money that gets spent in our segments in a 3D NAND environment, the proportion of WFE that gets assigned to the segments of our business is also increasing.

Our next question comes the line of C.J. Muse with Barclays Capital.

Question for you on gross margins. In terms of the 44% for Q4 guide, how should we think about the trajectory thereafter? And I guess what I'm getting at is, is there a benefit there in terms of deferred revenue contribution? How should we think about the synergies rolling in? And I guess, given your construct of kind of second half more heavily weighted than the first half, we'd kind of love to get a feel for how you see that trajectory through the calendar year.

C.J., this is Ernie. So at sort of even revenue volumes, I think you would see continued incremental progress on the gross margin line as the result of the synergies associated with that. Obviously, you're going to have an impact from customer and product mix as well. So I wouldn't say you would see any significant upward trajectory, but certainly, the opportunity exists for there to be some incremental improvement.

Okay, that's helpful. And then I guess as my follow-up, and forgive me if you've discussed this, but curious what you could share in terms of share repurchase with all the excess cash and how you plan on, I guess, using that going forward.

So certainly, we're going to be executing on the balance of the $1.6 billion share repurchase. As we talked about in our prepared comments, the vast majority of that will be complete in this calendar year. Beyond that time, we're engaged in discussions with the board about an appropriate way to ensure an ongoing return of cash we generate to our shareholders, and we'll be providing further updates to that as it becomes more clear to us.

Our next question comes the line of Jagadish Iyer with Piper Jaffray.

Two questions, Ernie. First -- on the first one, you guys have talked about 3D NAND. I just was wondering and you also talked about the build [ph] in n+1. So I was wondering how well are you guys positioned with all the 4 major NAND producers in terms of 3D NAND? And then I have a follow-up.

I'm not sure I got all of the question, but I -- certainly, in terms of the n to n+1 transition, for us in NAND, certainly, that is solid and there are penetrations assumed in the market share objectives that I've previously described in terms of kind of planar device. From a positioning point of view in the 3D structure, I think across the full portfolio of products, we've got the right engagements at the right time with the right people. As I mentioned, there's a lot of decisions still to be made, a lot of uncertainty, not just in timing, but in terms of the ultimate construct of the flow. And so it's a big part of the investment profile in the company, it's a big part of collaborating and engaging with R&D at the customers, and we're at every significant location we need to be.

Just on the -- as a follow-up on the foundry space, people are talking about this ramp in 20-nanometer in the second half. I just wondering how much amount of reuses there on the 20-nanometer are potentially do you think you could benefit from double patterning in that?

Well, the answer, I guess, to the second part of your question, there is really the 28 to 20 numbers that I already kind of disclosed to you. I said I thought 25% to 35% was the SAM expansion potential for etch, deposition a little less than that and clean is kind of to be determined. It's all about a yield play. So we really don't know how to answer that question until we get there. And there is a reasonable potential for -- in the range of 25,000 to 30,000 wafer starts of 20-nanometer capacity to be added next year. That's our assumption right now. And so it's predominantly a 28-nanometer year still. I mean, we exit this year at maybe at 240,000 wafer starts, and we believe that the 28-nanometer technology nodes will exceed 300,000 wafer starts ultimately and maybe get to as high as 340,000 or 350,000. We'll see. So it's really a 28 year, with initial 20. Answering the reuse question is pretty difficult at this point. I would tend to think reuse potential is very low. It's not an established model in the foundry, certainly in the 40 to 32 and 28 transitions. There was a little bit more than historically. But it's a tough place to do reuse on any significant scale.

Our next question comes the line of Terence Whalen with Citi.

This question relates to the foundry business. If we look ahead into 2013, and if we were to see a significant shift regionally toward a strengthening Taiwanese foundry away from a weakening Korean foundry, how would that affect some of your near-term market share goals in etch and deposition and clean?

Yes, I don't think it affects the near-term goals in the slightest, personally. But ultimately that's a by-product of decisions that customers make, not us. I do think if you kind of peel the onion of market share for Lam Research in etch, using etch as a proxy, we're reasonably agnostic to the choices and the transitions geographically across the base of foundries. And the obvious exception to that obviously is a company like Intel to the extent that it is -- have plans to or chooses to pursue business in a foundry model, which gets talked about in the press all time, and I have no idea if that ever happens. But that's the only one exception. Everything else, we're reasonably agnostic to, by virtue of where we have applications penetrated and the differences in the process flows between the foundries.

Terrific. And then the follow-up question I have is regarding the change in OpEx from roughly $284 million to approach maybe $305 million by year end '13. Just to understand, is that a fairly linear ramp? And also how much of that would be in cash OpEx versus stock compensation?

Terence, it's probably reasonable at this point, and I would stress that we haven't yet completed our 2013 annual operating plan where we would have that level of visibility. But it is reasonable to think about that on a linear basis. And it's reasonable to think about at least half of that delta being in noncash-related expenses, be it stock compensation or depreciation.

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

Martin, I had a question in terms of the front-end-of-line clean. As you guys try to get into the FEOL clean and your competition goes from front-end-of-line to back-end-of-line clean, are you guys using the same products? Are you guys developing new products to attack this market?

We're planning to do, we're -- as maybe the framework for answering the question, because the products that we're targeting to release is a calendar '13 product release for us, is building upon the established strengths of the company, which is to have a common architecture that we use to penetrate in many applications and then to have kind of like an options flexibility that allows customization to specific back-end applications and front-end. It is a new product, yes. It is much more productive than the platform that we're able to compete with today, and so it fundamentally changes the competitive position of the company on productivity. And we are working really hard to establish technical differentiation in the area of specific challenges that are relevant to getting traction. There's a lot of work ahead of us, but certainly the plans of the company are defined that way.

Got it. That's helpful. And just a quick follow-up, you guys talked quite a lot about 3D memory, and it seems interesting that it's finally coming to fruition at some point in the near future. What specific product lines of yours would be best impacted or best positioned when 3D NAND or even 3D DRAM happens?

It would appear like deposition, one, etch, two, and clean, three, in terms of the change from a percentage point of view in each of those segments. Obviously, the etch segment is the biggest segment. So the absolute dollars of change in etch may be quite similar to the absolute dollars available in terms of change in dep. But on a percentage basis, if you look to percentage dep in our planar structure to a 3D structure is the biggest proportional increase. And I've just said clean on the lower side, and the reason I said that is because until the customer attempts to yield, we're not really going to have the answer to where the value on yield-enhancing solutions is from a process flow.

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

Martin, going back to your NAND commentary, I'm just curious what gives you the confidence that we're going to see a kind of cycle in NAND spending that has happened many times in the past? The prices go up and your NAND customers are still [ph] spending. And I'm looking to next year, more and more of your NAND manufacturers are reallocating capital to M&A, and they seem to also be changing their strategy, different focus. So again, what gives you confidence that this time is going to be similar to what has happened in the past? And why wouldn't it be more different in the sense that more dollars of CapEx reallocated to M&A?

It's a tough question. I would say, first of all, I don't think anything happens kind of anytime soon. I think consistent with comments from other equipment companies, we're certainly in a lull in December and probably in a lull in March as well, and we'll see when we get to the June quarter. But relative to confidence, I don't think we have an extravagant characterization of bit growth in calendar '13. We're assuming 50%. 50% is the lowest percentage of bit growth in any of the last 4 years. That's kind of statement #1. Statement #2, at least to our analytics, there's only 20,000 wafer starts globally of unqualified, unused capacity. Unbelievably low. And so any demand that comes in is going to be a catalyst for spending. Now the spending can come in many forms, to your point. But at least to date, we have not learned anything from customers that would cause us to model the outlook differently. Obviously, if we do and when we do, we'll share that perspective with you. And at the end of the day, the NAND investment, and like any segment, frankly, is all about demand and supply, but it's also about the availability to upgrade the existing installed base. And in calendar 2012, we probably have seen 70,000 wafer starts of new capacity added by the end of this year in support of a 65% bit growth. And we're assuming about 150,000 wafer starts of addition next year, with about 100,000 wafer starts of less conversion in '13 than occurred in '12. So that mix has the potential to change the answer to the question. We've done our best to tell you what we think the outlook is, and our best estimate is 150,000 wafer starts of addition next year and 450,000 wafer starts of conversion. And the additions will cost about $280 million, $285 million, $290 million per 10,000 wafer starts and the conversions we think cost about $50 million to $60 million per 10,000 wafer starts next year.

Got it. And then one quick follow-up in terms of reporting. What would it take for you to provide some segment reporting, either revenue or revenue and profit margin profile?

Mehdi, providing segment data would be particularly challenging given that so much of that is available to our competitors, and there's a general trend in the industry that, that information is not provided, and we certainly don't want to be ground breakers there. So I don't think you should anticipate a change anytime soon.

I'd now like to turn the call back to management for closing remarks.

Thank you, Britney. 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 again, we appreciate your interest in Lam Research.

Thank you. Ladies and gentlemen, that does conclude our conference for today. We thank you for your participation. You may now disconnect.