Okay. While everybody is finding their seats, good morning and welcome to the 2015 Lam Research Investor and Analyst event. We have until 11 am this morning and we ask for presentations and Q and A and following that we'll have a coffee hour right next door. We'd like to thank everybody for joining us here in San Francisco and via webcast. I'll just cover a couple of items quickly before introducing Martin.
So as you can see in the slide accompanying my statements, statements made today that are not historical in fact are forward looking and are covered by the provisions of the Safe Harbor Act. Forward looking statements are based on current expectations and are subject to risks and uncertainties, which are covered in more detail in our SEC reports. Please be aware that actual results may differ from these forward looking statements and the company undertakes no obligation to update them. Great. So today we have 3 presentations for you.
Martin Anstas, our President and Chief Executive Officer will discuss our vision and the drivers of company and industry performance. Doctor. Richard Gaccio, EVP of Products will discuss our technology leadership with a focus on our strategy in the inflection. And finally, we'll have Doug Boettinger, EVP and CFO and Doug will discuss the drivers of our financial performance and provide you with an update on our financial model. Also with us today is Tim Archer, EVP and COO and he will be joining us for the Q and A session.
With that, I'll just remind you of 2 things before we get started. 1, please hold your questions until the end of the presentation and please place your phones on silent. And with that, I'd like to welcome Mark Nantzis, our President and CEO to the stage.
Good morning.
So it's always good when other people show up
to an event like this. So we're delighted to have you join us today. And I echo Audrey's comments. We really appreciate the opportunity to tell the story of the company and we really appreciate your interest in our company. So I was actually thinking this morning how many times have I done this.
I think this may be my kind of decade celebration. And in that context, one of the things that occurred to me when I woke this morning and started reading what some of you wrote yesterday is, I think this is the first time ever for me in 10 years where the comments from the competitor supported the investment thesis in our company so well. So for those of you that wrote, it's all about extra debt. Thank you. So joking aside, we're really proud of where we are as a company and delighted by the performance in the last couple of years.
And I would say probably more important in your context, we're excited about the future opportunity for the company. So we had a vision to do something pretty special to grow our company to outperform relative to the industry. And the evidence I think of the last several years is we've done that and we believe we can continue to do that and that will feature a lot in my statements prospectively. It will feature in Rick's session around the fundamental enablers of product and technology differentiation. And then Doug will provide some updated analytics for your digestion.
So I get asked all the time what I'm most proud of. And I would say, I am clearly very proud of the performance of the company. We strung together a couple of years' worth of growth in our company, which has outpaced the industry and I feel really proud of that. I feel very proud of the discipline and the creativity and the innovation that exists in our company. I feel really proud in the building of customer trust through a very significant period of inflections in the industry.
So at Lime Research, we have 5 very simple objectives. The first is to be number 1 in customer trust and every other objective we have like growth or profitability comes a lot more easily with customer trust. And the customer trust headline for our company in the last 3 years is a very positive one. And last but not least, I feel proud about the culture and the values and the authenticity of our company. So it's actually pretty easy to grow a company.
It's hard to grow a company profitably. And it's really a lot harder to grow a company profitably and sustainably. And in our humble opinion, as a leadership team in Lam Research, one of the most fundamental enablers for sustainable profitable growth is a set of values and a culture and authenticity to put the interests of the customer right at the front of every single decision we make in our company. And before I get to my slide, I'd like to do one last thing, which is to say a huge thank you to the leadership team that's assembled here today and the 7,000 employees in Lam Research that made story that we're about to tell possible. Without their belief and trust in our vision and our strategies, without their passion, their conviction and their ownership executing every single hour of every single day of every single week, month, quarter, this would not be possible.
It would not have been possible and it will not be possible, which is why our focus on differentiated values and culture execution is a very prevalent theme in today's presentation. So I'd like to begin today by saying thank you to the employees of the company. So calendar 'fourteen was a pretty extraordinary year. I have one slide on 'fourteen. So it was our 2nd year of essentially 20% growth year over year.
We added almost $1,000,000,000 to the revenues of the company. And to my earlier point, growth is maybe easy ish, take a lot of hard work, but profitable growth is what really counts if you're focused on increasing the value of your company. And we were able in calendar 14 to grow our revenues twice as fast as the industry and grow our profits, operating income twice as fast as our revenues. TSR as a company established an industry benchmark. There is no company in the top five that compares with the TSR performance of Lam Research in 2014.
That's a statement on the cumulative annual growth rates of revenues and earnings per share. It's a statement on the capital return policy and execution of the company and in calendar 14, the establishment of our first dividend. So the financial performance of the company in 14 was very strong. Execution excellence, we've talked a number of times about new product release activities. We had 3 press releases earlier this week that was another wave of that.
And Rick will talk some more in his presentation about how those platforms of growth further differentiate and strengthen the growth in the company. But we outputted probably the highest number of new products in the history of the combined companies in calendar 14. That's the platform for growth. Quality expectations of customers go up consistently and significantly, process windows narrow, predictability of results will also feature significantly in Rick's presentation today. And on time delivery is one of the most fundamental commitments we have to make to customers.
And in a period where we grew as fast as we did, our on time delivery to commitments was in the high 90s percent level around 97% in calendar 2014. So execution performance was really good and the market momentum was pretty good as well. We gained 1 to 2 percentage points of share in calendar 14 at a very fundamental level. So some of you have a lot of history with our company. So if you have enough history with our company, you'll remember Lam Research, single product company Etch.
Actually, if you have a lot of history, you'll remember something before that. But let's just focus on single product etch, which was the company 7 to 8 years ago. We would stand in front of this audience and say we actively compete for 14% or so of wafer fabrication equipment spending. In the 1st wave of an adjacent market growth strategy, we added clean and we said we now compete for 19% of wafer fabrication equipment spending. A couple of years ago, we added deposition to the portfolio of the company and at the day of the deal, we said we now compete for 25% of wafer fabrication equipment spending.
As a byproduct of process flow decisions in our customers, device architecture decisions of our customers, multi patterning 3 d device architecture are good examples. And our new product release activities in calendar 'fourteen, we were competing for about 27.5% of wafer fabrication equipment spending. Our sand box is getting bigger and our focus is on continuing to execute that strategy. Our market share of that at the 42% level and at the inflections, which is an important message today, it was an important message a year ago, we believe we have more than 50% market share in deposition, etch and clean combined. So excellent progress in calendar 'fourteen against the 'thirteen, 'seventeen objectives that we communicated to you a year ago.
Calendar 'fifteen, I think the trends and the message continues. So we're really pleased with the start to the year. We're executing on or ahead of our model. And in fact, one of the more prevalent parts of this conversation in the last couple of years has been, you're growing faster than you told us. What does that mean for the model?
And Doug's going to answer that a little later today. But we recognize in large part because of new product introduction, in large part because of the speed of multi patterning implementations and the bias of multiple patterning to spacer based solutions, the growth of the company has executed faster and the profitability of the company has executed to model as well. We have continued the trend of 2 times growth relative to the industry And our assumption still for calendar 15 is a $34,000,000,000 plus or minus 2 wafer fabrication equipment year. And we increased our dividends and have continued to execute the buyback. So financial outperformance is still an important theme of the company in the first half of calendar 'fifteen.
We have our earnings call in a couple of weeks' time. I did want to give you a perspective on the trajectory of the company in the second half of the year. So I'm going to make 2 statements. The first statement is, we expect to give guidance for shipments in September at or around the $1,600,000,000 level. The second statement I'm going to make relates to the revenues of the company first half and second half, which we expect to be reasonably balanced.
That was more distressing for me than you, I can tell you. And so does everybody hear that last statement? Revenues, we expect first half revenues and second half revenues to be reasonably balanced, Okay. We're a long way from December shipments, quite what they'll be. We don't know today whether they're the same as where we were in the March quarter or slightly less, we don't know yet.
A lot can change between now and then. But revenues for our second half balance, September shipments guide, we think around the 1 point $6,000,000,000 level. Execution, operational execution, financial execution continues to be strong in the company, record breaking factory output, on time delivery going up 99% level, on time delivery performance for the company in the first half of this year. The customer trust metrics of the company are net positive and we're increasing our commitments performance to the customer. Two illustrations, cycle time.
Cycle time reduction in the company is an important commitment to customers and we're driving all of our cycle times and lead times for delivery to customers down year over year. And our speed to support their ramp is a big commitment. So 12 week or less lead times is the commitment to the company and we're executing better to that today than we were a year ago. And year over year, we improved our cycle time installing equipment by 5%, which is a big component of speed to yield and speed to ramp for the customer. Market momentum continues to be positive.
My instinct is our market share momentum in calendar 'fifteen over 'fourteen will be greater than our market share momentum 'fourteen over 'thirteen. We'll talk about that in some more detail in a moment, which isn't to say we win everything because we do not. We win some things, we lose some things, we defend some things, we don't, we penetrate, we don't. But if we win more than we don't, we win long term. And that's been the story of the company and it's the story for the company that we continue to be focused on multi patterning, 3 d device architecture, very important headlines relative to sustainable outperformance for the company.
So if I kind of move away from 'fourteen and 'fifteen and think more holistically, what's the perspective on the company for the industry? Well, the first one speaks to kind of customer, customer speaks to the enterprise, speaks to consumer demand for electronics products. And we would characterize that as actually pretty healthy. So when we look at our customers, we see a lot of discipline. We see a lot of supply and demand balance.
And when we look at demand for consumer electronics, demand for enterprise electronics, we see things that are actually pretty inspiring. We see things that are very relevant to us, measured by opportunities to create innovation and competitive differentiation for our company supporting the roadmaps of the customer. And the intelligent device, the IoT, the connectivity, the ubiquitous device agenda that you all study much more than I'm sure we do is a very important headline for the company. We're actually agnostic to the device momentum per se, but what this translates into we believe is a statement of 8% to 9% IC unit growth, which is essentially consistent with the IC unit growth curves of the last 20 years. And IC units and wafer starts per month is our business.
This is the business of our customers. And the headline here is our industry and our customers' industry are the fundamental foundation to enabling the socioeconomic vitality of the modern world. That's pretty inspiring to be part of. And certainly, I think it's fair to say that the leadership team in the company and the 7,000 folks in our company that show up to work really hard every day have a passion and have a commitment and a humility about our role supporting that agenda. That's the substance of all of the financial and operational performance that I've already talked about.
A little close to home, this is a perspective on the segment disclosure for our company. So, kind of left to right, we haven't kind of got the numbers of wafer stats from an installed base perspective on the slide, but NAND flash is in the 1,300,000 to 1,400,000 wafer stats range. DRAM is in the 1,100,000 wafer starts per month range. And foundry logic ranging from 1,000,000 to 1,300,000 wafer starts. There's a couple of really important messages here.
I think we're generally in the range of consensus when it comes to bit growth 35 to 40 for NAND, 20 to 30 with stronger bit growth in the mobile space. Our expectation is this year is the transition year for 3 d NAND. So we absolutely expect the majority of NAND wafer fabrication equipment investment to be 3 d device architecture related this year. And we would anticipate approximately 150,000 wafer starts of ships capacity in 3 d this calendar year. Now some reasonable part of that will not be qualified for use, but shipped capacity we think for 3 d NAND is likely to be in in the 150,000 wafer start range.
And if I were to answer the question what's changed
since the last earnings call, I'd say
the 3 d device environment today is a little stronger than it was 2 months ago. And that 150,000 wafer starts reference point is somewhat indicative. By the time we get to calendar 2018, we believe that about 2 thirds of the installed base will be 3 d device architecture based. And the journey between now and then will be concentrated to very efficient conversion and where relevant new capacity additions on the 3 d device architecture beginning at the 32 layer and trending all the way to 64 and beyond in this timeframe. The DRAM space, super efficient, very cost effective and very legitimate as a scaling for the customer.
So this is the year where the 20 nanometer conversion is dominating the DRAM space. We would expect that by the end of this year, there's probably 2 thirds of installed base that is yet to transition to the 20 nanometer technology node and the 20 nanometer technology node for the customer gives them cost benefits and performance benefits. So there's high motivation to go execute that roadmap. And the DRAM space, although arguably it's the least certain technology roadmap, has kind of 2 to 3 generations in most of our customers roadmap after the 20 nanometer technology node. Foundry logic, 2 important messages.
1, this is the year of FinFET transition everybody knows, but it's also the year somehow related to the intelligent device and IoT message on the prior slide, where we're seeing a new wave of investments in 28 nanometer and above in a distributed base of foundry customers around the world. And so we do expect growth not just in the latest generation installed base, but growth also in the 28 nanometer and above installed base over the next several years. Our expectation in the foundry space for capacity at the end of this year that is 20 nanometer, 16, 14 capable is in the range of 200,000 to 210,000 wafer starts, again, shipped capacity into the industry. This is the land of opportunity for our company. This is the assumption set that we use relative to modeling our company, understanding our SAM, our served available markets, and also ultimately modeling our financial statements.
So for the purposes of your own analytics, this is an important slide because it tells you the basis of our assumptions relative to technology node transitions. A large part of today's conversation is going to focus on the dominant technology inflections through the calendar 2018 timeframe, 3 d device architecture and logic and memory, multi patterning and logic and memory and advanced packaging, 3 or 4 technology inflections depending on how you count them. In Rick's presentation today, we are also going to introduce to you the next set of inflections that we think are relevant to the growth of the company. And in the last several years, we have begun the transition of significant fundamental research and significant CNF investments, concepts and feasibility investments in the next wave of inflections because to a very large extent, the inflections that are relevant to the industry through calendar 'eighteen have been decided or at least there's a level of selection bias or decision making that defines the market for the next several years. So there's obviously an assumption set here relative to EUV.
There's an assumption set relative to the multi patterning and next generation transistor architecture. Rick will share many more details a little bit later. So what is our, again, strategic, more holistic perspective on the semiconductor industry? It's clearly consolidated. And in the semiconductor industry, it is still continuing to consolidate.
And in the supply chain, there's some consolidation and there's some not consolidation. But the headline is the value proposition of scale in light of the challenges of Moore's Law is very relevant to the success of our company, which is why we made a commitment to grow our company. Scale is an enabler to the success and the sustainability of profitable growth long term. 70% of spending we believe is top 5, 55% we believe is top 3. So a very concentrated spending, which means the fundamental dynamic for an equipment company is different today than it was 5 years ago, certainly 10 years ago.
And it's fair to say we don't have the crystal ball, but it's also fair to say that at least the recent last 5 years and our best assessments of the next 5 years imply that the world of cyclicality is behind us or minimally significantly suppressed, which introduces a new challenge if you're running an equipment company, because now it's a challenge of variability. You have to be able to create a flexible business model to create variability in your company from one customer to another, one roadmap to another, one device architecture to another. It's a different operational challenge a big part of what Tim Archer focuses on, our Chief Operating Officer, is enabling that business model transition. There were things that were important to us and valuable 10 years ago that are not so valuable and not so important to us today. They've been replaced by a new paradigm.
Complexity of the industry is going up. Complexity challenges, not just in terms of technical roadmaps, but in terms of economic roadmaps and the ability to control process. That agenda, whether it's measured by capital intensity, the number of layers in a device or the amount of demand outside of the leading edge that's emerging with the intelligent device roadmap. It's an entirely different gig. What does that mean for us?
It means opportunity. It means it is really important to innovate not just around the product technology agenda that Rick will present, but to innovate in terms of the business model that Doug will speak to and perhaps Tim will speak to in Q and A a little later to innovate the business processes and the organization and the systems and the processes of the company to innovate the collaboration agenda of the company. And one of our most fundamental headlines is collaboration and the Lam ecosystem, the community of partners that we have, peer companies, customers, suppliers, academia, that community is a strategic imperative to sustain the long term growth of the company. So innovation is a very broad prerequisite for sustainable growth in the company. These are the technology inflections that we've been talking to you about for the last several years that we'll kind of bring to closure at some level here today.
And it is not happenstance that every single column on this slide says deposition and etch intensive, okay. That was the vision. That was the strategic rationale associated with making the investments in fundamental research and concepts and feasibility to deliver the set of new products that are defining world class performance today in multiple patterning, 3 d device architecture and atomic level processing regimes, okay. Another very important slide. So this is a build to what Doug will present.
So one message has kind of changed this year for us. We've added 18. So last year this slide said 17 and the 18 reference is a little higher than the 17 reference. So for those of you that thought our message was we grow really well to 'seventeen and then we stop growing, that's not the message, okay? So the message is we believe we have sustainable growth, this set of inflections beyond 'seventeen and as Rick will communicate a second wave of inflections as well.
But a very important message I already stated is our market is higher than our baseline. This slide says the proportion of wafer fabrication equipment spending that is inflection based is going up. 2,003, it was about 10%, today about 33%. And by the time we get to the calendar 2018 timeframe, 55%. That's one of the most fundamental reasons why we claim we're one of the best, if not the best positioned equipment company to outperform in the next several years.
It's one of the fundamental reasons why the new product focus of the company has enabled the performance on the right hand side of this slide, which is the proportion of WFE that we're competing for. So I already delivered the historic message, which is the 14%, 19%, 25% message at the time of the Lam and Novella Steel, 27.5% last year, 28.5% to 29% this year. By the time we get to the calendar 2018 timeframe, we believe that 33% of wafer fabrication equipment spending will be our SAM, that's sandbox statements that I made previously. Another new number for you. So this slide last year said 2,000,000,000 dollars It's a slightly modified slide to the extent that it now speaks to a calendar 'eighteen reference point, not just 'seventeen.
But one of the messages today is the growth of the company has occurred faster because the market expanded and our share actually performed a little better than we were anticipating. Our best estimates of SAM expansion in the 'thirteen to 'eighteen timeframe at the wafer fabrication equipment spending level that was resident in that the 2013 timeframe is in the $3,000,000,000 range. And if you have estimates for WFE higher, then obviously that's the reason why there's a plus here. The items highlighted in yellow in the boxes speak to segment specific SAM growth. And with the exception of advanced packaging, which is the smallest and last but fastest growing opportunity.
Every other number on this slide is bigger today than it was a year ago. And that in large part is a multi patterning message. And a big part of that message is a spacer based, not litho etch, litho etch based patterning scheme. So a very positive message in terms of SAM expansion compared to 1 year ago. How do we compete for that?
We are not the total solutions guy. We're the unit process guy. Our vision and all of the focus of Rick's organization and all of the focus of our field organization of the customer interface is to establish Lam Research as the go to company in deposition, etch and clean. Unit process excellence is the commitment we make to our customers. We make the commitment to be a technology leader in those segments of the industry and a productivity leader in those segments of the industry and to do it faster than anybody else does.
That's our commitment. That's what we strive to do every day. Our focus and the intensity of our business today on the installed base and productivity solutions is at an all time high and it will continue to be that way. And there are a couple of illustrations here of the substance of productivity offerings in the portfolio of the company. There are many more and the productivity solutions of the company are obviously embedded in the financial statements that Doug will talk a little bit later.
And the last message on this slide relates to collaboration. It is not possible to do this all on our own. And we have consciously invested in an ecosystem around the company, a community of partners, customer suppliers, academia peers that create substance around integrated process knowledge across the portfolio of the company. Collaboration agenda will continue to be more fundamental. Market share, again, a couple of things on this slide that are ratified compared to the statements of last year and a couple of things that are perhaps a little bit more positive.
So the one statement I'm going to make today, which ratifies the claims of prior year is we believe at the inflections, we have more than 50% market share against a baseline for all WFE of the 42% market share level calendar 14%. So there's a promising market share story for the company in addition to the SAM expansion story we said a little earlier. I already delivered the message that we believe shipments based market share calendar 'fifteen over 'fourteen will be stronger than 'fourteen over 'thirteen. So our momentum to these targets is very strong. Deposition 4 to 8, same set of numbers as 1 year ago.
Etch 4 to 8 is slightly higher than the numbers we had a year ago. We had a 3 to 5 reference a year ago. Today, we have 4 to 8. There are 2 parts to our etch business that you are all aware of. We have very strong position in conductor etch and we continue to retain that strength and build upon it.
And the more significant area of growth for us in etch is in the dielectric etch business. So that's a big focus for the company. The clean business, we're not going to talk so much about clean today because we've chosen with limited time to focus almost exclusively on the technology inflections, which to the points that you have made is dominated by a depth and etch story. But there are a couple of things to say about clean. So the first thing is that we have some significant questions to answer with customers in the second half of the year.
I don't think that comes as a surprise. So we've got some kind of big events coming up in the next couple of quarters. In light of that, we have chosen for accounting purposes to be conservative about our estimates on the balance sheet and we are planning to record an impairment of goodwill on our SEZ purchase in the June quarter of up to $80,000,000 that will be a non operating one time expense to the company. That's a kind of heads up for what we'll talk about in a little bit more detail. That's headline number 1.
Headline number 2, great momentum with the product in the marketplace. So those two messages maybe don't sound like they relate to each other. The impairment basis essentially conservatively locks the growth closer to the low end of this range, not the high. So that's kind of the rationale for the accounting charge that I just explained. But the new products and Rick may add some comments later is getting great productivity, reliability is a better economic solution for our customers and for our company.
We're pleased about that. And similar to the Edge and the Deposition business, we actually have our strongest market share growth year in clean than any year in the last 4. So nice momentum, nice ratification of our prior market share ambition. My last slide. So this is at some level a celebration of some accomplishments to date, which I'm sure you give us maybe no credit for at this point.
And perhaps the more important message from this point forward is a statement about future opportunity. We believe with passion that the opportunity for our company is extraordinary. We believe our destiny is in our control. We believe that the opportunity is there and it's all about execution and execution begins with building customer trust. It begins with the values and the culture and the authenticity of our company.
It begins with the technology leadership, which Rick will present to you in a few moments. It begins with a collaborative and ecosystem approach to our business. It's easy to grow. It's hard to grow profitably. And it's even harder to sustain that profit growth.
That's our focus. That's our passion. That's our commitment. And we work hard each and every day to make that vision a reality. With that, I will transition to Rick.
Thank you.
Good morning. I'm really excited today to talk about extending technology leadership. I want to emphasize the word technology in there. I feel like I've been unchained and I'm allowed to talk more about technology than I have in the past. So hold on to your seats and let's go for a ride.
One comment I want to make though is that we say extending technology leadership on the back of our business cards, we talk about innovative technology, trusted productivity and fast solutions. We believe actually trusted productivity is actually a technology problem.
Cost
is a technology problem. Cost is a big problem for the whole industry. So being innovative technically to solve cost problems is what we need to do to sustain our growth. We're focused on technology through the inflections. Oh, sorry, pointer.
We've talked about these inflections before. Martin just talked about some of them. I'm going to focus today really on 3 d NAND first and then multiple patterning. That's where most of our growth has come from. Most recently and over the next couple of years, it's going to be the biggest source of our profitable growth going forward.
The rationale for going to 3 d NAND from the customer's point of view is obviously more bits. You're growing the number of transistors vertically, not just horizontally. So you get the highest bit density per unit volume now instead of per unit area. And that effectively keeps us on Moore's Law in terms of lowering cost per bit year over year. Multiple patterning is all about getting higher density, taking the limitations of lithography and really overcoming them and extending the capability to finer and finer pitches and finer and finer dimensions, again, enabling Moore's Law to continue.
So I'm going to focus on those 2 in this talk. First of all, 3 d NAND. So this is a schematic illustration of a 3 d NAND cell memory array. And the important point to take away from this and I'm going to show you schematically how we built this is the critical dimensions of the devices which is shown up here schematically as a wraparound gated structure is all controlled by the dimensions of the films that are put down. So the thicknesses of these layers, how uniform they are, layer to layer across the wafer within the device.
It's controlled by the dimensions of the holes that you have to drill. The most important one is this memory hole and we just had a press release yesterday on our Flex G, the latest high aspect ratio dielectric etcher is used by everybody making 3 d NAND to drill this hole. We also have a very large share of these slits which are used to isolate these devices in the other dimension. So the dimensions of this transistor are controlled by the film properties and it's controlled by the etch properties. I'm also going to talk about the contact metal fill.
For most of the devices being made, these are alternating layers of oxide and nitride. And you the nitride is sacrificial. It gets removed and replaced with tungsten in an inside out fill process, which is enabled by atomic layer deposition. I should mention that the etch process is also an atomically controlled etch process. But that inside out fill is done with atomic layer precision and is also an application where everybody making 3 d NAND devices today uses our products.
So I'm going to talk a little bit more detail now about the so called stack deposition, which is the beginning of the formation of a 3 d NAND structure. And as you can see here on your left, the key challenges first and foremost are making sure that each layer, each pair of layers is exactly the same as the one above it. Otherwise as you go through this three-dimensional structure, all the transistors will be different. And that's obviously not a very good thing if you're trying to make a memory device. The read and write times to each device want to be consistent and you want a very narrow distribution across the entire cell.
If you have any defects that particles that may get embedded in the film, that's obviously a huge challenge to overcome because as you stack more layers on top, you can actually get distortion and the conformality of those layers as you build up more and more. And the roadmap is to go from 32 layers to more than 100 layers in the future. So these challenges don't get easier, they just get harder. These depositions take a long time. So the other big challenge is how to do these depositions with the highest productivity in the smallest area.
Customers can't afford to put so many more deposition and etch
pieces of
equipment in their fabs, they only have so much space. So footprint productivity is a very key metric. So we're winning in this market because we have the highest footprint productivity that's illustrated here by our Strata Q system where we have 12 wafer processing stations in the smallest footprint in the industry. You can see a picture here of an actual deposition of 64 pairs of oxide and nitride. We win because we have best in class uniformity within the wafer, wafer to wafer, lot to lot, chamber to chamber.
Low defectivity is making these products in time for the customer, but the stress in the film, the nitride stress, the oxide stress have to be carefully controlled. Later on when you come in and replace the nitride with tungsten, the stress of that tungsten has to be very precisely controlled. Otherwise the wafers, what we say, they potato chip and you can't handle them, you drop them off the robot and things like that. All those things have to be met simultaneously. We lead in each one of those areas and that's why this is a very important part of our growth story.
Turning now to the etch process. So I'm going to show you a little video for how the etch is supposed to proceed. And we start off with deposition of a hard mask, that patterning of the hard mask is done with a litho step not shown here. And then we proceed to transfer that pattern into the multilayer stack. And that etch that I've just shown becomes that is the memory hole etch that defines the dimensions of the channel of every transistor in the device, billions of transistors fabricated at the same time with our etch process.
That is
done in our Flex product. We just had a press release as I said earlier yesterday about this product. We're very excited about it. The challenges to be overcome here as illustrated again on the left are while the video showed perfectly straight etching, the reality is things charge up that mask charges up in a plasma. You get differential charging as you go down the mask and all the way down through those layers.
The amount of charge that builds up on the bottom of a 64 payer stack very different than the charge on the top that causes ions to deflect, it causes electrons to deflect. How to overcome all that? We use a proprietary ion energy control system. We use atomic layer processes, which involve very rapid switching of gases and powers synchronously. And that's all enabled a differentiated way by the fact that we have chosen to use a small volume plasma, small volume reactor, essentially a chamber inside a chamber, which enables particularly fast gas switching.
But even if you pulse power, you need to evacuate the products from that pulse power discharge fast so that you can then cycle to the next step. So we do deposition, really atomic layer deposition, atomic layer etching all in the same system enabled by that proprietary design small volume confined plasma, which in fact Lam has had in its product portfolio dating back to the 1990s, early 90s. And it's been matured, obviously scaled to 300 millimeter, we were ready to scale it to 450. It's a very robust technology, which I feel is just now really coming into its own. It enables us to generate the best etch profiles.
It enables us to control the selectivity to that mask. And it has resulted in substantial growth in our dielectric etch market share such that we're number 1 in memory space for dielectric etching today. You can see here very rapid growth in the number of tools shipped. And you can see here an actual cross section of an etched memory hole structure. So we're very excited about this product.
The growth has not only come from 3 d NAND, it's also come from DRAM because the same product is what we use to etch the high aspect ratio capacitors in DRAM, which I used to think was the most difficult etch until we ran into this one with the alternating layers. That's just an etch through one material. So it's a little bit simpler even though the highest aspect ratios are a little higher. So now I want to turn to how we do the inside out fill. So we're going to replace the nitride with tungsten.
So what you see here, the green is the sacrificial nitride layer. It gets leached out usually today by a wet process. And then we apply an atomically conformal nucleation layer for the tungsten, which serves as a template for subsequent tungsten CBD that fills out that structure, as you can see here. And those tungsten lines now are the word lines in the 3 d NAND device. And the critical thing is to make sure that there are no voids.
And that sounds easy perhaps. It's not easy because you think about the structure and you've got to get into the middle of the structure, otherwise you have an open line and you have immediate yield loss. So you have to fill this structure from the inside out. I showed you schematically how that's done. You also have to wrap that tungsten around the memory hole which I showed you how we etched it.
It was subsequently filled maybe with polysilicon and you got to fill the tungsten around that pillar. So again, I said that everybody making 3 d NAND devices uses our product, the Altus ALD CBD Tungsten Deposition Systems, every single one of them because we have this atomic layer engineering control. We're able to fill without voids as you can see by the picture here on the right. You can actually see the tungsten here and you see the pillar is kind of in front of it, but that tungsten wraps all the way around. This shows you the ability to conformally on an atomic scale coat the whole structure with tungsten, which serves as the template.
And last but not least, again, these are long deposition processes and customers are very concerned about the cost per bit and we offer this in a multi station sequential reactor that whose architecture was pioneered by Novelis many, many years ago. And it's again another reason why we are winning in this market. Atomic layer engineering, void free filling even with high aspect ratios and a high productivity platform. You'll see those themes recurrent in everything that I'm talking about. So the net result here is that we are number 1 in deposition in etch and we have increased our market share overall in NAND.
If you look at 3 d NAND relative to 2 d NAND, right now we're forecasting overall a 7% to 10% market share gain. It's coming from being number 1 in the number 1 supplier in doing the stack dielectric deposition. It comes from being essentially the only supplier for the tungsten metallization and I didn't talk about it, but the back end of the line copper interconnect using our electroplating system SABR. We have a very strong position, essentially 100% position in the memory hole etch or the transistor channel etch and a very strong position in the less difficult slit etch using the same product. And we also have a dominant position in a dominant position in hard mask open and stair etch overall.
So very, very bullish about our position in 3 d NAND and excited about the future. Turning now to multiple patterning, which is another, as Martin said, huge opportunity for us to grow. It's being driven by our customers for higher density and feature shrinks. Multiple patterning has 2 big problems, similar actually to 3 d NAND and that there's a cost problem. And the cost problem comes about from the simple fact that you got more processes going on, a lot more processes in multiple patterning than you have in single patterning.
Whether they're litho etch repeated litho etch litho etch litho etch litho etch litho etch and so forth, or whether you're doing spacer based double patterning and I'll walk you through quadruple based spacer self aligned spacer based quadruple patterning in a second. So you have all those extra steps that adds cost, it adds complexity, it adds control problems, it creates a variability problem because each step has some distribution as shown here that adds to the total variability of the devices. That's not good. So anything we can do to reduce the variability of each step or perhaps reduce the variability of steps preceding the deposition in etch steps is a huge value to our customers. So let's see how we do that.
So I'm going to show a video now of how we do quadruple patterning. And the thing I want you to focus on most is how many times we're depositing something and etching something after starting with the initial litho. So here's the initial litho and we first transfer that pattern into the so called top mandrel. Then we deposit a spacer. Then we etch the spacer which will become the mask for the 2nd mandrel.
First we have to remove the top mandrel another etch step. Now we transfer that pattern into the bottom mandrel. And then we deposit another spacer, do another etch, do another mandrel removal, transfer it all the way down into the hard mask. The net result is a 4 fold decrease in the pitch between those lines and spaces. And I actually didn't count how many dep and etch processes that were in there, but there are a lot.
And that was just to make the mask. Now we have to use that mask here and transfer the pattern into the underlying film. So before we start with just one step, then we have all those other steps. Who wants to see the video again? I got time, right?
Okay, everybody wants to see the video. So I got to go back. Okay, should work from here. So again, we start from the litho, we etch the top mandrel, okay, that's 1 etch step. We deposit the spacer, that's 1 depth step.
Then we etch the spacer, that's another etch step. Then we remove the top mandrel, that's another etch step. Then we transfer the pattern into the bottom mandrel that's another etch step. Then we deposit a spacer that's another depth step. Then we etch the spacer, remove the mandrel, I'm going to run out of fingers here in a minute, and then we etch the hard mask.
I just made it. 9 different steps. Now there's actually more steps in there, but some of them we do in situ because those hard masks typically contain more than 1 film, more than 1 layer that's simplified in this picture. So let's start with the spacer deposition. We had a press release yesterday on this vector AL dioxide.
The key challenges, as I said, are variability in cost. Now there are 2 sources of variability when you do the spacer deposition. The one is the potential for variation in the thickness of this film center to edge across the wafer. You could also conceivably get some thickness variation within dye, but the bigger problem is within wafer. The other thing that happens is when you start this deposition depending on your chemistry and your conditions, you can cause erosion of the mandrel.
So we've kind of depicted that here, sorry, here this mandrel is thinner on the edge, let's say than in the middle. So minimizing that erosion to begin with, because any erosion of that mandrel is going to affect the final critical dimensions of what you're trying to pattern later on, but also minimizing that variability center to edge. So and then the second problem of course is again cost. There you saw a number of spacer deposition processes in multiple patterning. Those are extra steps compared to single patterning, customers highly motivated to keep the cost of those extra steps way, way down.
Why are we winning? We have best in class film uniformity within wafer. We have best in class mandrel dimensional control. I think that's probably the single largest reason why we have seen double digit growth in our vector ALD product in the last year alone. And last but not least, we have the highest productivity footprint.
In this case, we have 12 stations. The stack deposition was 16 in the smallest footprint in the industry. So we're dealing with a variability challenge, we're dealing with a cost challenge in a very differentiated way. Turning to the etch system that we use for multiple patterning, it's based on our Keio technology and I think the latest version is generation number 5, where we've continuously improved and added capability to this product line, but have based it on a common architecture that enables us to preserve the investment of our customers, their installed base investment by upgrading from generation to generation. The challenges here again are variability and cost.
The variability comes again from a center to edge problem in terms of the CD control as we transfer the pattern from that, first of all, defining the spacer and then transferring the pattern from the spacer mask to the underlying mandrels or the underlying hard mask. So there's a within wafer challenge. There's also a within dye challenge. And our approach here and again this was our 3rd press release yesterday, we talked about how we have shipped more than 200 of our Kios with the Hydra patterning system. So this is a subsystem that you can upgrade older versions to the system.
It's enabled us to substantially grow market share at a leading logic manufacturer as we said in the press release. And the reason is pretty simple. If you have incoming variability in the CDs as printed, let's say by the initial litho step that I showed, and let's say that variability is as high as 2 nanometers, it's not atypical. We can reduce that 2 nanometer distribution that's shown faintly here to less than half a nanometer by feeding forward the information from the scanner to the etcher making corrections with our hydro patterning system to take 4 times variability out, absolutely critical for multiple patterning, a big reason why we are winning in this space. The other thing you see that's really important is the within dye variability.
So the pattern on the left here was created by using more conventional plasma etching that is not atomic layer control that we announced last year at this time actually that we had developed this capability. And there may be some residual so called even odd variation in the initial CD that could come from how you printed or how the spacer was formed, that initial CD variation which is very difficult to see here if you just look at the initial mask is accentuated as you etch down in terms of different etch rates. This is called aspect ratio dependent etching and different profiles. But with the atomic layer etching approach, we can eliminate the transport effects that are responsible for this effect or this accentuation of the odd even CD variability and essentially wipe it out. And so again, the reason we're winning is best in class critical dimension uniformity.
It's enabled by atomic layer etching. We have a unique capability. This Hydra technology was first put out to the market almost 3 years ago. So, it took off a little bit slowly and now we're on a you know, the hockey stick actually happens sometimes and we're on one of those. But it was in development for 3 years even before we shipped the first one.
So this is not simple technology to reproduce. It has very sophisticated control algorithms embedded in the system. Those are the reasons we're winning. In addition, I said those masks contain multiple layers and we actually etch them all at the same time. That's called in situ processing.
It's a design principle that we have applied to virtually all of our etch processes for years because it leads to lower cycle times and lower cost for a customer to just leave the wafer in there, don't spend time moving it from chamber to chamber and etching the whole stack. And the Hydra system in particular gives us dynamic capability such that we can change the process conditions on the fly as we go from layer to layer in those complicated multi patterning stacks. So the opportunity in multiple patterning as Martin pointed out substantial and it's only getting bigger as we go forward. Despite the efforts of our customer and us to take costs out of multiple patterning to drive productivity up, There are going to be more and more spacer based multiple patterning steps across the spectrum of DRAM, NAND and logic going forward. Even with the advent of EUV, we see significant growth in the patterning space.
And make no mistake about it, we are collaborating with multiple people in the industry. We had a public announcement about a year ago of a patterning initiative with IMEC in Belgium. So that's just one of our collaborators. But we are working hard to enable EUV to happen. We think it's important for the industry that EUV be successful, get introduced as soon as possible.
We think the initial introduction will be at the 7 nanometer node in a limited way. It actually provides more opportunity for deposition and etch. We have some techniques to enhance the resolution of EUV and effectively lower the dose, which will hopefully enable EUV to be introduced at the 7 nanometer node in a significant way. But even with EUV, multiple patterning doesn't go away. Spacer based multiple patterning will not go away.
It's all question of what the relative costs are. There's a place for both of them and this plot represents what we think will happen in terms of the number of passes in total for deposition, etch and clean for foundry and logic from 16 nanometers down to 5 nanometers, we're looking at doubling the number of wafer passes in depth etch and clean and multiple patterning, assuming EUV comes in for both single patterning and there'll also be multiple patterning with EUV required by the time you get to 5 nanometers. So in summary, our leaders we have leadership in multiple patterning. It comes from being focused on variability reduction. It comes from being focused on cost reduction.
We have we are the market share leader in each and every segment of multiple patterning etch. I showed you many different kinds of etches in that simple cartoon, that simple video. We're the market share leader in each one of them. We I mentioned we have double digit market share growth in spacer deposition because of the ability to control the mandrel dimensions, the spacer dimensions itself, center to edge and because of productivity. We continue to focus on lowering costs and boosting productivity and lowering variability.
That's the key to keeping Moore's Law going for many more generations. And as Martin said, we have built an Lam ecosystem of collaborations and several of those collaborations are intensely focused on the multiple patterning opportunities and solving the problems for our customer, which again means reducing variability, reducing costs. And the net is we see an opportunity of 40% SAM growth in multiple patterning going forward. And last but not least, I just want to state that we are preparing for the next wave of inflections. Our strategy has been to identify inflections early, to work on solutions with our customers and other collaborators early, such that we can time the introduction of new products and capitalize on the fact that customers must upgrade, they must change tools, they must bring in additional tools.
For example, the Flex product that I talked about that defines the memory hole in 3 d NAND, there was no such etch at all, no need for that kind of reactor at all in planar NAND, completely new opportunity. So we're focused on those transitions in the industry driven by the needs of our customer. They create opportunity and we will be there with the right products at the right time. And I've illustrated what we think the next wave will be. There's still more patterning solutions out there.
In particular, there's something called directed self assembly. We have collaborations in this area to try and enable it. Again, it requires etch to pattern the DSA material. Nanowire, which is the technology we think that will come after FinFETs, probably at 5 nanometers and below. There are many more etch steps and deposition steps in making a nanowire than making a FinFET.
So there's going to be a lot of opportunity there. There's a lot of technology that needs to be developed to capitalize on that. The transition from planar NAND to 3 d NAND is just the beginning of what we see happening in memory space broadly speaking. There are many different kinds of memories that are being worked on, that we're working on with our customers. I've given one illustration here of magnetic RAM etch.
And you can see that we if you looked at these profiles years ago, you'd see a lot of slope and sidewall deposition. We made a lot of progress. We've had some breakthroughs both in terms of hardware and process to enable the kind of picture that you see here. So I used to think that patterning the MRAM would be the limiter for its introduction into the market. I no longer believe that that will be the case.
I think we can absolutely pattern these structures with high density. Interconnect metallization, there's a lot of efforts whether it's on the middle of the line contact edges or the back end of the line interconnects With the dimensions getting smaller, there are problems with potential problems with copper metallization, for example, for small vias. And we have already developed an electroplating cobalt product that's based on our electroplate highly successful electroplating copper product, SABR. And you can see that here. And then in the packaging space, we have a number of products.
We have an atomic layer deposition liner. Basically the vector ALD I showed you applies to high aspect ratio etching deposition lining for TSVs. We have a highly successful TSV Etcher called Syndeon that is capable of doing very straight high aspect ratio etches in silicon for 3 d packaging. And last but not least, again, we have an electroplating product SABR 3 d, which is now the number one product for electroplating on advanced packaging throughout the industry. And with that, I'll close and turn it over to Doug.
Thank you.
Okay. Thanks Rick. I remember when I stood up here last year I followed Dave Hemker and he had all those cool videos in there, right. Now I follow Rick, he's got all the cool animations. Next year, I'm putting animations in my stuff.
But all kidding aside, the stuff that you just saw from Rick is largely why you're going to see some pretty cool financials I think and is the manifestation of our performance. Martin showed you what's happening with our markets. The inflections are driving an expansion in our SAM that we believe grows again for the 3rd consecutive year our markets are growing as we look out to 2018. We're winning share because of the technical innovation that Rick just showed you. I'm going to show you what that means to the financial performance of the company.
So first let me step back and I don't know if this is a victory lap but look at what happened relative to Lam in 2014. Martin said some of this but I'm going to reiterate it, it's important. We're growing at double the rate of the industry which the graph shows you. We think WFE grew about 10% last year, we grew over 20%. Again, markets are expanding, the company is executing well, we're gaining share.
To me equally impressive, this has a lot to do with how we run the company. We grew operating income at double the rate of the growth of revenue. It's awesome to see leverage like this in the financial model. It's largely why our performance is ahead of where we expected relative to what I would have been talking about a year ago. Earnings per share up 45% and a pretty substantial return of capital to shareholders, $486,000,000 last year.
So when I step back and think about what we talked about about a year ago and what we've actually executed on, we've done better. And I'm going to show you how that compares to the financial model in a few slides. But to me when I look at this, this is industry best financial performance. It's got all to do with technology. It's got all to do with how the company is executing.
So I think 2014 was a phenomenal year. 2014 though wasn't a one off result. Always good when you're showing financial charts to have trends up into the right. All of these are up into the right. The timeframe that you're looking at and each of these bars represent 6 months, 2 quarters of combined numbers.
This is the duration of time since we brought Lam and Novellus together. So pretty substantial increase in the shipment profile of the company flowing through the revenue, leveraging the gross margin performance and then the steepest slope on the page again warms my heart as the CFO of the company is the bottom line earnings per share. You can see numbers that go kind of in the high $100,000,000 to $500,000,000 on a 6 month basis. That's the run rate the company is performing at today. So again when I look at this and think about it, this is best in class performance in the industry.
So I thought it'd be illustrative to spend a little bit of time and talk about well how are we doing this? What is enabling this to happen? So this has what this slide is trying to communicate to you is a lot to do with how we run the company. This is a lot of what and you can ask Tim when he comes up about how we do this. We've got something at the company we refer to as the land management system It defines how we run the company on an annual and a quarterly basis, so very predictable repeatable way.
It's got a lot to do with execution. So Martin described and Rick described for you I think the fact that our markets are expanding and we're gaining share. Our addressable market in terms of percent of WFE and you can see it on the page, gets above 30% by 2018, great place to be. On top of that, Martin showed you the objective that we have in etch to gain 4 to 8 points of share, in deposition also 4 to 8 points of share, and in clean 5 to 10. But it's got to do with the fact that we're in the right place and we're executing extremely well.
At the same time, again, part of how we run the company around that management system is doing it with control. It will balance the fact that we've got financial commitments to you guys and we have to deliver leverage in the model and we have. We have delivered leverage in the model which as you can see on the chart OpEx as a percent of revenue trending down. So again, pretty rigorous way by which we identify and this gets into the next one and allocate the right spending in the right place, prioritize it, adjust as necessary so that we make sure the inflections that coming up years from now are getting the proper funding. And that's a task that we do with a lot of discipline and a lot of rigor at a very detailed way.
At the same time, we've got an objective and an effort underway at the company to try to allocate more and more of our spending from the administrative area, from the corporate areas into R and D, so that we can place more bets and identify projects that we think have nice returns that will enable future revenue growth. So we're driving real hard to get more of the dollars we do spend into R and D and get as efficient as possible in the SG and A arena. And then at the same time, Martin referred to variability, kind of customer by customer, technology by technology and trying to be nimble and flexible about how we can prioritize those things, that's the flexible business model. And again when I think about all of the stuff that's on this page, this is the way we run the company. This is the land management system.
It's tried and trued and we've been using it for many, many years. So now let me talk a little bit about how we're performing financially. You'll remember at least those of you that were here last year and I see a lot of familiar faces, so I think most of you guys actually were here last year. Here's the model we put up last a year ago. Again these are scenarios.
So we triangulated around a wafer fab equipment spending level of $32,000,000,000 And in that we expected a little over $5,000,000,000 in revenue flowing down to 20% operating income in between 4 $5,000,000 $5.25 in terms of earnings. So I thought it'd be interesting to let's take the first half of the year and see how that stacks up. So what this shows you in this column is the number it says annualized, the number that's not annualized is the WFE number. So our outlook for this year $34,000,000,000 plus or minus 2, that's what we expected coming into the year, that's what we still expect today. But how is the company performing?
Performing really well in my opinion, right? $5,100,000,000 a year ago was the objective. Dollars 5,700,000 is essentially what the annualized run rate in the first half of the year was. We're executing to that financial model, 20% operating income flowing down to 20%. And then if you take the first half earnings and I should point out we're annualizing the June quarter at the midpoint of the guidance.
We haven't closed the June quarter yet, so I'm not signaling anything per se here. It's just the midpoint of the guidance. And if you just take that and annualize it, it's 5.7 dollars So not bad. Clearly we're running ahead of the model. So some of you might be here to hear what we have to say about the financial model.
Sit tight for a little while. Yes, it's time to give it an update. So I think I kind of described a little bit of what the bullets say. I'm going to transition now into a couple of things that I think are important to understand about the business of the company that we haven't talked too much about yet. And these are areas that are largely in Tim's domain.
But it's important to understand and those of you that follow the company know this, but an important part of the profit generation of the company is the installed base. These are spare parts. This is refurbishment, upgrades and services. Not as sexy as all of the stuff Rick went through, but a pretty exciting financial part of the company. This is about a quarter of the company.
This ebbs and flows quarter by quarter. It's not always 25%, but on average think of it that way. And we're targeting this part of the business to grow faster than the rest of the company. How is that? Why is that?
Well, 1, around the inflections the installed base grows and this part of the company benefits from that. That's 1. 2 is what the 3rd bullet on the page talks a little about is we're very focused on I characterize this as innovation in the business model where we're trying to identify things we can do for customers to increase the number of good wafers out from the installed base. So this is a win for both customers. I mean customers love this.
This is pure upside for them, more output, essentially free wafers in some ways and good for us because as we look at it we can identify things that we can do to make a little bit of money. So this is something that is a win win for customers. It increases customer trust and helps our bottom line. The third thing to think about in this area is Martin Martin had a slide that showed the growth of IoT. IoT or Internet of Things is going to drive an upgrade requirement I believe in the installed base which again I think is going to drive growth in this business over and above everything else.
So you can see on the slide, we're targeting this between 2014 2018 to grow at a CAGR of 13% as compared to the installed base growing at 10%. So not sexy in the way that you got leading edge technology here but a really important part of the financial performance of the company. So in another area, to me this is really impressive when you step back and look at what we've done relative to the supply chain and manufacturing. We are shipping record output today relative to the history of the company. At the same time, Martin described this already a little bit, we're improving the on time delivery of what we're doing for customers.
Huge win for customer trust, really important in terms of thinking about how we scale things here. At the same time, we're continuing to reduce and have reduced product lead times. It's phenomenal when you think about this, record shipments, on time delivery improving, lead times coming in. We're also spending time thinking forward to make sure that this performance continues. Things like consolidating the supplier base, rewarding stronger suppliers and streamlining things so that we deal with fewer people and that will enable a quicker ability of us to ramp.
And then finally at the same time we're increasing our offshore resources and capability. Two reasons, one it's a cost effective thing to do. Secondarily, it's close to customers. A lot of our customers are Asia based and so having day to day interaction with them helps. And you can see the line on the chart shows you that as a result of doing all this and scaling up the company, costs come down and this helps us deliver on the financial promise.
So these two things, again, your question is about it, you can ping Tim here when he gets up on stage, but this is critical to the performance that we've delivered and the performance we're going to deliver from a financial perspective as well as just pure operations. So I want to pause a little bit and talk a little bit about the balance sheet and a little bit about use of cash. This question comes up pretty often. Pretty healthy balance sheet at the company. We finished last quarter with $4,000,000,000 of cash on the balance sheet.
And for those of you that follow the company know in the March quarter we did our 1st investment grade bond offering raised $1,000,000,000 to shore up the onshore cash balance. That transaction is actually very well received from my perspective. And so when you think about this, it's actually pretty interesting to look at where the dollars have gone. And again this is roughly the timeframe between just before when we brought Lam and Novelis together through to today. These are cumulative where did the cash go.
1st and foremost the priority for cash is the profitable reinvestment in R and D and CapEx, right? You got to have lab tools and things like that and leaning at stuff. Again, this is how we make sure that we're winning 3 years from now to make sure that the next inflections that Rick signaled a little bit on that second to last or the last slide that we're going to win again in the future so that 3 years from now I'm going to show you another financial model that continues to improve. So that's one is make sure we're making the right investments and prioritizing where that goes in the business. And then second, when you look at what we've done with Calverturn, it's actually been pretty substantive since we brought the 2 companies together, dollars 2,400,000,000 returned to shareholders.
The biggest item here is the buyback. And I actually went back and looked at over this timeframe the average share price that we incurred to buy the shares back has been in the low 40s. So the fact that this was front loaded, we actually look pretty smart in hindsight with what happened there. And then most of you will know about a year ago we initiated the 1st dividend in the company's 35 year history. And a little over a month ago a pretty significant raise of that dividend by 67%.
So again just kind of to summarize, invest in the right areas of the business 1st and foremost. We generate lots of free cash flow and we've been pretty attentive to giving some of that back to shareholders. So let me transition to maybe what a few of you in the room are interested in hearing about. First I thought I'd set it up a little bit. So thought process around should we update the model or not, we're going to update the model but what we think about.
Obviously when you look at the 2014 model and our recent performance we've outperformed. If you look at us winning in the inflections I think we're pretty confident in the fact that we're going to continue to outperform. What we think about relative to this, profitable growth, Martin talked about this, making sure we're putting the right R and D dollars in the right place, getting as efficient as possible in the administrative areas of the company, trying to make sure we've got flexibility in the business model and then a continued focus on the balance sheet. So let me transition into how we're thinking about the model. I'm going to roll this forward to 2018 now.
So last year it was 2014, 2015 and then 2016, 2017 I'm going to roll it all forward a year. So first let me set it up a little bit, the 2015, 2016 mile. First and foremost I'd tell you these are scenarios. We're not necessarily ready to forecast WFE for next year. Martin actually had a chart in his slide that showed a 2016 WFE number.
That was an average of industry numbers by the way. So it's just not we're not ready to give you a 2016 number yet. I do believe that it's probably up next year but we'll give you the number when we get closer to the end of the year and we've done the analytics on it. But here's how to think about the profitability of the company. At a $35,000,000,000 WFE level, we now believe you've got revenue in the range of 5.8% to 6.3%.
Operating income at 22% and earnings between $6,000,000 $6.75 So let me roll this forward for you to 'seventeen, 'eighteen. The same $35,000,000,000 scenario again, it's a scenario. With the SAM expansion that we believe is in front of us as well as the share gains that we've got sight towards in a $35,000,000,000 scenario revenue moves up between $6,000,000,000 $5,000,000,000 $7,000,000,000 Leverage in the model, we believe we get to 23% operating income and earnings bump up about 1 dollars to $7 to $7.75 It's not hard to imagine that in this timeframe most of the industry publications have WFE a little bit higher. So I thought I'd give you a scenario where WFE is a little bit higher than 35. We chose 38.
So if in 2017, 2018 the industry spends $38,000,000,000 we'll target revenue in the range of $7,000,000,000 to 7,500,000,000 dollars I think we have ability to get to 24% operating income and that will deliver earnings between $8,000,000 and 8.75 So I think about the trajectory here, it looks pretty good. A lot of you guys are looking at this and saying, Doug, where's gross margin? You took it off the slide. True enough, I did. These numbers are put together largely with a consistent gross margin assumption to the model last year.
But we're really very focused on the bottom line and there's more than one lever to get to these bottom line numbers than gross margin. We're driving gross margin essentially to the same level that you would have seen 46% to 47% last year. But the bottom line is to us what is most important. So a couple of scenarios give you something to think about, something to update your models with. We'll be happy to take questions here in a little bit on this to the extent that you have them.
So let me summarize here. So I think we're really well positioned. Feel great about the execution of the company. We've executed ahead of the prior model, just gave you an update. Feel like the way we scale the company has been very impressive.
I think we are uniquely positioned around these technology inflections of 3 d NAND, FinFET, multi patterning and advanced packaging to take advantage of the growth in the market and the share gains that we're targeting. I feel really good about what we're doing with the P and L performance and returning cash to shareholders. So with that, we're going to transition into Q and A. I'll invite Martin, Rick, Tim up on stage and we'll be happy to entertain your questions for a while.
Who's got a question? I guess we should be matching our photographs. That's okay. First question over here.
Yes, we got Mike's in here.
Thanks a lot for the presentation. Martin or Doug, question on the algorithm of shipments turning into revenue is how should we think about when shipments turn into revenue? Is it different by customer? Is it any different than it used to be in terms of if you're talking about a $1,600,000,000 shipment number when that might turn into revenue or is there a dynamic where we don't get to a revenue number that's quite that high unless shipments go
a little higher? Thanks. Yes.
I mean in the old days, we used to have like a one quarter lag. So the shipments of 1 quarter would be the revenues of the next quarter. And today, really as a byproduct of what I talked about a little earlier and sort of cycle times and lead times and install timelines. I mean, we're just spinning much faster today. So a typical revenue quarter is going to see 50% to 60% turns of shipments that same quarter.
So it's spinning a lot faster than it did 5 years ago for sure. We'll take Tim next and then over here.
Thanks a lot.
So your numbers for 3 d NAND look similar to what applied put up yesterday. It looks like you're modeling something like 200 ks to 240 ks if I read the slide right, 2 20 ks to 240 ks added in 2016. And when I asked them the question, they said that about half of that is conversion and half of it is new Greenfield. So I'm wondering whether you've put thought into what the breakdown is between Greenfield and conversion that will consist of that incremental low 200 ks next year?
Yes. I mean the headline, frankly, there are 2 headlines relative to memory spending. And it's as relevant for planar as it is for 3 d, as is relevant for DRAM as it is for non flash. I mean, the discipline in terms of the supply and demand balance is really good. And the discipline relative to finding the lowest cost path to add the capacity is also really good.
And so the conversion kind of dominance is significant where it exists. It's actually really hard. It's much harder than you think it is to answer your question because every single customer is a kind of different way of executing this. Some people have a bunch of cleanroom space available, some people don't, some people have greenfield, some people don't. So we've done our best.
I think we got a blend of both honestly. So one of the headlines for us by the way is even if the language of a customer is conversion biased, because of what Rick presented, there is an addition of steps that is kind of relevant, which shows up to us for this kind of new equipment. So our SAM tries to deal with all of that complexity. It's a blend of both. It will be efficient, but there's a lot of new systems investment that's coming to the etch and deposition and that's our number.
Hey, Martin. Two quick questions. Number 1, in your guidance for the second half revenue being flattish it looks like compared to your WFE number for this year your revenue might be almost 3x the WFE growth compared to like 2x for the last couple of years. Is this year an anomaly or do you think something else is going on there? And I'll sort
of follow-up. Well, you've conveniently taken my revenue reference and applied it to a $34,000,000,000 reference plus or minus 2 to draw your conclusion. So I guess it depends on kind of where you are in the range. I would say the only headline that I really communicated in terms of what's changed since the last earnings call is the 3 d NAND piece is a little stronger. So the 34 plus or minus 2 definitely got a little stronger for us.
Our target is this twice as fast as the industry. If we do better than that, then obviously that's kind of icing on the cake.
And just as a follow-up for Doug, how do we think about the tax rate in your target model? Thank you.
Sorry, say it again, taxation? The tax rate, yes. Yes, there's a reason I give you a range because there's ranges on everything. I do think there's a bias for the tax rate maybe to tick up a little bit and that is what I've modeled in there. But the reason I range it the way I do is quite honestly I'm not 100% sure.
We're going right at the frontier and then back on the left hand side there.
2 of the ear also.
Okay. Those 2 guys there first, then we'll come to the front. 2 in the right also. Yes, go mic, hand the mic please. Okay.
Srini from Summit Research. The question that I have is your shipments for second half are going to exceed your revenues because there should be a lot of shipments to Japan for Toshiba and Elpida. So would you say that that is true?
We're probably going to talk a lot about that in the earnings call. I'm going to try and kind of keep our conversation today focused on the long term. And then I gave you a reference point. The reference point is the September ship number and the first half, second half balance. And you're right, deferred revenue kind of got stronger in the company.
It's definitely kind of relevant, but you have to wait till the earnings call for the rest of that answer.
Hi. Thanks for taking my question. So I guess my question is on ALD. Competitor obviously talk about ALD, the new entrant in ALD or re entrant in ALD and obviously there's one big competitor out there that has been in AOD for a long time. You guys talk about the success you have on AOD product.
Can you maybe break it down a little bit and talk about where you think you have strong have you have been strong and gained share? Is it in three d NAND? Is it in logic? And then which area you think there's more room for opportunity for you guys?
Yes. So most of our gains to date as I showed have to do with spacer based multiple patterning. And that's more in the DRAM space than not. Going forward, we have a couple of new products in development that are actually positioned at a variety of customers to address the need for lower temperature nitride films, lower K films of a variety of sorts. We have one reactor which has a again, we focus a lot on productivity.
And ALD like ALE involves a lot of switching of gases back and forth. So a lot of effort on chamber volume is where we expect to establish some differentiation in those, for example, in the low temperature nitride space. We also have a new reactor to do low k spacers, which probably will have an impact broadly speaking across memory and logic and foundry in the near future.
I think one more there as well, please. Yes, it's Mehdi Hosseini from Susquehanna. Two questions, one for Doug. What is your inherent assumption for yen given now that tail is a standalone company? And for Martin, is there any opportunities for acquisition in advanced packaging to help you consolidate that segment of the market and better scale the technology?
So, Mehdi, it was relative to Paymapel, what is our expectation? That was the question.
No. Expectation inherent expectation for yen. When it comes to competitive landscape, what are you expecting for Tokyo Electron now that they're standalone and given the fact that yen has depreciated by 20%, 30%?
Yes, good question, Manny. By and large, customers aren't making decisions with a 1 or 2 quarter change in currency as the thought process about technical differentiation. It's about cost of ownership. And yes, the yen may benefit that particular competitor a little bit. But at the end of the day we win because of technical differentiation and cost of ownership advantages.
And honestly and Tim you can comment. I don't think we've really seen any impact from the yen dollar variation relative to competing with that customer.
No, I think that's pretty fair. As Doug said, any difference in yen is quickly overwhelmed by other benefits of the tool, so.
2nd part of the question. So, one of the things I said is the advanced packaging opportunity for the company is the smallest, the fastest growing and the last to show up. So to the extent that there are inorganic or organic opportunities, we have a little bit more time to go execute. The specific answer to your question is maybe the threshold for doing deals in this industry is really high. I mean, it is a hard thing to go to get done.
And you have got to be able to articulate clear value and benefits to the customer. And if you can't do that, it isn't worth doing. So for me, there's a maybe always and it's the last of the inflection SIM. We have more time, but it's a tough threshold. Maybe, Tim, you want to give a few comments on advanced packaging generally?
Sure. As Martin talked, it is one of the key inflections and it's an area that we've seen tremendous strength because like the other inflections, depth and edge and clean intensive, in fact, the last 12 to 18 months, we've seen our revenue more than double in that space. And there are some unique characteristics about that market. As I think by the virtue of your question, it is more fragmented, meaning the types of competitors we have in the space are a little different. And one thing that we've done, which has been quite successful is we've actually established a dedicated team within our company that focuses on the advanced packaging unique requirements of both products and markets.
And I think that has helped us develop a set of products now that are quite well targeted to that space. And it's about it contributes we expect about 10 percent of the future SAM growth and we expect to do even better than that as we continue to gain share.
Hi, Martin. Thanks for taking the question. My question is regarding your market share position. Like when I look at AMAT and Tel both have put a target model saying that they will gain a substantial amount of market share over the next 3 years, you are saying something similar. How do you think like it plays out?
Do you see that there is a difference in competitive dynamics going forward or it's just that it's always been a very competitive industry and there's no fundamental change going forward? Well, I
mean, it's a competitive industry, yours has been and always will be. And I think through the entire supply chain, the kind of risk reward profiles are going up, the stakes are higher, the complexity is kind of higher. So at a very fundamental level, we can't really opine on our competitors' disclosure anymore than we think they're valid opining on ours. We're really focused on building competitive differentiation. We believe in etch, which probably is the basis of your question.
We are gaining more share in this year than we did last year. So our momentum is positive and more of our momentum is in the dielectric etch space than conductor, but conductor is also strong and growing for the company. Thank you. We'll go one more over here and then we'll go left side.
Good morning. Harlan Sur with JPMorgan. You guys talk about the inflection technologies on some of your earlier slides, I think exiting this year, 3 d NAND about 10% penetrated 20 nanometer DRAM about 35% penetrated, FinFET pretty small. If I look at that and I think about 2016 and let's say under the assumption of a normalized 9% to 10% unit growth rate for next year, do you guys anticipate WFE spending up in 2016?
So I would say yes, but how much kind of too early to really tell. I mean,
I think the fundamentals
of the discipline on the supply side and supply demand balance, the capital intensity conversation we touched environment of maybe a more positive macro environment for sure.
And just a follow-up question. On the announcement yesterday on the Kiel Hydra and the wind that you have, I think you guys announced like a fivefold increase in the application base on your customers' 10 nanometer process over 14 nanometer. You just give us a rough sense in terms of how much that translates roughly in terms of dollar increase per unit of capacity spend by this customer? I got it. No.
Nice try, Harlan.
Nice try, Harlan. Hi, Atif Malik, Citigroup. Martin, question on the China opportunity. Can you talk about your bookings or shipment funnel? How does it look like for next year?
And what has changed in the last 6 months with respect to China memory opportunity?
Yes. I mean, so at a macro level, position in China for us from a share perspective is very good. We have kind of very good participation in the business. If your question is biased to kind of the IT 2020 agenda and how that ultimately kind of plays out for our industry or our customers' industry, it seems and perhaps yesterday's kind of message reinforced it that the primary focus of that initiative was semiconductor device and design and maybe there's a play in equipment. But we've been living in a world of competing with a domestic alternative in China for 7 years plus now.
And at a regional level, China is one of the strongest market share positions in our company. Kind of current momentum, any headlines, Tim?
Yes. I think that China is a place where many years ago we focused on building very deep collaborative relationships with R and D. And the partnership they have with Lam, I think, is very trusted and highly valued. I was there last week and there's clearly a strong ambition to grow in that market and I think Lam will play very significant role there.
And Doug, you did about 50% ASR last year and you're exposed to all the inflections and what not. I mean, what's holding you back from putting a long term target for total shareholder return for the company like a fixed number?
You're asking a TSR?
A TSR number for the long term model.
To me the TSR is largely going to be dependent on the P and L performance of the company, which I think we've just very clearly established as it pertains to the cash piece of that, right, the dividend portion. I think you've seen kind of what our intention is, which is to grow it over time. I don't feel like we need to put a numeric target out there except to communicate to you we know we need to continue to drive this and we will and we expect to be in line with the best in class in the industry.
I want to kind of get back to Harlan's question actually. So one of the things that I will say is, our objective and I think everybody has the baseline here to know that the history of the company in the logic space in etch was not where we wanted it to be. It's a multi technology node transition. This isn't a pendulum swing. It takes time.
But our objectives over 3 to 4 technology nodes is we've got a market share at the logic and foundry community that is very, very equivalent. And what we tried to describe yesterday is a meaningful step along that journey. But it's not throwing a pension, it takes a lot of hard work and a lot of time. I saw a hand up. Go here, please.
Yes. Thanks for taking the question. So first question, you talked about a 15% SAM expansion between now and 2018. And so curious when we start thinking about 7 5 nanometer adoption of EUV, how should we think about the implications there, Q? And what are the sort of upside, downside in terms of number of layers and what that would mean, whether flat line, grow, decline, etcetera?
And then I have a quick follow-up.
You want to take it, Tuner?
Well, as I said in my talk, even with the advent of EUV, which we think will be limited to a small number of wafer passes at the 7 nanometer node. The increase in multiple patterning that relies on deposition and etch is going to grow significantly. If EUV is a little bit later, then obviously there are more passes available to us. But even with the advent of EUV and when it's used as a single patterning wafer pass for say you're doing holes or cut masks or block masks. We have developed some techniques in collaboration with others to lower the dosage of the EUV required, which effectively translates directly into a boost in throughput.
That can enable EUV to be introduced at the 7 nanometer node. I don't think it can come in earlier than that. And it's more business for us. So maybe a little less spacer based double patterning and I emphasize little, but there'll be more applications. So net result is the growth in the number of wafer passes of a factor of 2.
It could be higher than that. I don't think it'll be lower than that. So I think
we will change the kind of content of the presentation a little bit this year. So last year we called out the EUV passes. This year we've kind of represented our marketplace. So we talked about the multi patent steps in etch and deposition. But I think that 7 nanometer baseline is 5 to 8 EUV passes, is that our headline?
So that's our assumption that now is inversely presented in terms of etch steps. What Rick also said, which hopefully is very clear to everybody, we're changing kind of the economic intercepts all of the time. I mean, we have an investment to Rick's point around enabling EUV. We also have investments around extending immersion, right? I mean, every single day that passes, we're changing the economics of multi patterning for the reasons that Rick presented.
And it's a huge opportunity for competitive differentiation. It's a big focus for us.
Very helpful. And then I guess as a quick follow-up, you maintained your WFE outlook $34,000,000,000 plus or minus for this year, but it looks like you uptick your 3 d NAND equipment shifts to 150, I think you were talking about $1,500,000 before. So curious where have you seen
It's up to minus. And the basic message I try to convey it is in our 34 plus or minus 2, we've gone up a little bit since the last call. So there isn't a fundamental message in terms of DRAM or foundry or logic pushes and pulls. There's pluses and minuses.
Yes. I think if
we had to identify one area perhaps where at least our own internal expectations may have gotten a little ahead was in a little bit in the foundry coming in this second part, but not at the leading edge. It's a little difficult to forecast that slight off leading edge demand, but it is very minor as Martin suggested.
One there and then we're coming in the middle.
Yes, Martin. Just quickly on the logic side of the house for WFE. A lot of discussion that maybe Moore's Law is moving from a 2 year cadence to a 3 year cadence on logic. Wondering you agree with that and if that is the case, what does that do to either the absolute level of WFE in a given year or the volatility of WFE in any given year?
I'll let Rick speak to the first piece of that. Yes. It sure doesn't feel to me like Moore's Law is slowing down at all. I see things actually accelerating. The pace at which we're going to go from 32 layers to 96 plus layers in 3 d NAND is going to happen on the cadence of every year.
I see 7 nanometer likely being accelerated not pushed out. So I don't think in the foreseeable future, we're seeing a slowdown in Moore's Law. And
from a WFE perspective, we I should have said it when I was presenting my slides, Doug kind of bailed me out, but I gave a slide with a multi year WFE statement, Slide 10, I think, not ours. It was just an average of the Gartner numbers and VLSI numbers. So I just kind of gave you that reference. We've already kind of articulated, we think maybe 16 is stronger than 15. We provided a capital intensity statement, which kind of shows the economic substance of the technology nodes conversions here.
And we have scenarios. So hopefully with all of your expertise and competency, you'll answer your question much better than we ever would.
And then Martin, as a follow-up, in your prepared comments, you talked about unit growth going forward kind of approximating what it has been historically, but you kind of threw an incremental driver in there around IoT. And I'm just kind of curious, I think the perception is that IoT is not really a driver of leading edge silicon. Is that wrong? Or do you actually see kind of capacity needs at N minus technology which could help bolster the WFE forecast in future years? Sure.
Let me take a shot real quick and then Martin can supplement. Right now when we how we have modeled IoT within our forecast is much more thinking about it from a service opportunity or service and upgrades perspective, at least through the first round of IoT. What IoT eventually becomes, I think none of us exactly know. But it's an opportunity I think for us to create SAM and growth opportunity for the company that's actually not reflected within kind of the traditional WFE capacity. We are establishing a refurbishment and upgrade center in Asia, specifically with the idea that customers, whether it be IoT or other use perspectives, are going to be looking for low cost, quick turn ways to upgrade equipment.
And we think that's going to bear fruit. We've also developed some productivity upgrades, which we think will be of high interest to fabs that really for 10 years or more probably have not implemented significant upgrades. They're going to want to do IoT at the lowest possible cost. We think we've got services and upgrades that can help them. And so I think that's the way we see it and the way we've modeled it in at least the near to medium term.
And I do personally agree with your premise that it's going to drive memory demand, right? You're going to have to store somewhere all the data that gets created. That's going to be by and large leading edge silicon. You're going to transmit that data. That's by and large leading edge silicon and you're going to have to analyze it also leading edge silicon.
So I think it's a net positive in addition to what Tim said.
Yes. So two quick questions Martin. First you had seen a big spike in the etch market over the past 2 years. So my question is if you look at from 2015 to 2018, how do you see the trajectory of the overall etch market? And can you provide some clarity between front end and back end or dielectric or conductor?
Where are we going to see the big next big growth amongst these sub segments? And can you contrast with the deposition market growth during the same period? Thank you.
I can't answer it quantitatively in the way you've asked it. I mean, what we've tried to do is speak to the portfolio of the company, right? I mean, we gave very specific SAM expansion numbers for etch and deposition and clean combined. What Rick tried to present is the etch story and the debt story both have very significant inflection consequences in patterning. In the 3 d NAND space, the deposition piece is probably an accelerator relative to the etch, but the etch piece is significant to some expansion as well.
So, we're not going to break out elements of the portfolio. Hopefully, the headline for the company and the presentation that Rick made gives you confidence that it's very relevant. Each of these inflections is relevant for all elements of the portfolio, and we're working hard to strengthen the competitiveness of the products. Sorry, there was a second part to
your question as well. Okay.
1, and then is there anybody over here? Okay, 2.
Thanks. I have a question to Rick on the ALD product. Applied presented a product which looks similar to multi sessions sequential deposition. The change with the change being I think they have a different gas plumbing for each session. So I'm just wondering if you're familiar with that system and can you compare that to the Vector platform and are you can you implement the same strategy in your Vector platform?
We have some familiarity with that product. And it's a technology, it's called spatial spatially based ALD, if I understand it correctly. And yes, it has some, I would say, somewhat superficial similarities to the vector approach. We've looked at that technology. We opted not to pursue that approach.
I think there are some significant engineering challenges, especially with respect to keeping variability down and control, very tight tolerances and so forth. So we've opted for actually exploiting our vector based design, which uses a it's time based atomic layer deposition where you process the whole wafer at a time. And I think that has significant control of benefits. And for the bulk of ALD applications, we actually think it has productivity benefits as well.
You mentioned magnetic RAM, resistive RAM and cross point. When do you expect U. S. Technologies to ramp?
I'm hesitating because I don't I have to be careful what I say because that's really our customer's decision. But I would say that new memory you'll see introduction of new memory in the next few years, whether it's specifically what kind of RE RAM or whether or not it's MRAM, I don't think I'm at liberty to say. I'm not sure anybody knows exactly when. But I think this is something that's going to happen over the next you'll see it happen a little bit gradually over the next 3 years. And then after that, I think it'll be much more substantial growth.
I think related, one part of answering that question is the slide that I presented. I think it was the technology roadmap slide or the land of opportunity is the way I described it. There's 3 generations or so in the 3 d NAND roadmap and the DRAM roadmap, both in front of where we are today from a technology point of view. So lots of legs with kind of known products and technology. We'll take one more question and then I'll wrap this or maybe no questions.
Sir, one more.
Any more questions? Okay. So I wanted to just wrap this thing by saying what I said at the beginning. We really appreciate you taking the time to spend with us today. We know you'll have lots of choices.
Appreciate your interest in the company. And as Audrey mentioned at the outset, I think in the room, this side of the conference room, we will be hanging out. Dave Hemker, our CTO will also be joining us. If you would like to ask us more questions, we're available for the next hour or so. Thank you very much.