Welcome to the Applied Materials Earnings Conference Call. During the presentation, all participants will be in a listen-only mode. Afterwards, you'll be invited to participate in a question and answer session. I would now like to turn the conference over to Michael Sullivan. Please go ahead, sir.
Good afternoon, everyone. I'm Mike Sullivan, Head of Investor Relations at Applied Materials. We appreciate you joining us for our third quarter of fiscal 2017 earnings call, which is being recorded. Joining me are Gary Dickerson, our President and CEO, Bob Halliday, our current Chief Financial Officer, and Dan Durn, who will be our next Chief Financial Officer. Before we begin, let me remind you that today's call contains forward-looking statements, including Applied's current view of its industries, product roadmap, share positions, revenue growth, profitability, and business outlook. These statements are subject to risks and uncertainties that could cause actual results to differ materially and are not guarantees of future performance. Information concerning these risks and uncertainties is contained in Applied's most recent Form 10-Q and 8-K filings with the SEC.
All forward-looking statements are based on management's estimates, projections, and assumptions as of August 17th, 2017, Applied assumes no obligation to update them. Today's call also includes non-GAAP adjusted financial measures. Reconciliations to GAAP measures are contained in today's earnings press release and in our reconciliation slides, which are available on the investor relations page of our website at appliedmaterials.com. Now I'd like to turn the call over to Gary Dickerson.
Thanks, Mike. With revenue and profits at all-time highs, I'm pleased to report that Applied Materials has now delivered earnings records for the past five quarters. We have tremendous momentum and a very positive outlook for the future. Our markets are growing with a broader set of demand drivers. Across a wide variety of industries, companies are making huge investments to drive transformative changes that shift economic value towards technology. At the very foundation of these emerging trends is Applied, and we are playing a larger and more valuable role advancing the innovation roadmaps in semiconductor and display. Our broad portfolio of technologies and businesses, combined with our strategy and investments, means that Applied has never been in a better position. This gives me confidence that we will continue to expand our available opportunity, gain share, and outperform our markets.
In today's call, I'll outline key industry trends as well as our strategy for sustainable, profitable growth. I'll then provide additional details about our markets and major businesses. Bob will then discuss how we're translating our broad portfolio of capabilities and products into differentiated performance for Applied Materials. After that, Dan will provide his initial impressions about our future opportunities, as well as our outlook for the next quarter. I'll start with the backdrop for today's discussion. These are incredibly exciting times for Applied Materials, and I strongly believe that we have more opportunities today than at any point in our history. There are three main reasons for this. First, our markets are strong and getting stronger. Pervasive demand for electronics means that our markets are getting larger and substantially less cyclical. New demand drivers are emerging that layer on top of traditional computing and mobility.
The Internet of Things, big data, and artificial intelligence will transform industries over the coming years. In healthcare, transportation, manufacturing, and retail, competition is increasingly dependent on capturing, transmitting, understanding, and storing data and images. In these industries and many more, value is shifting towards semiconductors and displays. To realize the potential of IoT, big data, and AI, major technology inflections are needed to advance the roadmaps in logic, memory, and display. These inflections are enabled by materials innovation, and as the leader in materials engineering, Applied Materials has a fundamental role to play. Second, Applied Materials is better positioned than ever before. We've aligned the company around our innovation leadership strategy, and we've made significant investments to accelerate R&D. We have developed a strong portfolio of differentiated products with more in the pipeline. It's the company's breadth and depth of products and capabilities that sets us apart from the competition.
We have, by far, the largest exposure to industry inflections, and we're combining our skill sets in deposition, removal, materials modification, inspection, and metrology to deliver innovative new solutions. Third, we built a platform that gives us sustainable advantages over the long term. We've done this by strengthening our technical and management teams and putting in place a company-wide operating system, the product development engine, that delivers repeatable success. Across the organization, I see stronger execution. Our product success rate is higher. We are transferring new technology to market faster, and we're using our breadth more effectively. I'll now describe the major inflections and investments within our markets. In memory, near-term market fundamentals remain strong, driven by high-performance storage for data centers and increasing smartphone content to support new features, including 3D cameras.
We expect healthy investment in memory to continue as the explosion of data storage requirements created by IoT, big data, AI, and streaming video has only just begun. To keep up with demand, customers are aggressively pursuing their 3D NAND scaling roadmap, which has four major levers: increasing number of pairs, shrinking film and stack heights, multi-tier schemes, and lateral scaling. All these approaches are enabled by advanced materials engineering and expand Applied Materials's opportunity. In foundry, market dynamics are equally healthy. In the near term, we see capacity additions at trailing geometries to meet growing demand for sensors and IoT devices, as well as strong investment in the leading edge. Looking ahead, advances in artificial intelligence are beginning to drive significant architectural changes, and customers are positioning themselves to win the major inflections in high-performance computing.
In data centers designed for AI workloads, we see logic content growing at least twice as fast as memory. As technology complexity increases, capital intensity is also growing about twice the rate of memory. In foundry logic and DRAM, dimensional scaling of devices remains a challenge with two primary areas of focus, resolution and placement accuracy. New innovations in patterning are becoming increasingly important as customers move to advanced nodes. First, self-aligned multi-patterning techniques, SADP and SAQP, are needed in conjunction with EUV lithography to drive the resolution roadmap. Regardless of the rate of EUV adoption, we expect our opportunity in traditional multi-patterning to grow. Applied has great momentum in this market and has gained 16 points of share since 2012. Second, new patterning approaches are being developed to address placement errors.
As devices scale, the accuracy of vertical alignment between interconnect layers has a significant impact on performance and reliability. Placement errors cannot be addressed by advanced lithography alone and will require materials-enabled solutions. Applied has unique technology in this area, and we expect our opportunity to grow significantly over the next several nodes. Looking at the market as a whole, our outlook is incrementally more positive. We now believe that wafer fab equipment spending for the calendar year will be up 20% or more compared to 2016. We also expect 2018 spending to be higher than our current estimate for 2017. In these forecasts, we made conservative estimates for spending in China. In addition, within wafer fab equipment, the ongoing shift to materials-enabled solutions in memory and logic opens up new opportunities for Applied. We now address 64% of the total market, compared to 53% in 2012.
Beyond semiconductor, our outlook for the display market has also strengthened. Display is growing even faster than wafer fab equipment as customers make multi-year investments to address large inflections in both TV and mobile. In TV, a major push to new Gen 10.5 substrates is underway. These huge 10 square meter substrates are ideally suited for manufacturing larger format screens, 60 inches and bigger. We now expect 13 new Gen 10.5 factories to be built over the next several years. At the same time, mobile OLED investment is getting stronger as customers prepare for broad adoption of OLED in smartphones. OLED enables new form factors that result in larger display area per smartphone, further expanding the overall market. I'll now provide brief updates for our major business groups.
Our semiconductor leadership businesses, where we have strong market share and highly differentiated products, are having an outstanding year, fueled by investment in leading-edge foundry and expanding applications in memory. Demand for our PVD and CMP products is especially strong. We expect the PVD market to grow by around 45% in 2017 as foundry customers adopt new interconnect technology, and the CMP market is on track to grow 40% this year on top of 40% growth in 2016. Our CMP revenues were at an all-time high this quarter and on track to exceed $1 billion for the fiscal year. In our semiconductor growth businesses, we have great momentum in markets that are growing rapidly. Over the past several years, we have gained significant market share in both etch and CVD. Our combined etch and CVD revenues are at record levels, and we're in a great position to drive further growth.
Service is an important part of Applied's portfolio that has grown significantly. This year, we're seeing an acceleration of that growth. We delivered record performance this quarter and for the year, we expect revenues to be up more than 15% over 2016. Demand for our Service solutions is driven by our rapidly growing installed base, plus our customers' need to shorten ramp times, improve device performance and yield, and rapidly optimize their factory output and operating costs. We are investing to expand the services we offer, and we are confident in our ability to sustain our growth in this business. In Display, as technologies and manufacturing are becoming more complex, Applied is in a unique position to enable the roadmap.
This year, Display revenues are on track to increase more than 50% over 2016. Based on our strong position, we expect to grow Display at least another 30% in 2018. Before I hand the call over to Bob, I'll quickly summarize. Applied is executing well, setting new records, growing our available market, and gaining market share. I would like to thank our employees for their passion to create value for customers and Applied. I'm confident that together we will raise the bar further. Applied is in a great position, working at the sweet spot of major trends that will drive tremendous shifts in economic value over the coming years. Finally, our breadth and depth differentiate us from our competitors. We have wider exposure to major industry inflections, and we are bringing together the broadest skill set and technology portfolio to create new solutions for our customers and accelerate innovation.
Now, let me hand the call over to Bob.
Thanks, Gary. Today, I'll add my comments about the business environment, describe how our capabilities and products are driving competitive and financial performance, and invite Dan to share his perspective along with our Q4 guidance. As you contemplate the markets, including record levels of demand for semiconductors and Displays, I'd like you to know three things. First, I feel increasingly confident that our markets are becoming materially and sustainably larger. Gary explained the big picture, and I'll add some data regarding the nearer-term environment. In semiconductor, two-thirds of NAND demand is driven by smartphones and SSDs. NAND demand for these devices is projected to grow by more than 40% in 2017 and 2018. Smartphones and servers drive the majority of the DRAM market, and demand from these devices should be up by 30% or more in both 2017 and 2018.
In foundry, our largest customer recently announced that its seven-nanometer tape-out expectations have nearly doubled, and we continue to see new investment at trailing geometries as well. The largest drivers of WFE demand are strong. In display, OLED is quickly becoming the preferred technology for leading brands, yet the installed capacity supports less than 40% of the smartphone market. An increasing technical complexity in semi and display gives us a growing opportunity in services. In short, I expect sustainable growth for our markets and particularly for Applied. Applied is uniquely well-positioned in this improving environment, and we're driving growth and competitive performance across our markets. Applied has incredible breadth, and we're taking advantage of it across our semi and display businesses to help our customers solve their highest-value problems with solutions based on unique combinations of our technologies.
The investments we've made in our new product pipeline are already impacting the market. In 2017, we're delivering strong double-digit revenue growth in every semiconductor device type. We project our share to be up for the sixth year in a row, and our share is balanced at greater than 22% across memory, logic, and foundry. In patterning, we see continued growth and share gains because we're shipping newly designed products that address both line shrinks and edge placement errors. We've also put our services business on a growth trajectory, adding $1 billion in annualized revenues since 2013, and we've built a unique and more diverse growth engine in display. In 2012 and 2013, our annual display revenue averaged $657 million. We expect to deliver a similar amount in the current quarter. Our display business is also becoming more profitable, and we're now targeting operating margins in the high 20s.
Applied has the broadest capability, largest served market opportunity, and strongest new product pipeline of any company in the industry. Third, perhaps best of all, Applied's innovation engine and operating discipline are driving higher profitability. The value that our technology brings to the industry is increasing, and that's reflected in our gross margin, which has grown by five points over the past five years. In Q3, our gross margin was the highest in nine years. At the same time, we remain focused on expense control. In Q3, our non-GAAP OpEx to sales ratio was 17.9%, a record low. As a result, we delivered our first quarter with $1 billion in operating profit, and we had record cash from operations equal to 36% of revenue. Next, I'll comment on our Q3 performance on a year-over-year basis.
We grew company revenue by 33% and non-GAAP gross margin by 2.9 points. We grew non-GAAP operating profit by 67% and increased non-GAAP operating margins by 5.9 points to a 16-year high of 28.7%. We grew non-GAAP EPS by 72% to a record of $0.86 per share. Turning to the balance sheet, we grew cash and investments to $8.3 billion, and 40% was onshore. We repurchased nine million shares of our stock and returned $482 million to shareholders, including dividends. Turning now to our segment performance on a year-over-year basis. We grew semiconductor systems revenue by 42% to a record $2.5 billion and grew non-GAAP operating profit dollars to a record of $920 million. We grew services revenue by 20% to a record $786 million and increased non-GAAP operating profit dollars to a record of $215 million.
We grew display revenue by 31% and increased non-GAAP operating margin by 2.6 points. In summary, I'm proud of the accomplishments of our teams, and I'm pleased with our top and bottom-line growth. Applied's innovation engine is delivering sustainable momentum for profitable growth. I'll hand the call over to Dan.
Thanks, Bob and Gary, and the entire team here at Applied Materials. You've given me a warm welcome and a great introduction to the technologies we're preparing to launch into a growing market for our products. I'm incredibly impressed by the depth of our technical talent. Applied has a world-class team, and I believe we can do anything we set our minds to. Gary and Bob talked about new advances in technology that are going to make dramatic change to our industries and to our lives. I saw this unfolding at my previous companies, and I'm excited to be here because these changes are enabled by semiconductors and displays, which begin with what we do at Applied Materials. I'll share more of my impressions about Applied at our Analyst Day in New York.
One thing that's already clear to me as a financial person is that Applied has a tremendous portfolio. The leadership semi businesses Gary talked about have some of the strongest product positions and profitability I've seen anywhere in this industry. The semiconductor growth businesses Gary mentioned are dramatically outpacing their markets. Etch is up 5X over the past five years, growing at two and a half times the rate of the market, and that's just one example. Display is growing faster than semi, and we have technologies in the pipeline that will enable us to grow beyond what we can do in OLED smartphones and LCD TVs. Services are a great way to drive growth and diversify the company in areas that are particularly stable and cash generative. Applied's portfolio is incredibly valuable.
I'm excited to be a part of this team, and I'll use all of my energy to help drive Applied's innovation engine and growth. I'll provide our fourth quarter guidance. In Q4, we expect our overall revenue to be in the range of $3.85 billion to $4 billion. The midpoint would be up nearly 19% year-over-year. We expect our semiconductor systems revenue to increase by about 14% year-over-year. We expect services revenue to grow about 17% year-over-year, and display revenue should be up by about 48% year-over-year. We expect non-GAAP gross margin of about 46%. Non-GAAP operating expenses should be $685 million ±$10 million, and we expect non-GAAP EPS to be in the range of $0.86 to $0.94, the midpoint of which would be up by 36% year-over-year.
To help you with your models for the calendar year, I'll share some of our early views on Q1 of fiscal 2018. We believe semiconductor systems revenue in Q1 is likely to be higher sequentially and higher year-over-year, and we believe our display revenue in Q1 is likely to be lower sequentially but higher year-over-year. I'll summarize by putting the company's guidance and momentum into context. In Q4, we expect the second highest semiconductor equipment revenue in our history, with continued double-digit growth momentum year-over-year. We also expect to set new record levels in services, display, and for the company as a whole. We expect record earnings per share. We also look forward to another year of growth in 2018 based on strong customer pull across the portfolio as we supply our enabling technologies to drive the new data economy.
I look forward to sharing our longer-term financial targets at the Analyst Day in New York on September 27th, and especially look forward to meeting many of you there. Let me turn the call back to Michael to start the Q&A.
Thanks, Dan. To help us reach as many people as we can, please ask just one question at this time. If you have an additional question later, please poll the operator, and we'll do our best to answer it later in the call. Operator, let's please begin.
At this time, I would like to remind everyone, in order to ask a question, please press star then the number one on your telephone keypad. We will pause just a moment to compile the Q&A roster. Your first question comes from C.J. Muse with Evercore ISI. Your line is open.
Yeah, good afternoon. Thank you for taking my question. I guess the biggest question here is sustainability in memory. It looks like you're guiding WFE around $43 billion ± this year, and roughly 70% of that looks like it's coming from NAND. Curious, as you look at that part of the market, can you talk to your visibility to both greenfield and shrink plans as well as your thoughts around demand elasticity and how you're thinking about spend from that business over the next 5+ years? Thank you.
Sure. I'll take a shot. Overall, we expect WFE in 2018. I'll drill into the memory too. We expect overall WFE in 2018 to be up from 2017. By device type, we think logic and DRAM should be higher next year. We think foundry and NAND should both be strong next year. I think the 70% of total spend in NAND, I thought you said, C.J. Muse, sounds a little high because we have NAND at less than 50% of total spend. I'll drill into the deeper dive on what's going on in the memory markets. Let me talk about DRAM and NAND together. I'll drill into NAND also. I think there's three important things to remember about customers, one about apply to both NAND and DRAM. 1, if you look at NAND and DRAM customers, very profitable now. I'll give you the numbers.
2, second thing to remember is demand is strong for both DRAM and NAND, looks sustainable. The third thing is actual wafer start additions have been modest. I'll give you the numbers. Finally, remember about Applied Materials, we've gained eight points of WFE share in both DRAM and NAND over the last four years. Let me give you more specifics. In terms of customer profitability, which was the first talking point I made, DRAM WFE spending as a percentage of total EBITDA for our customers used to be about 26% on average. That's DRAM WFE spending as a total of their total EBITDA. We think that's going to go down to about half, 14% this year next. You know they're very profitable. When you look at the numbers, it's very profitable. NAND similarly has gone down. Used to be about 26%.
We think it's down about a third this year next to 19%. That's also sustaining more higher growth rates in NAND. Okay? The second thing to remember is bit demand. If you look for smartphones and servers in both 2017 and 2018, we see 30% bit growth demand. Now, it goes down a point or two when you layer in PCs and stuff, but the big drivers are smartphones and servers. We see 30% bit growth demand or gigabytes. Secondly, in NAND bit demand, we see 40% growth rate in 2017 and 2018 each for smartphones and solid-state disk drives. On capacity additions, if you go look at the data over the last basically five years. We see this year basically the same, NAND, DRAM wafer capacity has been basically flat.
They get some bit growth, but because of their more layers, the actual wafers out of the building kind of goes down. It's kind of flat, not a lot of capacity added. If you look at NAND capacity additions, they spent a lot of money, but the actual capacity additions have been somewhat modest, a couple hundred thousand wafer starts up to 1.6 million now, and we project that they need about 2 billion in a few years. 2 million wafer starts per month, 2 million a couple of years after that. If you look at the three big things, customers are very profitable, both DRAM and NAND. Secondly, bit growth and bit demand is still very healthy, and we see it next year, too. Three, actual wafer capacity additions have been moderate. I think NAND's pretty good for the foreseeable future.
Great. Thanks, C.J.
Your next question comes from Farhan Ahmad with Credit Suisse. Your line is open.
Hi. Thanks for taking my question, and congrats on the great results. Just quick clarifications. On the June quarter, do you think it'll be a record level for SSG or not? Secondly, any color on the first half of next calendar year? Your commentary on 2018 seems to be fairly confident. Is it fair to say that the early trends that you're seeing in 2018, they are looking pretty strong right now?
Yes, it's strong. We said in the call that Q1 was going to be a good fiscal quarter for us, we said semi was going to be incrementally stronger than the comp and display not quite as strong as the comp in Q4 that we just guided to. If you look at, we were close to record SSG this quarter. I think the record was last quarter, Q3. I think there's a pretty damn good chance we're going to beat the record in Q1.
Thanks, Farhan.
Got it.
Wait, there's a second question. Did you have a second part to that, Farhan?
Yeah. No, I was just the first half of next calendar year , is there any visibility that you have that makes you confident on the next year overall?
What we said on the call, number one, was that we think next year WFE is up. We also think our position's really good, and we have a strong degree of confidence in our fiscal Q1, and we don't see that problematic after that. We think it's a pretty good year.
Thanks, Farhan.
Your next question comes from Atif Malik with Citigroup. Your line is open.
Hi. Thanks for taking my question. Good job on the results and guide. I just want to clarify the 30% year-over-year growth for display, is that for fiscal 2018 or calendar 2018?
Fiscal revenues.
Okay.
Calendar is good, too. I just don't look at it that closely on a calendar basis.
Can you help us out in how much of that is kind of market growth? I think in the past you've talked about an $18 billion display spending this year. Is that mostly market growth, or are you baking in contribution from new products? Thanks.
Yeah. If you look at it in 2017, 2018, our total revenues are going to be up in about 30%. If you go look at the market in 2017, 2018, it's going to be healthy, and our position's going to strengthen. What we alluded to a little while ago, that we're going to have some new product introductions, in fact, we already had one about a year or so ago, we're going to have more. I think we'll get some contributions from the market and our position.
Thanks, Atif.
Your next question comes from Harlan Sur with J.P. Morgan. Your line is open.
Good afternoon. Congratulations on the solid execution and outlook. You guys have discussed the growing number of programs for large screen LCD in your forward pipeline. A few weeks ago on their earnings call, LG Display discussed spending about KRW 20 trillion of OLED CapEx over the next four years. That's about $18 billion, and much of it is actually for large screen OLED, and I think the view is that OLED TV volume is going to grow from 2.5 million, 3 million next year to something like 6 million by 2020. Can you guys just remind us on your position in large screen OLED, your views on how you see the market evolving? Clearly, I think it feels like it's the next big driver after the mobile adoption.
Sure. I'll give you some factoids on the overall market, which you touched on, Harlan, and then a little bit on the OLED, and then Gary might chime on that one, too. I'll give you a factoid. I don't think we've mentioned it. From our call, which I think was three months ago, our expectations for the total market of display equipment has gone up for all five years in our window, 2017, 2018, 2019, 2020, and 2021. All five have gone up. We see strength across the board, basically, in both TV and mobile, a lot of big TV factories and increasing strength in mobile. It's a very strong market position, even more so than three months ago. In terms of our product position, we're going to continue to introduce new products.
In terms of OLED TVs, what they're kind of doing is this white backplane stuff with a color filter, I guess. It's not the same OLED as you have on mobile, but it is a version of OLED. Our position's pretty good there, but it's not the big technology inflection which has as much percentage growth and TAM availability for everybody as you have in true OLED on the mobile phones.
This is Gary. We've been increasing our share with pretty much all of the customers and certainly in this inflection also, we have an increasing position, so it's a very positive driver for us longer term. Relative to the inflections and kind of the market outlook, we're still in the early innings of OLED mobile, so as that continues to get built out, that's a big positive driver for us. This adoption of larger screens in TV, we talked about 13 Gen 10.5 factories. That will be built out over the next several years. If you look at the investments that large technology companies are making, VR, AR, automotive, there are other types of display technologies, foldable displays, those other inflections that could be more significant if you look out at a 3- to 5-year timeframe. We have good exposure to all of those changes.
Thank you.
Your next question comes from Krish Sankar with Bank of America Merrill Lynch. Your line is open.
Hi, thanks for taking my question, and, hey, congrats on the great results, folks. I had a question on display. Samsung has been going around saying their display CapEx is going to be down next year, but you guys are guiding to up 30% for revenue, and I understand you're getting some new products introduction. A two-part question, how much of the growth next year is driven by new products versus the existing product line? Along the same path, can you split up your display revenue by LCD versus OLED for this year and next year? Thank you.
Well, first I'll do the market, then I'll do a little bit on us. We think the total spending by display manufacturers is up this year and up again next year. We think it's pretty strong in both TV and mobile. The big TVs are driving the CapEx spending on TV, and then mobile is mostly this OLED driver. We see upside in total, we see upside in TVs next year, and we also see some upside potential in mobile next year. Part of that is you see a proliferation to mobile phone display manufacturers beyond the big one in Korea. If you think of the factoid we gave you the last couple of quarters, if you go back to our sales of OLED equipment in fiscal 2016, roughly two-thirds was to the biggest manufacturer of OLEDs.
We've now proliferated to 10 different people buying some equipment from us. The big guy, in terms of commitments of orders this year, fiscal 2017, is significantly less than 50%. What you're seeing is next year can grow because TVs are good and you have a proliferation of more people making the mobile phones.
Thanks, Krish.
Your next question comes from Toshiya Hari with Goldman Sachs. Your line is open.
Thanks for taking the question and congrats on the strong results. Gary, I was hoping you could elaborate a little bit on your view on China into 2018. You talked about having relatively conservative assumptions embedded in your WFE forecast for 2018, what exactly are those assumptions, and what would be the bull case over the next 12 to 18 months? Thank you.
All right. Thanks for the question. China is one of our strongest regions in both semiconductor and display. Our business has grown a significant amount. If we look at 2017 versus 2015, we're up 2X in semi, 50% growth in service, and 50% in display. In 2016 to 2017, we anticipate we're going to be up more than 20% overall in China. We track all the projects. We have very close relationships and very high share in China, so we're tracking all the projects for all of the customers, and we have leading indicators for the most likely investments. There's a lot of discussion about investment, but we qualify that based on these leading indicators. Based on all of that, 2018, we think we'll be up an incremental at least $1 billion, probably in the range of $1 billion-$1.5 billion in 2018.
Longer term, we think that this number is going to continue to grow because the strategic nature of the investment in China, they're trying to grow the percent domestic content, build a secure supply chain, and
It's a long-term strategy. The investments are not as efficient. Many of the investments next year will be in pilot lines in China, but we believe that this market will continue to grow over time, and our position is very, very strong in both semiconductor and display.
Thank you, Toshiya.
Your next question comes from Joseph Moore with Morgan Stanley. Your line is open.
Great. Thank you. I want to ask about the growth in services. You talked about 15%, seems pretty good. I know you've had the strategy of kind of trying to move that opportunity. Can you talk about how much of the growth you're seeing is just because of the significant increase in the install base of tools versus some of the other things you might be doing to drive that growth? Thank you.
Sure. If you look at it, if you go back at a longer-term historical perspective, we used to be kind of flat around $2 billion a year back around 2013. This year, we're gonna do about $3 billion. If you look at the rate of growth year-over-year, it's 15.9%. If you look at the quarters of the year, you'll see the momentum picks up even throughout the year. I'm pretty confident that next year is a strong year for us, and probably the momentum is as strong, maybe even a little better than this year for several reasons. One, if you look at your service business, it's a function of two things, the market, your position, and your strategy to attack the market. Three things I guess, market, product, and strategy. If you look at the market, WFE's pretty damn strong.
It's a pretty big number, this year, next year, that provides a toolset that you can service. If you look at percentage of the market for us, we used to be 17.8% of WFE just for 300-millimeter tools back in 2013. Last year we were, what, 22, and we're gonna go up again this year, we got a bigger piece of the market. Thirdly, if you look where we gain a lot of share, we gain a lot of share in places like etch, which is pretty good entitlement for service. All the dynamics around the market, our position, the products we're selling is raising our service entitlement. Secondly, what we've done, or thirdly, that was market and our position. In terms of our strategy has improved more and more.
If you look at it, instead of selling parts and ad hoc labor, we are selling long-term service contracts. We talk about getting 1,000 extra service contracts per year. That means we have a more predictable, more sustainable, and greater value for the customers and ourselves that again raises our service entitlement. My guess, the rate of growth in the service business could be pretty strong next year.
Thanks, Joe.
Your next question comes from Romit Shah with Nomura Instinet. Your line is open.
Yes, thank you. It seems like the DRAM industry is at a crossroads here whereby, with the conversion to 1X bit supply growth is gonna decelerate into the teens range, and that's well below what people are thinking in terms of DRAM bit demand, call it 25%, 30%. It would seem like next year we do see a capacity increase, and I know you guys are positive on that segment for 2018, and I was hoping you could just share your assumption for capacity additions in DRAM.
I'll give you my opinion, Gary can jump in. If you look at the dynamics, three out of our four big memory customers do both DRAM and NAND. I personally believe they have slightly different strategies in DRAM and NAND. If you look at NAND, the elasticity of demand for NAND is probably pretty damn big. There's a lot of growth in NAND, and they're going to eat into hard disk drives. They're going to spend there and grow their addressable market and make a bunch of money. If you look at DRAM, I don't think the elasticity of demand's quite as big. If you look at shrinking, you nominally get 28% more bits, roughly, per wafer, but because there's more layers, you get less wafers out of the same building bits because more steps. That's where the capital intensity comes in.
You're sort of getting 14% growth in kind of output of a factory. It's getting expensive to keep growing effective bits. Now what do you do? Well, if you have a total company strategy, you're probably going to invest more heavily a little bit in greenfield for NAND to grow the addressable market and make money. You're probably going to do more heavily weighted on conversions for DRAM. What we see, our estimates in 2017, you probably got about 75,000 wafer starts adds and converts at 240 to 250. The net adds is probably even a little less because I think it's like 50 or maybe a little less, because you're kind of effectively shrinking your output a little bit with all those others because of more layer count. We look at 2018 similar numbers, kind of 75 adds and converts at 290 maybe.
It's more weighted to conversions than adds because of the economics.
Thanks, Romit.
Your next question comes from Patrick Ho with Stifel. Your line is open.
Thank you very much, and congrats, and a personal congrats to you, Bob. As the CFO, you're leaving on a very high note.
Yeah, I want to say it's been a lot of fun, and Dan's taking me out drinking every night this week.
Looking at the foundry market, clearly you're benefiting right now from capital intensity trends going to the 10 and 7-nanometer node. Could you discuss maybe how that node may develop over time? Could you see that node being potentially as large as the 28-nanometer node, which has been very profitable for the chip makers? Is it possible that this is the next big node that we see, and this could be a multi-year event for the foundry segment?
I think the answer is yes, and yes. I want to give you more than a one-word answer. If you look at its peak 28 nanometer, where I think peaked out at 335,000 wafer starts. Earlier in the year, about nine months ago, our estimate that seven nanometer would peak out, it's kind of 10 going to seven. Seven, I'll call it, would peak out at, originally we said 250, then 260, then 280. We're currently at 300,000 wafer starts. Might be upside to that, because what you get is more and more applications for seven-nanometer devices. There's going to be all the typical processor stuff, mobile phone stuff, but you're going to get more and more into this cloud stuff. You're going to get into the cars with probably 7nm class. It's probably going to go long and big, right? Sort of like 20 nanometer.
The good news, there's a bunch of good news in there. One, it's big, that's obvious. Two, it's kind of long. You start to roll forward and you say, "Hey, does a lot of this roll to five, or does it stay at seven?" Because if you think of the product lives in places like the car industry, which aren't big now, you stay at a node for a while. There's probably not as much reuse, my guess, down to five, right? The third thing which is really interesting is if you look at capital intensity is picking up from 28 to seven, you'd say, "Well, okay, Bob, it's 335 to 300, you're down 10%." Capital intensity from 28 to seven nanometer nodes is up over 90%. To make those wafers, there's a lot more equipment sales. Our position's really good.
It's going to be a bunch of equipment sold. The company's going to make a lot of money. It's going to be a big, long node.
Thanks, Patrick.
Your next question comes from Edwin Mok with Needham & Company. Your line is open.
Okay, thanks. I take my question. Also congrat to you, Bob. Just quickly on margin, I guess a two-part question. One, I noticed that your service revenue just hit a record, as you suggest, but margin actually came down. It was actually lower-
I'm sorry, Edwin-
Than last quarter.
Wait, Edwin, you pinged out a little bit. Would you repeat the question, please? I just couldn't hear it clearly.
Sure. No problem. I guess two-part question on margin. First is service margin. I noticed it came down this quarter. Wondering why, what's the driver for that? Display, with the revenue increasing by quite a bit this quarter and you guys guiding for higher revenue, is it possible that display get to the Semicap margin level sometime in the future?
First I'll do service margins. He's looking at operating margins. Here's the trend line and they can pull up service. Let me get the data here. Q3, Q4. I think service gross and operating margins are up a little bit in Q3 and up more in Q4, actually, and our trend line is they're going to go up. I think what you'll see, I'll give you some longer term perspective of where we're going. That might help. We used to be 23%, 24% operating margins in service, and then in full year 2016, we were probably 26.5% or 27%. This year, the guide is like 28%, frankly, and we're going to trend up a little bit more. The operating margins are trending up, gross margins are trending up some, too. I think generally the trend's pretty good.
What's driving the trend line in margins? Revenue growth's pretty darn good. Gross margins are trending up a little bit, and you get drop-through because you get the percentage gross margin and you get the leverage of the revenue growth. I think operating margins are going to trend up and gross margins trend up as a service.
One other thing I would add in terms of service is we've really done a great job in driving lower cost. We're moving a lot more content into lower cost regions. Customers are also facing big challenges in the technology ramps for logic and also for memory. Our opportunity, Bob talked about earlier, in service contract is also increasing our revenue growth. We're driving lower costs, we're driving greater value for customers, and that is increasing our overall operating margins, and giving us a momentum going forward in the service business.
Display you asked about, Edwin, also. Display operating margins and gross margins are tending to trend up. If you look at operating margins for display from 2012, they were about 9.8%. 2013 was 16.8%, 2014 was 24%, about 25% in 2015. 2016 was about 20%. 2016, we were investing. This year will be about 26%. Next year should be up with the volume. I think we're going to do pretty well on display. Now, if you look at the longer-term model, which we'll show on Analyst Day, last year, the Analyst Day model, we showed that generally, our operating margins and gross margins are higher in semi, but services and display are both, they're all increasing. What you'll probably get is high 20s in service and display and higher 30-somethings in semi, and the overall average goes up.
Yeah. Thanks, Edwin.
Your next question comes from Weston Twigg with KeyBanc Capital Markets. Your line is open.
Yeah, hi. Thanks for taking my question. I just wanted to dig back into the visibility you have with your large memory customers. The reason I ask is just six to nine months ago, your outlook for the year was substantially different. That implies very low visibility at some of these large customers. I'm wondering if that visibility has actually improved or if really you're commenting on the trends as you look into 2018.
That's a heck of a good question. I'll give you some historical perspective. We do forecasts historically based on, we talk to the account guys, we talk to the customers, and then the marketing guys look at trend data and underlying drivers. When we do a kind of a one-year forecast, it's a little bit more weighted towards talking to the account guys. The customers' communication of second half visibility to us is moderate. Sometimes they don't know, and sometimes they don't want to disclose it, frankly, because it's a competitive issue for them. My belief is that the Semicap industry probably underestimated the second half of this year. If we'd done a lot more deep thinking on it, we might estimate a little higher number in the second half. It was one particular customer who was much bigger than expectations.
If you go look at it, we've spent, me personally included, have spent a lot more time on root cause drivers. Not just on what customers tell us, but what are they selling? What are the devices going into? What is the content of the phone? What is the content of the PC? What's the content of the cloud? For instance, if you look at the delta from what we thought ended 2016, it was content. It wasn't unit count phones, it was content of phones, content in the cloud. That is continuing through this year. If you say, what device type is up this year from earlier expectations? It's across the board, but it's NAND, DRAM, and foundry. NAND's probably the biggest increase, but at one particular customer in particular.
We're spending a lot more time going to root cause drivers of what's driving demand. I probably have more confidence than I did a year ago about the outlook.
Okay. Thanks, Wes.
Your next question comes from Craig Ellis with B. Riley. Your line is open.
Thanks for sneaking me in and taking the question, Bob, wishing you the best as you move forward in an evolved role. I'm going to ask you to take another longer-term view of one of the end use areas as you did with Foundry, except this is in DRAM. What I was hoping you could do is just look out beyond the next few shrinks that we're likely to get to some of the alternative architectures. The question is really around the equipment intensity that AMAT might be seeing with any of the successor architectures that would be there for DRAM. Is it higher? Is it lower? Is it a mix? What's the company seeing longer term for the extension to non-volatile memory as we know it? Thank you. For volatile memory as we know it. Thank you.
I'll start with more general stuff. Gary'll do the more technical stuff. I generally am optimistic about the industry and the various device types, but there's going to be granularity to some of that. I believe that there's drivers, that there's greater and greater valuable applications for silicon than there used to be, because you're basically, A, leveraging all the physical assets in the universe more efficiently through silicon. Two, I think a lot of the incremental consumption of entertainment and healthcare will be leveraged by silicon. I think the cosmic drivers, whether it's on automobiles, hotels, healthcare, you name it, silicon's going to play a big role in this. Okay? Bad drivers. Now that said, how you make money in life is two ways: scale and differentiation.
If you look at the scale of these things, they're very big, including scale up at the cloud, where they have huge volumes. The differentiation is only solutions. What does that mean? I think you're going to have more different tape outs of more different type of devices than you've had before. I think you'll see the ASICs chips, you'll see the graphics processor units. I think you'll see different types of high-end, low-end, different types of DRAM, and I think you'll see different types of memory. I think you're going to get a proliferation of more devices and more designs, which is all good for us, frankly. Then if you go to, what do I think about alternative memory? I think NAND's gonna grow for a while into hard disk drive.
I think DRAM and other forms of memory, there'll be ones that segment the market and solve different products because the volume up there and the differentiator is so valuable that you can go to different design types. I don't think it's going to be one shoe fits all feet. Then thirdly, in terms of capital intensity, capital intensity by device type is up across the board from a few years ago, 30%-90%, right? I don't see that trend changing for all the new devices either. Okay.
I would also agree with Bob that you've got big investments being made by leading companies to drive transformative changes that shift economic value in major industries. That's going to create a great opportunity for memory, for high-performance computing, all of those different areas. If you look at what's happened in memory over the last few years and what's driving the market today, tremendous innovation in the architecture driven by materials innovation, what we do better than anyone else. We're working with very large companies on certainly extending current memory technologies, scaling 3D NAND, but we're also working on new memory technologies, I really believe that the pace of innovation is going to accelerate, really at the foundation of that is Applied Materials with materials innovation.
I'm pretty optimistic on the overall market opportunity, I'm also even more optimistic about our position to grow as these changes happen.
Thanks, Craig. Operator, I think we have time for just two more questions, please. Your next question comes from Jagadish Iyer with Summit Redstone. Your line is open.
Thanks for taking my question. Gary, you talked about Gen 10 fabs in the display side. I was wondering how much of it is going to be OLED, and as a bigger picture, if this large-scale transformation to OLED happens, are we going to see a new baseline to your display revenues? Thank you.
The Gen 10.5 factories are really focused on TV, and all of the 13 projects that we're tracking right now are not OLED factories. They're, again, large area. We're seeing an increase in terms of the area consumption for large screens, and it's way more efficient to go to these larger factories from a cost perspective. That's what's driving that. I mean, certainly we're also very optimistic about mobile OLED and the opportunities that we have there. We're still in the early innings, but the driver for Gen 10.5 is larger TV screens. I don't know, Bob, if you want to add anything else?
I think that's true. I do think there's a hell of a lot of upward directional arrows in general for the market. The new baseline of revenues is increasing for display, which was one of your second part of your question, Jagadish. If you look at the inflections, it's not just bigger TVs, in my mind, it's not just OLED. Those are really powerful. I think there's other inflections, like on your simple phone you have now, as you go to the wraparound, take the top and the bottom off the chin and the forehead, you get 40% more surface area. We sell surface area, whether it's layers or size. I think you're going to have foldable in a few years, which doubles the surface area of a phone or an iPad, I guess.
Eventually, you're going to have some OLED proliferation, as you mentioned, TVs, whether it's white OLED or something further down the road. All of these lead to more spending. The question people would ask a year or two ago is, "Gee, whiz, historically, you might have $32 billion in WFE spending and $8 billion in display spending. Shouldn't it stay at 25%?" No, because the drivers are very different. Display is huge application across the world in TVs, mobile phones that everyone has, and the layer count to increase the visual content of that is doubling. I think the potential display market's pretty big.
Great. Thanks, Jagadish.
The next question comes from Toshiya Hari with Goldman Sachs. Your line is open.
Great. Thanks for taking my follow-up. I had a question on gross margins heading into next fiscal year. Bob, you talked about display growing 30% year-over-year. I'm guessing, at least at this point, you don't have as high expectations for the semiconductor business. Do you think you can improve gross margins into next fiscal year? Thank you.
Our goal is to increase gross margins every year. I think this year we've done particularly well. We're going to be up a couple points, almost 3. We're going to be up to 2.9 this year or something like that. We're up a lot, okay? That's a lot of things. I think we've executed by product across the line. If you look at our gross margins by every single product in the company, every BU in the company, which number between AGS, service, and semi, like 13 businesses, every one of the gross margins is up from 13 to 17. Across the board, good execution. Mix this year's pretty good. It's a good semi year, but it's also a pretty big display and services year.
My take, there's an opportunity to go up next year, but we haven't run the models, but it'll be hard to do as well as we do this year. My average goal, and Dan's got to go through the numbers, what I've said in the past, is our average goal is 7/10 of a point a year. If we do anything like that, we're already ahead of the 2019 model. I think there's still room.
Great. Thanks, Toshi, for the question. Dan, would you like to make any closing remarks before we end the call?
Sure, Mike. I think it's probably good. Summarize a few key points from today's call. First, I hope you share our excitement that Applied's markets are substantially larger and more attractive than they've been historically. We expect another record quarter in Q4. We've got line of sight to continued momentum in Q1 and far beyond. Second, Applied is uniquely positioned to outperform our core market, further building on our leadership in semis, growing the unique opportunity we have in front of us in display, and expanding the service businesses. Third, we're more profitable than we've been, and we have many levers to keep driving the profitable growth. Lastly, I look forward to meeting all of you in New York at our Analyst Day on the 27th.
Personally, I can't wait to show you how we're going to keep raising the ceiling on our performance in the years to come. Mike, now back to you.
Hey, great. Thanks, Dan. We'd like to thank everybody for joining us this afternoon. A replay of this call is going to be available on our website beginning at 5:00 P.M. Pacific Time today. Thank you for your continued interest in Applied Materials.
Where's the [inaudible]?
This concludes today's conference call. You may now disconnect.