Ladies and gentlemen, thank you for standing by. Welcome to the ASML 2016 fourth quarter and annual financial results conference call on January 18th, 2017. Throughout today's introduction, all participants will be in a listen-only mode. After ASML's introduction, there'll be an opportunity to ask questions. I would now like to open the question and answer queue. If you'd like to ask a question, please press star one to register, and star two to withdraw that question anytime during the call. Your questions will be answered in the order they are received. If any participant has difficulty hearing the conference, please press star zero for operator assistance. I would now like to turn the conference call over to Mr. Craig DeYoung. Please go ahead, sir.
Thank you, Arend. Good afternoon and good morning, ladies and gentlemen. This is Craig DeYoung, Vice President of Investor Relations at ASML. Joining me today from ASML's headquarters in Veldhoven, the Netherlands, is ASML's CEO, Peter Wennink, and our CFO, excuse me, Wolfgang Nickl. The subject of today's call is ASML's 2016 fourth quarter and annual results. As always, the length of the call will be 60 minutes, and questions will be taken in the order that they're received. This call is also being broadcast live over the internet at asml.com, a replay of the call will be available on our website. Before we begin, I'd like to caution listeners that comments made by management during this conference call will include forward-looking statements within the meaning of the federal securities laws.
For a discussion of risk factors, I encourage you to review the safe harbor statement contained in today's press release and presentation found on our website at asml.com, in ASML's annual report on Form 20-F and other documents as filed with the Securities and Exchange Commission. With that, I'd like to turn the call over to Peter Wennink for a brief introduction.
Thank you, Craig. Good morning, good afternoon, ladies and gentlemen, thank you for joining us for our fourth quarter 2016 and 2016 annual results conference call. Before we begin the Q&A session, Wolfgang and I would like to provide you with an overview and some commentary on 2016, the fourth quarter, beyond. Wolfgang will start with a review of our annual 2016 fourth-quarter financial performance, with some added comments on our short-term outlook, I will complete the introduction with some additional comments on key 2016 accomplishments and some of our near-term expectations.
Thank you, Peter, and welcome everyone. 2016 was a remarkable year for ASML, both financially and strategically. I would like to first highlight some of our financial accomplishments, and then finish with our view of the coming quarter. 2016 was a record-breaking year in many financial respects, with total net sales, gross profit, net income, and earnings per share all reaching record levels. In addition, we finished the year with the highest backlog ever, which combined with our current business view, allows us to look forward to another great year in 2017, where EUV becomes an integral and growing part of our system revenues, contributing significantly to our top-line growth through the balance of this decade and beyond. Turning to our Q4 results. Net sales came in at EUR 1.91 billion.
Net system sales accounted for EUR 1.22 billion, driven by logic, which represented 61% of net system sales, with memory returning to strength versus Q3, accounting for 39% of net system sales. System sales included EUR 144 million of EUV revenue in line with the guidance given during our earnings call in October. Net service and field option sales for the quarter came in strong as expected at a level of EUR 684 million, driven by ongoing strong demand for holistic lithography options, high-value upgrades, and our growing install base. Furthermore, we closed the acquisition of HMI in November, and net service and field option sales include about EUR 25 million for this new and exciting part of our business. Our gross margin for the quarter came in at 47.2%.
This includes starting the amortization of intangibles as well as the effects from the fair value assessment of HMI's inventory as of the closing date of the acquisition. The negative impact on gross margin for both of these purchase price allocation related items was approximately one percentage point. R&D expenses came in at EUR 287 million, slightly higher than guided due to both the R&D expenses of HMI and the start of our partial funding of ZEISS SMT for our High-NA EUV program. SG&A expenses came in at EUR 107 million, also slightly higher than guided due to the inclusion of HMI. We also had an impact from foreign currency revaluations on transactions and balances relating to the HMI acquisition. You may remember that this was an unfavorable effect of about EUR 28 million in Q3, as reported during our last call.
For Q4, we had a more than offsetting favorable effect of about EUR 83 million. These effects are reported in the interest and other line in our P&L. Moving to the order book. Q4 system bookings came in at EUR 1.6 billion for 44 systems, including six 3400 EUV systems. Strong bookings continued in the logic sector in support of the 10 nanometer ramps and in support of EUV insertion at the seven nanometer node. Memory booking strengthened further from its strong Q3 level, supporting expected strength in memory shipments continuing in 2017, driven by DRAM. Continuing order flow for EUV systems brings our total year-end EUV system order book to 18 systems. Our overall systems backlog now stands at nearly EUR 4 billion.
A major driver was our free cash flow of EUR 1.1 billion in Q4. As we experienced in the last quarter of 2015, we saw a significant level of early payments from customers, which will impact Q1 2017 cash flows. Also, as already mentioned before, we closed the HMI acquisition during the quarter and also issued a EUR 750 million bond to support part of our planned strategic investment in ZEISS SMT, which is expected to close in Q2 of 2017. With that, I would like to turn to our expectations and guidance for the first quarter of 2017. We expect continuing sales strength in Q1 with total net sales of approximately EUR 1.8 billion, of which an estimated EUR 30 million will be deferred EUV revenue. Foundry shipment strength supporting 10 nanometer ramps will continue in Q1 and will be firmly supported by memory shipments.
We also expect to ship our first NXE 3400 EUV system in the quarter. We expect to record the revenue for the system in the third quarter of the year since this system will ship in a non-final configuration. I would also like to mention here that one of the EUV systems that we expected to ship early this year, postponed from Q4 last year due to a customer readiness issue, will not ship this year as originally planned. Due to other extenuating circumstances, this customer has now decided to place a system upgrade order for this tool and will take delivery of it in 2018, where it will add to two other systems at the customer site to be shipped this year.
This leaves our system output plan at 12 new systems and our shipment plan at 13, considering the one additional system that missed delivery for material availability reasons in Q4 2016. We expect our Q1 service and field options revenue to again come in above EUR 650 million, driven by continued demand for holistic lithography options, high-value upgrades, and our growing install base. For now, we will report HMI revenues under field options and services. Gross margin for Q1 is again expected to be around 47%, including the effect from the purchase price allocation for the HMI acquisition. The negative impact of these purchase price allocation adjustments for Q1 is more than one percentage point. The impact for the full year is about EUR 90 million and will reduce to about EUR 40 million per year from 2018 onwards.
R&D expenses for Q1 will be about EUR 320 million, and SG&A is expected to come in at about EUR 95 million. The uptick in R&D spend is driven by the inclusion of HMI and accelerated investments in pattern fidelity metrology. Our contributions to SMT's High-NA development, our own High-NA development acceleration, and the strong U.S. dollar. As a reminder regarding our share buyback program, last year we purchased EUR 400 million worth of our own shares before the program was paused during our acquisition of HMI. It remains paused for the time being as we close our planned investment in ZEISS SMT. The transaction is in the regulatory approval process in the required jurisdictions. We already received the approval in South Korea and expect the approval from Germany and China in time to close the transaction in Q2 2017.
Finally, an increase of our annual dividend from EUR 1.05 to EUR 1.20 will be proposed at our annual general meeting of shareholders in April. With that, I would like to turn the call back over to Peter.
Thank you, Wolfgang. 2016 was indeed a remarkable year for ASML, not only financially, as highlighted by Wolfgang, but strategically and product technology-wise, as we continue to prepare ourselves for continued growth in the coming decade and beyond. I'd like to take a moment to highlight some of these key events. Firstly, our acquisition of HMI will enable us to move from simple imaging and imaging placement significantly in the direction of full pattern fidelity control, which is a key requirement and value provider for our customers at seven nanometer and below. This, along with an expansion of our current products into broader types of applications, allows us to lay the foundation for future growth in our holistic lithography products group.
Secondly, we brought a key supplier partnership to the next level by agreeing to acquire a minority stake of Carl Zeiss SMT, our major critical optics supplier, for the purpose of not only securing the extension of EUV imaging technology, but also strengthening the current combined businesses of the two companies to improve cooperation and governance. The co-investment of about EUR 760 million over a six-year period centers on the R&D investments of next-generation lenses and related capital expenditure, also referred to as High-NA, which is critical in meeting our customers' imaging roadmaps throughout the next decade. Lastly, probably most significantly, the industry has turned a corner on EUV. Throughout the year, we continued to execute on mutually agreed performance milestones, which allowed our customers to grow confidence in the technology and tool performance.
This resulted in customer decisions to allocate their most critical layers of the next-generation nodes to EUV, beginning with the industry seven nanometer logic node. With customers backing this confidence up with orders, we ended the year with an EUV backlog of 18 systems, as mentioned. Together with the anticipated Q1 orders, this will cover our 2017 and early 2018 EUV output. As mentioned before, our output capability for 2018 will be around 24 systems. Based on the timing of High-Volume Manufacturing introduction of the advanced logic and memory nodes, as announced by our customers, it is realistic to assume that this production capability will be fully utilized by our customers. This means we would expect a continued order flow in the coming quarters.
As we move into the next phase of industrialization of EUV, our focus will remain on continued improvement of key HVM performance metrics, very much in line with what customers expect of our DUV offering, above 90% availability and productivity per tool specifications. Furthermore, we are heightening our focus on our supply chain to provide the required number of EUV systems on time and on preparing a competent and sufficiently sizable EUV field service organization capable of supporting our customers in the volume manufacturing insertion plans. With respect to our core product lines, we continued to develop industry-leading imaging systems, evidenced by the introduction of our latest immersion product offering, the NXT1980, which has demonstrated the fastest ramp of a new product in our history, with 46 systems shipped in 2016.
For our holistic lithography product line, 2016 was also a good year, where we launched new products penetrating new large customer accounts, promising significant new business in the years to come. Together with the aforementioned progress on EUV product performance and related business opportunities, we feel confident in further anchoring our leadership position in the semiconductor equipment marketplace. Turning to the short term, as Wolfgang mentioned, we expect to build upon a record 2016, seeing further significant opportunities to grow in 2017. We see a continued ramp of the foundry logic 10 nanometer node, as recently confirmed by one of our large foundry customers, with memory strength driven by expected stronger bit growth in 2017. In addition, service and field option sales are again expected to continue to grow in 2017.
This will continue to be driven by sales of scanner system upgrades and our growing holistic lithography product offering. In 2017, we'll see the first real impact of EUV system sales, with recognition of systems shipped in the calendar year, as well as pieces of remaining revenue recognition from systems shipped in the past. For 2017, our opportunities and challenges are crystal clear. The commitment of our customers to EUV is now apparent and evidenced by our significant and growing EUV backlog. Supporting our customers' intent on moving EUV into volume production in the coming 24 months will remain our number 1 priority in order to ensure that our customers can deliver their next node transitions as planned. ASML remains committed to do everything within our capability and power to bring EUV to manufacturing readiness as soon as possible. With that, we would be happy to take your questions.
Yeah, thank you for letting us take some extra time to review in some depth key events of 2016 and their impact on our future. The operator will instruct you momentarily on the protocol for the Q&A session, but beforehand, as I always do, I'd like to ask you kindly to limit yourself to one question and one short follow-up if necessary. This will allow us to get as many callers on in the hour as possible. Now, Arend, could we have your final instructions and then the first question, please?
Of course, sir. Thank you. Ladies and gentlemen, at this time, we will start the question and answer session. As a reminder, if you have a question, please press star one to register and star two to withdraw your question from the queue. If you're using speaker equipment today, please lift the handset before making your selections. One moment please for the first question. First question comes from Weston Twigg. Please state your company name followed by your question.
Hi. Thanks. Pacific Crest. My question is related to the memory strength. You mentioned that it should stay strong through 2017, particularly on the DRAM side with bit growth. Was just wondering if you're seeing new DRAM capacity being installed, given the pricing trends, or if these are tools that are being shipped in the end and then replaced in the DRAM fabs? Yeah, it's more of the latter. We've seen significant relocations from DRAM capacity into NAND throughout 2016, already starting in 2015, and that's really filling it all back up. Like we said in the previous call, we saw quite a significant capacity drop in 2016 in terms of wave starts per month in DRAM, which is a double-digit drop, and that's being filled up with new DRAM capacity.
Okay. I guess as my follow-up, are those upgrades on the NAND side, are those being upgrades that are contributing to your field and service options, or is that a different revenue line?
Always.
Are those even being upgraded?
It is the first. These are the upgrades that are in the field service and option line.
Got it. Thank you.
The next question comes from Kai Korschelt. Please state your company name, followed by your question. Please go ahead.
Good afternoon. It's Bank of America. The first one was, Peter, just to clarify your commentary around the EUV shipments next year. I think I heard you say, it's realistic to expect 24 tools. What would be the sort of puts and takes on kind of whether you would ship capacity or maybe potentially slightly less in terms of adoption and capacity? The second question was on the Chinese CapEx, I think you said something on the video on your website. Some of the data points that have been making the rounds or the press, suggest that there could be tens of billions of EUR spent on new memory fabs in China. I'm just wondering in terms of the phasing, timing or magnitude, how much visibility do you have on those projects? Thank you.
Yeah. Okay, thanks. On 2018, the determining factor is not so much our capability or the performance of the tool. It is really, as you can understand, the next nodes of our customers have complexities in there that are not only driven by our capability or our lithography capability. This is really based on the introduction timing of our customers. When our customers or all our major customers make public statements about when they want to introduce their next nodes and start using EUV, we take those statements, and actually, we don't only take the public statements. To be honest, we have very detailed discussions with those customers on those public statements and what that means. We get very detailed plans of when they need what tool when.
That is really driving the, let's say, realistic assumption that, when we have 24 units to sell, we will sell 24 units when they stick to their plans. This is basically what it is. There's nothing more to it than that. Customers will decide. That could be that if one customer says, "I need to, for whatever reason, need to do an introduction six months later," there will be a six-month rescheduling. That's what really drives the business. China CapEx. Yeah, it's interesting. There was a lot of interest following every comment that comes out of China on the tens of billions that will be invested in all kinds of fabs all over the place. You also have to look at what impact that will have on our short-term business.
Short-term, yes, there are many ceremonies of fab openings or fab extensions, I would say. The number of real new fabs is limited on, let's say, Chinese-owned companies. Of course, there are the foreign companies that open fabs in China. I think the reference to the tens of billions are really Chinese-owned companies. That is definitely a promise. When we look at the next 12 to 18 months, it is good. There's some good logic opportunity, there's some memory opportunity there, but it's all within the realms of good business and not the extraordinary growth that some people are portraying. That will very likely happen. It's very likely also going to take a bit more time than the 2017, let's say, slash 2018 timeframe.
Okay, great. Thank you.
The next question comes from Sandeep Deshpande. Please go ahead, state your company name followed by your question.
Hi, Sandeep Deshpande, JP Morgan. Just a quick question, Peter, on the order intake. In the fourth quarter, you took orders for six EUV tools. At the same time, we have some expectation that you will sign volume purchase agreements with some of your large customers for these EUV tools. Is it now that you will be taking these EUV orders as a normal part of the business, as you seem to have done in the fourth quarter? Or are we to still expect volume purchase agreements in the next few months? I have one quick follow-up question on the 2018 shipments.
Yeah. On the order intake in Q4, that was based on a volume purchase agreement for the, you could say, the 3,400 delivery. It does not include yet the additional options that they need on those 3,400. You could say if there's a fully signed and closed volume purchase agreement with that customer, not entirely because the options are still under discussion, what they need and the economics of it. Yes, those orders were indeed taken under the agreements that we have on the pricing. Yeah, and the pricing models. That's just for one customer. Other customers will follow suit as we continue in 2017. For those who will introduce EUV later, those VPAs will be also signed later on in 2018. As you said, it's the normal course of business.
Following on to 2018 shipments, you're going to have this capacity for 24 tools, potentially, you've already got six of those orders at this point. By which point do you need to get all your orders so that that capacity will be readied for the customer? Because you've said in the past that you take almost a year to ship these tools, and then there is a time which the customer takes to install the tool in their own facility and get it stabilized.
Yeah. Well, the time to ship and install is about a year, so it's not plus. It's not for us a time as a year to ship the tool. What our time is, we can install an EUV tool in about three to four months. We have currently an internal cycle time in the factory of about six months. All in all, for us, it takes about nine months from start of the tool to get it installed at the customer site. The customer needs to, of course, start to qualify the production. The issue is in the supply chain. In the supply chain, we need lenses, we need large mechanical modules that need to be produced, and this is limiting our output capability right now.
However, having said that, to your earlier questions, we are in very intense and deep discussions with customers on Volume Purchase Agreements. It is also clear that those customers need those tools at a certain moment in time. Taking into account our cycle time reduction plans, it is not absolutely necessary throughout 2017 that we need to keep onto this two-year type line. That will go down. That will be shorter because we will reduce the cycle time as volumes go up. However, to be able to ship 24 units in 2018, those systems need to be booked by the end of the year, so in this year. If you ask me when should those orders come in, is it Q1, is it Q2, or early Q3? I don't know yet, it's not that important.
What I do know is that if we need to ship 24 units in 2017, those orders need to be in.
Thank you.
Next question comes from C.J. Muse. Please state your company name, followed by your question.
Good morning, good afternoon. I am with Evercore ISI. I guess first question, can you walk through how we should think about the gross margin for EUV through calendar 2017? Particularly interested how we should think about the inclusion of deferred 100% margins through the year. Thank you.
Okay. I will take that, C.J. We painted the picture at our investor day as well. The starting point is a gross margin, if you take everything into consideration, of about -75% in 2016. Our objective is to get this business breakeven from a gross margin perspective in 2017. In 2018, you would make another significant step forward. We are thinking somewhere in the 20s also. Of course, there are a few things that contribute to that. You are right, some of the catch-up revenue that essentially comes at no or low cost will help. It also will help that this year we are shipping the 3400, which has a list price that is about EUR 20 million higher than the 3350s. We are also going to continue to make progress on the cost side. We still have significant field wanting to do. We are making progress on that front as well.
If you then take into consideration also the service business, where, as you know, we are charging per wafer out. We got a growing installed base that is not productive in churning out a lot of wafers. We are still spending a significant amount of money on this without real revenue coming in. If you take it all into consideration, we are targeting around breakeven. As it relates to the total business, we were about 45% gross margin in 2016. I think we are going to continue to make progress in both businesses, EUV going from -75% to about breakeven, but also the non-EUV business, because the mix is shifting towards applications and more higher value systems, will make progress as well.
Overall, I think for your modeling purposes, you should assume that the total company gross margin will somewhat go down because you're growing the revenue significantly on the EUV line. I should expect a little bit of a step back there before then in 2018, we are marching towards the 50%+ that we're targeting for 2020.
Just as a quick follow-up there, are you thinking closer to like 43%, 44%? I guess as my follow-up, on the foundry side, orders ex-EUV came in fairly weak, I think sub EUR 300 million in the December quarter. I guess how should we think about the timing of a pickup there? Is that wafer really an issue and about the trajectory as we contemplate 10, seven nanometer ramp?
Without tying it down to a specific number, I think with everything that we gave you between Peter and the shipments on EUV and the margin, I think you're in the right ZIP code there on the gross margins. Relates to orders, first of all, we are thrilled, be it an overall backlog of EUR 4 billion. Having said that, same story as in prior calls. We publish the backlog because it gives you a decent structural view on what's happening in the business. You see memory picking up, you see a lot of EUV orders coming in. We always say that just merely looking at bookings or backlog should not be your main input parameter in the outlook. If you know the order at the end of the day, in particular for the more matured business is merely an administrative act.
We have VPAs with all of our customers, every other week, we get a detailed sales forecast and all that what's giving us. We were not the slightest bit disappointed about bookings last quarter.
Excellent. Very helpful. Thank you.
The next question comes from Timothy Arcuri. Please state your company name, followed by your question. Please go ahead.
Cowen and Company. Thank you. I had two. On EUV, continued availability. You had two tools that showed a four-week average of greater than 90% versus one last quarter. There was no improvement in productivity. Is that because it's not your focus right now? I guess I would've thought maybe you'd seen more than just one tool producing 1,500 wafers per day over that three-day period.
Let me answer that, Tim. First of all, the reason why we put those metrics in there, is one, to be able to communicate to you, but also to our customers. What we believed at that time when we created those metrics, the most relevant milestones were that customers used to get convinced that EUV was going to be the choice of their next generation leading-edge litho production. Having said that, meeting those targets, and not only those targets, it was the continued, let's say, showing of those targets throughout the year that actually raised the customer's confidence to the level to also publicly state, "We're going to use EUV." They followed that up with orders, as you have seen. Now, just for your information, we are over 200 watts, which will actually provide us way more than 1,500 wafers per day, and that's what we've shown the customers.
It is about the confidence that we will be at a high-volume production requirements by the time that customers need it. That was shown with those targets. The targets going forward, I would like to relate to are the targets that they need for high-volume introduction, like I said in my introductory comments. It's like with deep UV. We're moving into that direction. Like Sandeep said, it's business as usual. We have to be, and we will be over 90% availability when they need it, when they start HVM production. We will be at the productivity of wafers per day as specified by the tool specifications, i.e., 125 wafers per hour.
This is where we are, and this is why we're absolutely confident that we're going to get there, and that's why it's the last time we're going to give you these targeted numbers because those milestones have been met, evidenced by the customer orders.
Got it. Okay, Peter, thank you. Then, I guess, just as it relates to backlog on EUV. You had 18 systems in backlog. You have slots for 12 this year, so obviously, six of those are going to ship next year. Since you have 24 slots, give or take, next year, does that mean that backlog can only be 24 exiting this year? Is the policy such that if you get an order, even if it's going to ship 18 months from now, you're going to put it in backlog so that the backlog exiting this year can be actually a lot higher than the 24 slots that you have next year? Thanks.
That's correct. Like I said, currently, in 2016, we had a 24-month lead time. That will go down. It's not going to be 12 months by the end of this year. It's very likely that there are going to be 2019 orders in there.
Got it. Okay, Peter, thank you so much.
The next question comes from Amit Harchandani. Please state your company name, followed by your question.
Good morning and good afternoon. It's actually Amit Harchandani from Citigroup, thanks for taking my questions. Two, if I may. Firstly, my question is with respect to the technical milestones that you referred to earlier. Could you maybe talk about what are the key areas of improvement that you're working on within the tool that need to be completed this year or are on the verge of being completed that would take you to the targets you've talked about for commercial introduction in terms of the technological progress? Secondly, could you also give us a sense of what's happening in the wider ecosystem, particularly around defectivity and any other complements within the ecosystem? If you could kindly share any updates around those. Thank you.
Yeah. Make some improvements, some extensions. Taking out some of the quality issues that we know what to do. That actually brings us to those targets that customers need for high volume introduction, 90% plus availability and productivity at spec. On the ecosystem, I think two things, resist and pellicles. Resist, good progress. We get progress reports every now and then, and over the last couple of weeks, we got some good progress reports on photoresist and on photoresist sensitivity, on line edge roughness data and information that gives customers good confidence that by end of 2018, 2019, we're going to get what we want. On the pellicle. Pellicle, we have started to outsource the pellicle production to a supplier that actually should make the pellicles for the industry, for our customers.
Initially, there was a process that yielded low because pellicle still had some defects on it, on the pellicle itself. As you can imagine, if the pellicle is a membrane sitting in front of the photomask, you don't want any defects on that pellicle. Because they're big. Now that everybody is looking for. It is not our main concern. What our concern is, yes, that the supply base of those pellicles will be maturing also. So we get a constant flow of defect-free pellicles. If you look at the progress that we have made over the last six months, that gives us the confidence that also by the time when we need the H volume, the HVM requirements, we will be there.
Thank you, Peter. Just a couple of clarifications on what you said. Firstly, with respect to the technical progress, with respect to your tool. Just so that I understand correctly, there are all incremental improvements really around lifetime extensions, but there is no radical improvement that you need to do. All of that is behind us in 2016.
Correct.
Would that be a correct statement to make?
Correct.
Okay. Secondly, is the topic of actinic inspection tool when it comes to EUV behind us, or does that still come up in your conversations? Do you think the workaround is pretty much accepted now by all the customers who are looking to move on with EUV?
That's correct. We do not have any discussions on actinic inspection at this moment. The workarounds that we currently have, either through the pellicle use or the on-wafer inspection using e-beam tools, that is really what the solutions are that customers are currently using. The discussion on an actinic inspection tool over time and in the next decade might come back, but we'll see how effective the current solutions are.
Thank you very much.
Next question comes from Gareth Jenkins. Please state your company name, followed by your question. Please go ahead.
Thanks. It's Gareth from UBS. One follow-up, please, and one question. Just to follow up on memory. Your slightly more positive tone on this, does this include, in addition to the 1980s that you're talking about, some KrF business? Secondly, I just wonder whether you could talk about your expectation in terms of conversion at the seven nanometer node for your large foundry customers from 10. Thank you.
Gareth, on the last question, could you be a bit more specific on your last question on the conversion? What do you want to know?
I'd just like to know the sort of level of conversion that you expect between the 10 and seven node, given the commonality between the tools. Would you expect it to be more or less than we saw at prior nodes?
Okay. On the memory, predominantly yes, it's going to be predominantly 1980s. We always ship, if you add some extra capacity, some KrF. Going back to an earlier question, this is really backing up the installed capacity that was relocated to NAND, which were basically immersion systems, where they were upgraded in the NAND space. It's predominantly NXTs, but we always have some level of capacity space that is in existing fabs, which also need some KrF, but it's largely NXTs. The level of conversion from 10 nanometer to seven. 10 nanometer, there's always a level of conversion from 10 to seven, which would include upgrades. The level of commonality from a platform point of view is always there, but it's really the performance on overlay and focus that drives really upgrades.
When there is a reuse of a existing body in the 10 nanometer space or a previous node onto the new node, then you see upgrade business. This is a part of the business that we see growing in 2017. We did see a number of upgrades in the logic space, but also in the memory space from, let's say, previous platforms to the newest specification of the NXT platform really happening. That is part of the business growth that we see in service sales and options. Not much different than previous nodes. It does lead to a lot of new business in terms of system upgrades.
Can I just follow up and just ask what your expectations for the sort of 10s? I think you've always talked about wafer starts on the 10.7 geometries combined. I wondered whether there's been any change in the thought process around the starts on 10.7.
No. Currently not.
Okay.
The only thing that we can say is that, of course, we discuss with business plans. I only say that in flash anatomy, confidence that our customers have in being a big node, driven by more than just the smartphone applications. They're all talking about, and it's real, because, well, about customers in the automotive space, customers in the space that is dealing with artificial intelligence, and augmented reality, virtual reality, big data analytics. They're seeing customer applications in that space, and that is driving their confidence on the node side. They keep repeating it to us that, they strongly believe based on what they see in tape outs going big node.
That's great. Thank you.
The next question comes from Francois Meunier. Please state your company name, followed by your question.
Yes. It's Francois from Morgan Stanley. Yeah, I understand all the question around the gross margins and there are zillions of effects around it. There's one I would like to understand a bit more. I think you guys have been talking about PPA having a negative impact this year of around EUR 90 million. It's actually a non-cash impact. When you guide for 47% gross margin in Q1, actually, the cash gross margins is more like 48%-48.2% or something for Q1. Is that the right way to look at it?
Yeah. You got to be careful with the cash because there are other cash, non-cash related items. In principle, you're right. We're guiding to about 47%. If you just look at the two elements, one being the amortization of intangibles, we said that's going to be about EUR 40 million per year, that's linear, so it's about EUR 10 million. Then we said there is EUR 50 million that results from the revaluation of the inventory to a fair value or market value at the time of closing. We got to work ourselves through this, and as you can imagine, this is going to be a bit more skewed to the front of the year rather than the back of the year. Therefore, there is, as you state, a probably closer to 1.5% impact on the gross margin in Q1.
In other words, had we not done the acquisition and the deal with the purchase price allocation, we would have been north of 48% in the first quarter. You've got that correct.
Okay.
I wouldn't associate it with cash and non-cash because there's other stuff.
There's other things going on. Okay.
Yep.
Okay. Very interesting. Thank you very much.
Cool. Thanks.
Next question comes from Farhan Ahmad. Please state your company name, followed by your question.
Hi, Credit Suisse. Thanks for taking the question. I have a question related to EUV. What are some of the risk factors that you see going forward? Is there anything that you have to deliver for EUV to be adopted, or should we take these orders as a sign that EUV is now at a point that we can count on its delivery in 2018? Also, can you remind us again on what the lead time is for EUV going forward?
I think, ultimately the proof of the pudding is when customers give you an unconditional commitment to pay you a lot of money. Which I think has happened. I think that decision, which I think most customers have been public about EUV, when they want to use it and how they want to use it, and just following this up by orders. I think it is true that, somebody asked the question earlier, said, what could happen to make 2018 a year where you're not going to fully ship your production capacity? I just answered, one of the things is, I don't know, customer roadmaps might change, but what they are telling us today, and we know the number of layers that they want to use EUV on, then we have to use that capacity. Only if the customer change their minds, things will change.
That's not the case today. On lead time. Lead time, like I said, lead time 2016 was just a two-year. We had a supply chain that we had to kick out of hibernation. Well, they're now awake, I can assure you. Lead times will also compress somewhat also throughout 2017. I'd love to have a lead time by the end of the year of about 18 months.
Got it. Thank you. Can you talk about how do you see the linearity of 2017? Some of your peers have indicated like there is a stronger first half relative to second half. Is that something you see also? Related to the China 500,000 wafer starts, is any of it hitting this year or next year?
Two questions. On the linearity It's too early to say. Generally, we have lead times of about six months for our tools, customers probably give us a very clear indication of the next six months, which looks very good. The rest of the year is a bit driven by expectations, that's always a bit more uncertain to a certain extent. That might be the reason why some of our peers focus more on the first half, it's just visibility. On China 500,000 wafer starts next year. We don't see 500,000 wafer starts next year. It's too much. It will be there in the end, but I said it earlier, it's going to take a bit of time. With some of our customers, we've been talking about building new fabs now for two and a half years.
In that same time frame, our larger customers built the fab, we're shipping tools, while we're still talking about the others. This has to kick into a different gear also. The 500,000 wafer starts next year is absolutely not what we expect. Good business. Yes, we will expect shipments into new pedestals for our tools, into fab extensions and perhaps a new fab, but nothing to the level that you just mentioned.
Thank you, Peter.
Next question comes from Jagadish Iyer. Please state your company name, followed by your question.
Yeah, Summit Redstone. I have two questions. First, Peter, if you look at your immersion revenue systems, it has been pretty much stable through the last three years and in fact has trended up. I just was wondering if, as EUV starts to progress, how should we think about the immersion system trajectory over the next 12 to 24 months? Then I have a follow-up.
Yeah. The next 12 to 24 months, I do believe that when you look at what is driving our customers' business, it is 10 nanometer logic, and it's memory. That still needs those immersion systems. I think next 12 to 18 months, I don't think you will see a lot of change. On the longer term, when complexity of chip design increases, the number of layers will also go up. Now, the increase of the incremental layers are very much the critical ones, which is going to be EUV. As EUV progresses in terms of maturity and productivity, also EUV will grow into the realm of deep UV, leading-edge deep UV. No matter how you look at it, there's also, I think, clearly discussed at our capital markets day, immersion and deep UV will be with us forever. Yeah?
Also means over the next 10 years or so, going to be a very significant part of our business. For the next 12 to 18 months, it's going to be the majority part and the key part of our business, given the fact that those nodes that they are being designed into are the nodes that we're currently using and ramping, which is not an EUV node. 10 nanometer is an immersion node, and the high teens DRAM is an immersion node. That will be with us for the next 12 to 18 months.
Okay. Briefly, on the six EUV system orders that you got as a VPA, is it fair to conclude that you have met the 7-nanometer initial insertion specification with this key customer? Thanks.
Yes.
Okay, thank you.
Next question comes from Andrew Gardiner. Please state your company name, followed by your question.
Good afternoon, it's Barclays. Thank you, guys. Just one on your outlook for 2017. You've given fairly clear messaging around what you see on the logic space and on the memory space, and of course, services and options continue to grow. Just the statement around significant revenue from EUV for the first time. If I go back to the capital markets day in late October, you seem to be indicating at that point, something on the order of sort of below EUR 1 billion mark as a combination of rev rec at shipment, as well as the deferred revenue coming through. Is that still a reasonable assumption, given better visibility in how you see EUV trending for this year?
Within services and options for this year, now that HMI is closed and you're in the integration process, what are your expectations for that business over the next couple of quarters? Thank you.
Okay. I'll take that, Andrew. Hi. EUV, what we said before continues to apply. We said we can ship a maximum of 13 tools, which is the 12 plus one carryover. Revenue recognition is now close to shipment, or with shipment for a majority. Yes, there will be some catch-up revenue from last year, where we shipped systems with no revenue recognition. Without tying it down too much, but I think the number will have a one in for sure, and somewhere north of EUR 1.2 is clearly within the realm of possibilities. Service and field options will continue to grow, even if you start with excluding HMI. We grew 78% this year, and also based on some of Peter's comments on the upgrades, we think it will grow at least by that level. It could go 10% or so year-over-year.
Then we have HMI, which was only in for two months last year, and we're not intending to break this out in the future. You know from their standalone reporting that they should be somewhere in the EUR 200 million ZIP code. That is incremental.
I think you will see us announcing new products during the year, they will not lead to any significant revenue in the year. EUV, field options and services, and HMI are all growth drivers. Then you have the rest of the business that is stable, and in some cases, up a little bit. Like we said in our prior remarks, it should be a pretty good revenue year in 2017.
Thanks for the clarification.
The next question comes from Douglas Smith. Please state your company name, followed by your question.
Hi, it's Doug Smith from Agency Partners. I was wondering, can you break down the 18 EUV systems backlog into your foundry memory and IDM groups?
Yes, we could. We generally say that we're not doing it, because it would be very customer specific, because you could easily say who is who. That's not what I want, but there are a few memory orders in there, and that's just less than a handful, and the rest is mostly logic and IDM.
Okay. Just a clarification, were you saying that the six EUV orders in Q4 were all from one customer?
No. There were five from one customer. One was an additional one.
Okay, five from one customer.
Four memory customer, one memory customer, and one logic customer.
Okay. The one that was in this group of five was what you called a kind of quasi-VPA.
Sorry, I have to correct you. I think it was all logic. Sorry. Six were all logic, but with two customers, one had five, one had one.
On the 18 systems, we have said before that it's five customers in total. It's not like only the three.
Right.
Yeah.
It's five customers in total that have orders in with us.
Right. It's the one that had the five, which were the thing that you called a kind of a quasi-VPA.
Yep.
Yep. Okay. Thank you.
Ladies and gentlemen, we have the time for one last question. If you are unable to get through, as always, feel free to contact the investor relations department, and we'll get back to you as soon as we possibly can to try to help. Now, operator, if we could have the last caller, please.
Of course, sir. The final question comes from Robert Sanders. Please state your company name, followed by your question. Please go ahead, sir.
Yeah, good afternoon. Just a question about the 3400. The shipments that customers have ordered, are you going to upgrade the source to 250 watt at a later stage? Is that a free upgrade? Then the second question would just be on the HMI business. It does seem to be tracking below expectations from June, when you acquired it. I was just wondering what's the update there on the outlook and how that business is tracking. Thanks.
No, there's no major source upgrade. It's just cranking up the power. All that is not necessary. The source is the source, and we'll be capable of doing 250 watts. At least, it's going to be above 205 watts. Whether it's 250, it doesn't really matter because at 205, we're at 125 wafers per hour, and that's the throughput specification. 250, by the way, is also not, with this particular source design, is not our end target. I think with this particular design, we can go higher. We can go 300 watts and above. On HMI, below expectations?
Yeah. There is no significant difference in what we have seen already during our due diligence time, and since this is a growing business. More importantly, the roadmap going forward is well-aligned, not only within us and HMI, more importantly, also with the technology folks at the customers. We're looking forward to a significant opportunity, like we said it at our October call, which could be up to EUR 1 billion by 2020.
Yeah. Rob, you have to remember that in 2017, we still have the majority of the HMI sales are, you would call the standalone HMI sales. What we're really looking at is, you may remember the presentation that we did at the time of the acquisition, that the area where we believe we will have a significant growth opportunity is the combination of the holistic lithography or the computational lithography competencies of ASML with the HMI capabilities, creating a new product. That is where we think there is going to be a big market and a big growth opportunity. That's not for 2017. That will be 2018 onwards.
Thanks a lot.
Okay?
Yep.
Thank you.
Good. Well, thank you, everybody. On behalf of ASML's Board of Management, we'd like to thank you for joining us in the call today. Operator, if we could have your formal conclusion to the call, we'd appreciate it. Thank you.
Of course, sir. Ladies and gentlemen, this concludes the ASML 2016 fourth quarter and annual financial results conference call. Thank you for participating. You may now disconnect your line.