Ladies and gentlemen, thank you for standing by. Welcome to the ASML 2017 first quarter financial results conference call on April 19, 2017. Throughout today's introduction, all participants will be in a listen-only mode. After ASML's introduction, there will 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. If you would like to withdraw a question, please press star two at any time during the call. Your questions will be answered in the order that 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. Go ahead, please, sir.
Thank you, Peter. 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 CEO, Peter Wennink, and our CFO, Wolfgang Nickl. The subject of today's call is ASML's 2017 first-quarter results. The length of the call will be 60 minutes, and questions will be taken in the order that they are received. This call is also being broadcast live over the internet at www.asml.com. A transcript of management's opening remarks and a replay of the call will be available on our website shortly following the conclusion of the call. 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 meanings of the federal securities laws. These forward-looking statements involve material risks and uncertainties.
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, and 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 first-quarter results conference call. Before we begin the question-and-answer session, Wolfgang and I would like to provide an overview and some commentary on the recent quarter, as well as provide our view of the coming quarters. Wolfgang will start with a review of our first quarter financial performance, with some added comments on our short-term outlook. Then I will complete the introduction with some additional comments on the current business environment and on our future business outlook. Wolfgang, if you will.
Thank you, Peter, and welcome everyone. 2017 is off to a great start with a stronger than expected quarter. I would like to first highlight some of last quarter's financial accomplishments, and then finish with our view of the coming quarter. Turning to the Q1 results. Net sales came in at EUR 1.94 billion. Net system sales accounted for EUR 1.22 billion, nicely balanced between logic and memory. With the addition of HMI products, we are now including metrology and inspection equipment in the system sales, what was previously reported as service and field option revenue. This also means that metrology and inspection system orders are from now onwards included in our bookings and backlog numbers. This provides more visibility of our current and future system business in this product group.
Net service and field option sales for the quarter came in much stronger than expected, at a level of EUR 728 million, driven by mixed EUV and holistic lithography upgrades. As noted, year-to-date HMI system revenue are now reported in net system revenue. Otherwise, the service and field option revenue would have been even higher at approximately EUR 790 million. Our gross margin for the quarter came in at 47.6%, slightly higher than guided, driven by a higher top line and a favorable mix. Gross margin includes the amortization of the effects from the fair value assessment of HMI's inventory as of the closing date of the acquisition. Overall, OpEx came in as guided, although R&D expenses came in slightly lower at EUR 315 million, and SG&A expenses came in slightly higher at EUR 99 million. Moving on to the order book.
Q1 system bookings came in at EUR 1.9 billion, including orders for 3 3400 EUV systems from 2 customers. Strong bookings continued in the logic sector in support of the 10-nanometer ramps and in support of the EUV insertion at the 7-nanometer node. Memory bookings strengthened further from its strong Q4 level, supporting expected year-on-year growth in the memory sector in 2017. The continuing order flow for EUV systems increases our EUV backlog to 21 systems valued at EUR 2.3 billion. Our overall systems backlog now stands at EUR 4.5 billion. In addition, we also have 4 EUV upgrade orders valued at approximately EUR 200 million. This will bring these 4 NXE systems to NXE: 3400 performance. In total, we have 14 3300 and 3350 systems in the field, which are candidates for upgrades. As a reminder, system upgrades are not included in our system backlog. Turning to the balance sheet.
Order over quarter cash equivalents, and short-term investments came in at EUR 3.84 billion. As already mentioned in January, we saw a significant level of early payments from customers in Q4 of last year, which resulted in a negative free cash flow of EUR 212 million in Q1. As a reminder, in Q2, we have several extraordinary cash outflows, which will bring the overall cash balance back to our target level. Assuming approval at our AGM, we will pay a dividend of EUR 1.20 per ordinary share, or approximately EUR 515 million in total to shareholders. We also have a bond maturing in Q2 with an outstanding value of EUR 238 million. Lastly, we expect to close the acquisition of 24.9% of Carl Zeiss SMT during the quarter for EUR 1 billion.
Based on our current business view, we see a continued strong demand for DUV, holistic lithography, and EUV products throughout the year in both memory and logic. Our view is supported by our highest backlog ever. With that, I would like to turn to our expectations and guidance for 2017. We expect continuing sales trends in Q2 with total net sales between EUR 1.9 billion and EUR 2 billion, including an estimated EUR 200 million of EUV revenue. We plan to ship three NXE:3400s in the June quarter. Our EUV shipment plan for the full year includes 12 systems and is back-end loaded. We expect our Q2 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 installed base.
Gross margin for Q2 is expected to be between 43% and 44%, driven by the recognition of EUV system revenue. Excluding the EUV revenue, gross margin would be approximately at the same level as Q1. Q2 gross margin also continues to carry the effect from the purchase price allocation for the HMI acquisition. The negative impact of these purchase price allocation adjustments for Q2 is more than one percentage point. The impact for the full fiscal year is about EUR 90 million and will reduce to about EUR 40 million per year from 2018 onwards. R&D expenses for Q2 will be about EUR 315 million, and SG&A is expected to come in at about EUR 100 million. As a reminder, our share buyback program remains paused for the time being as we close our planned equity investment in Carl Zeiss SMT.
The remaining approval from China is expected in time to close the transaction in Q2 2017. Finally, as mentioned before, an increase of our annual dividend from EUR 1.05 to EUR 1.20 is submitted for approval at our annual general meeting of shareholders on April 26th. With that, I'd like to turn the call back over to you, Peter.
Thank you, Wolfgang. As Walter highlighted, our business continues to perform well. We started the year with a very strong quarter, and we expect this positive momentum to continue throughout the year. While Walter reviewed our current quarter performance and outlook for the coming quarter, I would like to provide some additional commentary on our markets and our longer-term outlook, as well as provide a few highlights on our product portfolio. As seen in our first quarter results, logic demand remains solid, and our memory demand continues to strengthen, with DRAM largely compensating for weak spending in 2016. Logic demand is driven by continued ramp of 10-nanometer, with memory demand driven by DRAM 1x-nanometer node and additions of 3D NAND capacity. The strength in shipments to China this quarter was driven by existing Chinese and non-Chinese customers.
As for new China business, we are in discussion with multiple Chinese logic and memory customers regarding timing of system demand for their new fab projects. We expect shipments to support pilot production in these new fabs starting in 2018. While it is still too early to provide quantitative guidance for 2017, our directional view, as expressed last quarter, remains largely unchanged. However, in terms of potential magnitude of our business, it now appears that memory demand will be up significantly as compared to prior year. On the ASML product side, let me start with an update on our EUV business. We started shipment of our NXE:3400 system, which will be the EUV workhorse in volume manufacturing over the coming years. Furthermore, we continue to make progress towards our 125 wafers per hour productivity and 90% availability commitment.
At the SPIE Advanced Lithography Conference in February, our customers presented their latest results confirming our progress on these metrics. The status of the EUV infrastructure was also presented by our customers, while there is still work to be done on things like pellicle, there appears to be no major roadblocks for EUV insertion in the time frames as indicated by our customers. Regarding demand, we took three EUV production orders from two different customers this quarter, bringing our total EUV backlog to 21 systems. As Walter mentioned, on top of this, we booked four orders for a total value of around EUR 200 million for upgrades of EUV systems currently in the field to NXE:3400 production specifications. By the way, these orders are field upgrades and do not show in our recorded order backlog.
EUV order flow continues while we work to finalize a PPA with at least one of our major customers, which will translate into additional orders over the next quarters. As customers continue to assess timing of their roadmaps and firm up layer adoption, we are beginning to get a clearer view of EUV demand for next year. The average analyst demand expectation stands at around 20 new systems shipped in 2018, which, given our current views, is reasonable, while we still have the option to build up to 24 systems next year. In deep UV lithography, demand for our TWINSCAN NXT980 immersion systems continues for both logic 10 nanometer and DRAM 1x-nanometer nodes, bringing the installed base to more than 60 systems. We are also seeing strong demand on our KrF platform, where we boosted the productivity of our XT860 system further to 250 waves per hour.
For our 3D NAND customers, we released new options that improve focus and alignment performance on the high topography layers typical for this application. To maximize capital efficiency, a number of customers also upgraded their immersion systems through significant enhancements to productivity, imaging, and overlay. These upgrades drove significant growth in our option business, which will continue to drive growth through 2017. In holistic lithography, we continue to ship our most advanced YieldStar 350 metrology systems to our customers supporting qualification and ramp up to 10 and seven nanometer logic node, as well as the 1x-nanometer DRAM node. In addition to YieldStar metrology systems, we are also shipping HMI eScan systems that are now reported as part of our systems revenue, as Wolfgang mentioned. The integration of HMI is progressing well, and customer interest in our pattern fidelity products remains high.
In summary, a great start to the year with a very solid quarter. Strong EUV demand, service and options business showing further growth momentum, and continued EUV order flow provides a clear indication that this technology has now become a part of our mainstream business. We expect the positive industry environment to continue resulting in a very good year for ASML. With that, we'd be happy to take your questions.
Thanks, Peter. Ladies and gentlemen, the operator will instruct you momentarily on the protocol for the Q&A session. Beforehand, I'd like, as I always do, to ask you to kindly limit yourself to one question with one short follow-up if necessary. This will allow us to get as many callers on as possible. Peter, could we have your final instructions and then the first question, please?
Yes, sir. Thank you. Ladies and gentlemen, at this time we will begin the question and answer session. Again, if you have a question, please press star one to register for a question and star two to withdraw a 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. The first question comes from Mr. C.J. Muse. Please state your company name, followed by your question.
C.J. Muse, Evercore. I thank you for taking my question. I guess first question, was hoping to get an update from you on your expectations for adoption of EUV in DRAM. I saw in the slide deck that you reiterated 1Y. Would love to hear how your discussions are progressing, and how you're seeing adoption, and what kind of layer count we should be assuming at that first part of the EUV adoption curve.
Okay. Yes, like we said, we are talking to logic and memory customers. On DRAM, the expectation is indeed, 1Y to be specific, it's mid-teens, 16, 15-nanometer DRAM. That will be the introduction node. That would involve one to two layers. You have to remember that, of course, the DRAM market is quite a significant market. One to two layers is a decent start for EUV technology at this node. We're talking at least to one customer very specifically. Other customers have shown similar interest, and they will follow suit.
Okay. Very helpful. I guess as my follow-up, Wolfgang, if you could give us an update on how you're seeing the trajectory for EUV gross margins. I guess within that, would love to hear how you see the cascade effect of 100% gross margin is as it comes through this year, then how we should see reaching that 20% target into the calendar 2018 timeframe. Thank you.
C.J., the story there is pretty much unchanged from what we communicated at our capital markets day in late last year. We were at minus 75% or so last year. Our objective is to get to 40% by 2020. We're targeting breakeven this year. There are several components to get this accomplished. One is simply volume. We're shipping three times as much as last year. As you can imagine, it's the same factory producing that will help. The second one was mix. Last year we shipped NXE:3350B. This year we're shipping NXE:3400. As you know, from prior calls, the list price on those is approximately EUR 20 million higher. The cost is not EUR 20 million higher, that will help.
Of course, we have the service business, where we've talked before that we're charging per wafer, which of course ultimately is a lucrative business model. As we are not churning out a lot of wafers right now, but have to man the systems, has a significant dilutive effect on gross margin. Last but not least, we're progressing on the learning curves. As always, in our own factory, the number of hours it takes to put one of these things together. Also with our suppliers as well as work that we have to do in the field to upgrade these systems and bring them to the latest level. I think the expectation of 20% is still good for next year as we again go from 12 systems to around 20 systems, as mentioned earlier.
We made some progress last year where we can now do partial revenue recognition upon shipment. It's partial revenue recognition because in case we have some performance criteria that will be met later, we got to defer some of the revenue. On the other hand side, it will help that some revenue that we deferred in the past will come in at no cost. We believe, in this year, the deferral and what's coming in is roughly offsetting each other. We said also in our communication that the revenue should be somewhere in the EUR 1 billion-EUR 1.2 billion range. I think we are on the right path to get to 40%, and volume is the biggest driver of this.
Peter, can we have the next call?
Okay. Yes. The next question comes from Mr. Farhan Ahmad. Please state your company name followed by your question.
Hi, this is Farhan Ahmad from Credit Suisse. Thanks for taking my question. My first question is on the orders that you have this quarter. The order level is pretty high and strong. I just wanted to ask how sustainable do you think the order trend is, and in particular, on the memory orders? Hello?
Yes. Go ahead.
Yes. In particular on the memory orders, the orders were significantly higher than any time in last 2 years. Can you just a bit more color on, is it driven by capacity additions or 1x-nanometer and any kind of visibility on your NAND or DRAM?
Yes. Just making a small note here, NAND or DRAM. Yes, yours in Q1. Yes, I think that's sustainable for the next few quarters, given the output that customers are giving us. Yes, it is significantly higher than the last 2 years. We have to remember that what we saw over the last 2 years was especially weaker order flow from DRAM-type customers, because the DRAM market was over the last 18 to 24 months not that strong. It had to do with the fact that in 2014, we ended beginning of 2015, M14 of Hynix and then Line 17 of Samsung came online, which were big fabs that we actually saw when you added them up, was quite a significant step up in capacity. We haven't seen any new fabs coming online of that size since.
What we did see, that has to do with your NAND, DRAM question. What we did see is that given the economics of the DRAM in our markets over the last 2 years, we've seen leading-edge litho capacity being used in NAND production. That is actually what happened, and I think we mentioned that on previous calls also. Bringing the wafer out capacity down with double-digit %. Now, that has created a situation clearly where demand and supply of DRAM end products was somewhat unbalanced over the last nine months, which led to increase in DRAM prices, which is not unsurprising that now customers are backfilling that capacity that they actually used to basically help produce NAND over the last couple of years. It also is an entry into your question on NAND versus DRAM.
It makes it all very opaque if you have NAND and DRAM sitting next to each other, depending on the market situation, leading-edge litho can be used for either or. That's why we look at memory now, and we said it also before as one segment. Looking at it, let's say from a demand point of view, what we currently believe, and also looking at last year's {inaudible} drop numbers, 30% in DRAM, 40%+ in NAND. Looking where the capacity situation is today, looking at the price situation, I can fully understand why memory customers are filling up their open spots in their fabs to make sure that they have sufficient capacity to fulfill the demand of their customers. Sorry for the very long answer, but you asked three questions, so that's why I had a long answer.
Thanks. If I can just squeeze a quick question. Last month, ASML signed a MOU with Shanghai Micro Electronics Equipment Company. I believe they are one of your new emerging competitors in China. I just wanted to understand what exactly is the terms of agreement. Is it something where you're just supplying the standard light sources, or is there more to it?
No, it's a different kind of cooperation. We're not competitors. They build lithography type machines for a different part of the market. They are largely in the packaging market. What we have done is effectively created a memorandum of understanding to start working together, whereby we actually use them as one of our suppliers in Not so much in the lithography market, but more in metrology systems area. We have a cooperation which is focused on making sure that they will get a better understanding of how you manage a modern, complex supply chain. It is not directly focused on areas where there could be a logical competition between the two companies. It's in different areas.
Thank you, Peter.
The next question comes from Mr. Sandeep Deshpande. Please state your company name, followed by your question.
Hi. My question is regarding your revenues for 2017. Peter, when we look at the consensus ahead of today, it's about approximately EUR 7.9 billion. Last year it's about EUR 7.6 billion. Wolfgang, you said that you would do revenues between EUR 1 billion to EUR 1.2 billion in EUV. If you took EUR 1.1 billion at the midpoint, it would mean that this year, the additional revenue outside EUV is about EUR one or EUR 100 million. With DRAM particularly looking much better, are you more positive than EUR 100 million incremental revenues from the memory market this year? I have a short follow-up.
I think Wolfgang can answer this.
I'll do the revenue piece. If you look at what we said last quarter and what we said this quarter, if I go exclusive EUV first, we had a very strong year in logic, and we continue to believe that's going to be a flattish year, which is good news. We have a memory business. I think we both used the word significantly up versus what we previously believed. Last quarter, we said right now think it's flat, but it could be up, and now we think it's significantly up. I think that's more than EUR 100 million, definitely. We have our EUV business, which was only EUR 350 million or so, which, like we said before, think it's going to be between EUR 1 billion and EUR 1.2 billion. Don't forget field options and services. Even with the adjustment that we made, we made EUR 2.1 billion last year.
If you add up the two quarters, the actual and the guidance, we're almost at EUR 1.4. We are at a significantly higher run rate there. Without giving a quantitative guidance, I'd say the EUR 7.9 is very much on the conservative side. I'll give you one other way to look at that may be helpful for you, Sandeep. If you look at the first half, the actual and the guidance, you deduct that we only have EUR 200 million in EUV revenue in there. You know in the second half, the bulk of the EUV revenue will come in. We've not indicated that our non-EUV business is going down. If you do these two exercises, I think you'll get the pretty good range that is quite a bit higher than EUR 7.9 billion.
Thanks, Wolfgang. Peter, just one quick follow-up on EUV. Market will be focused on the top three customers. What about the next tier of customers with EUV? Are you engaged with them, and when would they start placing significant orders? Thank you.
We think we have orders in our backlog of maybe more than the top three. We have an additional two, that makes it five, we are in discussion with two others, very close. I think that is now spreading, it's clear, of course, that the top three is then leading the pack in terms of speed of EUV introduction. The others are clearly following.
Thank you.
The next question comes from Mr. Pierre Ferragu. Please state your company name, followed by your question.
Hi, it's Pierre Ferragu from Bernstein. If I think about what I heard on the call is very positive developments regarding EUV. You have already 21 orders in the backlog. You have more in the next couple of quarters. Next year, like 20, 22 or so is probably what we're going to see. At the same time, I heard also that the DUV business is likely to do well in 2018 as well, mostly because of China generating a new area of demand in DUV. When I look at consensus expectations, basically, if I assume there is about EUR 1 billion of EUV revenues with about 20 tools, that implies the DUV business would be down about EUR 1 billion in 2018. Do you think that kind of pullback makes sense and would reflect maybe like a sustaining of the rate of the 10-nanometer node?
Do you think 2018 could be another very strong year in DUV?
Good question. Although what you're asking me is to just give you some reasonable financial feedback on what we think 2018 is going to look like, which I'm not going to do. Qualitatively, I don't see anything at this moment in time that would bring the DUV business significantly down. There is continued strength in the memory business. When I look at the strength of the logic business, leading-edge logic needs a performance memory, so it needs DRAM. That's clear. I don't see major new DRAM fabs coming online within the next 12 to 18 months. That means that all available slots in factories that can take leading-edge DUV will take leading-edge DUV, given the fact that I don't see that end demand will go down significantly. That means memory will stay strong. You mentioned China.
I think China longer term, and I think I also said it on the last call, leading to longer term, definitely is a big opportunity for the entire industry. However, in 2018, what we're seeing in terms of memory projects coming online will be focused on, let's say, finishing the construction and then putting the first pilot lines in, which will not drive a big addition to installed memory capacity. It will take time. I think all those memory projects are all new. You could call them greenfield, and not only are they greenfield fabs, they are greenfield companies. It's going to take a bit of time. The positive thing is that, of course, all those choices for technology and for later fab equipment will be made in the next 12 months timeframe.
It's going to be strategically very important to be in China and to actually sign up those Chinese customers. The rollout of that capacity will not be in 2018. In 2018 will be pilot production, and you see an acceleration of that rollout in 2019 and 2020. All in all, we do not foresee a DUV market in 2018. I think what we are currently seeing in the end markets simply does not support such a negative view.
Thanks. That's very clear. A very quick follow-up on China. How would you qualify demand you anticipate there? It's a 2019, 2020 story, okay? Is that mostly leading-edge tools? Are we going to look at high-end immersion tools mostly, or is it going to be more lagging edge in terms of demand in China?
I think you will see probably both, but I think the emphasis will be on the leading edge.
Okay.
Leading-edge logic and leading-edge memory. China currently has a lot of installed capacity, which is currently 45-nanometer and 28-nanometer, and also 28-nanometer fabs will still take tools in 2018. It's a bit of both, but I would say the emphasis will be on the leading edge.
Thank you.
The next question, Mr. Gareth Jenkins, please state your company name followed by your question.
Thanks. One question and a couple of follow-ups, Gareth Jenkins, UBS. The question I have is around the light source. I just wondered whether any of your customers are asking for Gigaphoton light source rather than your own {inaudible} solution on EUV? Just a couple of follow-ups, if I could, Peter, on other issues.
Well, if they would be interested in using EUV way into the next decade, they would be asking for a Gigaphoton, most of them aren't. They want it now, or they want it soon. It's the only available EUV source, and Gigaphoton is potentially something for the future, but I would say way into the next decade.
Okay, thanks. I just wanted to clarify, EUR 200 million of EUV revenue in Q2, EUR 800 million to EUR 1 billion in H2. The EUR 800 million to EUR 1 billion, that will come at zero gross margin? Is that the expectation or allowing for some catch-up, it will come at zero gross margin?
Yeah, for modeling purposes, I would not distinguish between the first half and the second half. It's making it too complicated.
Okay.
Next call, I think is next.
Yes. Okay. The next question is coming from Mr. Mehdi Hosseini. Please state your company name followed by your question.
Yes. Mehdi Hosseini, Susquehanna International. Pete, can you please give us an update on HMI? When would you expect the dual beam product to be available for evaluation? I have a follow-up?
Well, we will see the first integrated product, let's say the combination of the ASML product and competencies with HMI, shipped to the first pilot customers by the end of the year. The end of this year, beginning of next year. Also, what we call dual beam is what we refer to multi-beam. Well, our plan is to have that available relatively soon, but we first need to finish the development. It won't be this year. It will be 2018 at earliest. More important is, I think, that we have the whole concept of using the ASML holistic lithography capability with the inspection capability of HMI and combine it into one product. That will be available much sooner. I think that's the most important part. Then multi-beam is just a natural extension of this product.
Like I said, this combined product, the first shipment is targeted to be end of the year to the first pilot customers.
Sure. Then I have a follow-up on the EUV 12 system shipment this year, maybe as high as 24 next year. Would it be fair to assume that these systems are going to be for 7-nanometer, maybe insertion for a few layers, and then shipment in 2019 and on will include maybe one or two layer for DRAM, and as you look into 5-nanometer logic, the number of critical layers that would use EUV would go higher, and that's how we go to doubling playing out. 12 systems this year, 24 next year, and 7-nanometer introduction, and as we go into 2019 and 2020, the doubling is driven by DRAM and higher critical layer as you migrate to 5-nanometer logic. Is that fair?
Well, I think it's a little bit more complex than that because 7-nanometer is not ending in 2018. I think as you will see the production ramp for our customers in 2018, 2019, and it will continue into 2020, which will be seeing layers. These node layers will be on top of each other, like we discussed earlier on these calls. What we will be seeing is that you almost see an acceleration of a capacity ramp for a new node by the leading customers. It's almost like a camelback. It's like a hump. Then you have a very long tail. Now, that long tail on 7-nanometer will be extending into 2020 and 2021. It's not by the one or two leaders in logic, but there are also followers.
We'll move into that 7-nanometer node, and they will start using EUV in that timeframe. We will see layer upon layer. This is not the nice clean-cut end of the 7-nanometer node followed by a 5-nanometer node. It's going to be layered. On memory, you said 2019, DRAM one or two layers. Yeah, in production, I think we'll see probably earlier adoption of one or two layers in DRAM by 2019. This will also be some of the systems that you referred to, the 12 systems in 2017 and the 20 potential 24 systems in 2018 will be used for DRAM, one or two-layer production on this mid-teens node.
Is that why they haven't finalized the options part of the package? What they have booked so far is just for the equipment and actions are finalized later?
This is specifically to one customer, where we have the tool price is fixed, the options is really the commercial negotiations of what do you want in terms of options, which is specific to the production process and to the, you could say, the design of the customer specifically. Those are options that are add-ons that they need in their production process, which, of course, for us is good business and for the customer, always an area where they think they can have some purchasing advantages. That's just a matter of time. I don't think we can draw any firm conclusions other than that.
Got it. Thank you so much.
The next question, Mr. Andrew Gardiner, please state your company name, followed by your question.
Good afternoon. Thank you. It's Andrew Gardiner from Barclays. I think I was just interested in following up a bit more on some of those comments you were making on the EUV decision-making process. You highlighted during your prepared comments the public comments from all three of your lead customers acknowledging the progress that's been made on EUV, not only by yourselves, but also across the ecosystem that was fairly well publicized at SPIE. Yet, we're still not seeing all three step up sort of equally in terms of orders. I think TSMC has been pretty clear and consistent with what they're intending to do over the last couple of quarters. We can see some of that in your backlog.
Samsung also seems to have been clear in their statements, yet, as you just highlighted, we don't yet have the Volume Purchase Agreement or indeed the kind of volume orders that you'd highlighted or that you had expected in the first quarter. Finally, Intel, they were the first to sign a VPA two years ago now, yet they still haven't really acted on it, and they had publicly said they would do so when the technology was ready. Again, publicly, they seem to admit that it is, yet they haven't stepped up in terms of their order rates.
Given what you've described in terms of the 2019 ramps, I'm just wondering why you think we're seeing such sort of different positioning from these lead customers and the closing order, the window later this year, where's your confidence level that the two who are lagging are going to step up over the next, say, two quarters?
Well, to be honest, I don't think customers are really lagging, one. They have different areas of focus, they have different roadmaps and timing of roadmaps which they would like to execute, also the size at which they want to do it is also different. This is not a homogeneous pack. This is a heterogeneous, three different customers, three different roadmaps, three different ways in which they want to execute this. I think you're absolutely correct when you refer to the public statements, which are very convincing. Not only the public statements, as you can imagine, we are in very close contact with them, and we are also seeing what they are doing in preparation of that EUV ramp. That's real tangible. Those are factories, those are EUV pedestals.
That's EUV infrastructure that's being built and being put into place, into those fabs. That is also tangible, which is not that visible, but to us it is. On the negotiation process, that's also different per customer. For instance, one of those customers doesn't have a Volume Purchase Agreement yet, they have orders in our backlog, they order tools without a VPA. The Volume Purchase Agreement is really to determine, in their best interest, what the pricing is on a certain volume. That is a commercial negotiation. Which, as Mehdi asked you in the previous question, is also a matter of how many options do you want, and what do you want to pay for those options? These things are commercial negotiations, which are not fully detached from the planning of what they want to do, it is a different process.
You have the planning process for the production, which we're pretty close to, which actually drives our own production planning. You have the commercial process. That commercial process ends when it ends, and this is where we are. I think in summary, three different customers, not homogeneous in the way they look at their roadmaps, the timing of their roadmaps, not homogeneous in the way that they negotiate, and they're not homogeneous also in the speed at which they want to do something and the speed of their ramp. This is what we have to take into consideration. Again, I try to put some color on what we said before, and what we said before is that we see this ramp of EUV very clearly coming on the 7-nanometer, which we make maybe 12 units this year.
We should now say around 20 units next year, and we could see a further doubling in the years thereafter when we look at the customer roadmaps and their execution planning. I think that is the most important to mention right now.
Yeah. Thank you for that intel. That's very helpful. Also, just a quick follow-up for you, Wolfgang. The guidance that you Sorry, the 1Q installed base revenue and the guidance you just gave for 2Q, clearly very rapid growth there. Are you still saying that that sort of revised installed base revenue of about EUR 2.1 billion last year, is that still an 8%-10% revenue growth target for 2017? Are you tracking significantly ahead of that at the moment, given the one-half guidance as you've already mentioned? What kind of growth you would be thinking there?
Yeah. It certainly looks like. We said in average, we think this business is going to grow 10% over the next couple of years. As customers are optimizing their installed base, they're using some of these performance-enhancing options. It's very capital efficient for them. You'd be right. I mean, EUR 7.8 billion plus EUR 650 million, I don't see this deteriorating much in the second half. You can do the math. It will very likely be above 10% this year.
Okay. Got it. Thank you.
The next question is Timothy Arcuri. Please state your company name followed by your question.
Thank you very much. I guess I have two. First of all, Wolfgang, I wanted to ask you about the 2018 EUV comments.
Yes.
We have gotten some real bad feedback from your line. Is there anything you can do your end?
You have a horrible cold. It's one of the two.
Better?
That's better.
Try again. Okay, great. I wanted to ask about the 2018 EUV comments. I know you still have 24 slots. That has not changed. I think last call, you said that you'd be totally full for those slots and even have some backlog by the end of the year for 2019 shipments. I guess I'm wondering why you'd only ship 20 systems next year versus 24. It would seem like you have to ship as You're totally full on those slots. Why wouldn't you ship all those slots versus only 20? I'm just getting some questions from investors that it seems like a downtick, and I just wanted you to address that. Thanks.
Yeah. I think when you look at the 20 units, you have to remember that we also booked four upgrade orders. There are customers that are taking those tools, those 3350s and 3300s out of the R&D. They want them upgraded and put them into production. That adds another four. We always said, when we look at our production capacity, and we look at the roadmap and the timing of the roadmaps that our customers are currently talking about in terms of their roadmap ramps, and look at that timing, said, "We will have to use that 24 capacity." Well, four of them, at least, are now being upgraded. That's basically taken out of R&D and then put into production, which is probably an efficient use of their capital.
It also means that we need to allocate some of our production people to fab, to upgrade in the field, not in our factory, but in the field. Which also has an impact on our own capability, because you actually have people in the field, competent people, production people that are in the field doing these open heart surgeries. It is a mix now of, I want to say, new systems and customers wanting those upgrades, because it's capital efficient.
Got it, Peter. Thank you. That's helpful. Very helpful. I guess my follow-up was just on how to think about China timing. I know you said that you expect shipments to support pilot production, I think you said in 2018. For the existing China projects, does that mean that you'll ship tools in the second half of this year for pilot production for them next year? Or does that mean that you won't ship tools to these guys until next year? Thanks.
It's been a mixed bag. I think the first tools will ship towards the end of this year, and then we'll continue shipping in the first half of 2018. Before that's all installed and the production process is qualified, you won't see any output out of these pilot fabs, I would say, way in the second half of 2018.
Great, Peter. Thank you so much.
The next question, Mr. Amit Harchandani, please state your company name, followed by your question.
Thank you. Amit Harchandani from Citigroup. Good afternoon, gentlemen. I just wanted to come back to the topic of EUV shipments, orders, and lead times. I was hoping you could help me with some math here. You've got 21 tools in your backlog at the end of Q1 quarter. My understanding is the lead times are, including the supply chain, about 18 months. If you're looking at shipments cumulatively in 2017 for the rest of the year and what you said for 2018, it's about 30-32 tools to be shipped. Does that mean you need around 9-10 orders coming through by the end of June? If you complete the orders coming later and your lead times get shorter, allowing you to still meet the shipment target by the end of 2018? Thank you.
First of all, I think you'll see, independent of any VPAs, you'll see a decent order flow in the next quarter or the quarter we just started. It's not a hard and fast rule, right? Where you have 18 months. It's not just the order that's the communication between us and the customer. We're sitting together with these customers on a weekly basis. Just because we don't have a piece of paper that says VPA, because some final detail on a term and condition is not done, means that we don't have a very good view on what the customers need in which quarter in the next year. It comes down a little bit to trust as well, right? Where you trust that the forecast is eventually translating into an order.
Of course, in the long run, you've got to have the discipline to have the orders coming in, because we can certainly not be the inventory holder of the industry. I think we're in pretty good shape. You'll see orders into with pretty good forecasts on it.
Yeah. I think, in addition to what Wolfgang said on it. I said in earlier calls, it is our focus to have the orders needed for shipment in 2018 in the backlog by the end of this year. It is exactly what Wolfgang said. It is not a piece of paper that drives us to start ordering lenses and the long-lead-time items for the EUV source. It is really the weekly connection that we have with our customers and the almost weekly, or you could say, monthly update that we're getting on their EUV planning. That is driving this. We will be looking at getting in 2017 all the orders that we need for 2018.
Thank you. Just maybe as a follow-up on the lead times, can you give us a sense of how we should think about the lead times contracting as we look out over the next two, three years? What are the key parameters that would help you get them down besides, of course, experiential learning?
Yeah, I think it is experiential learning. It is the learning curve. We will be looking at halving the lead time over the next three years. Like I said, till the end of the decade. By 2020, we should have half the lead time. Which is, by the way, this is what we did with DUV. With DUV, as a matter of fact, we're reducing the factory cycle time in our factory and the overall lead time of a DUV system still. It's a continuous process. But it is continuous learning. It's almost exponential learning. That's what it is that will drive the lead time down.
Thank you.
The next question is coming from Mr. Douglas Smith. Please state your company name, followed by your question.
Hi, it's from Deutsche Bank. The question I had actually is on High-NA. I saw on one of the slides at SPIE a diagram of what High-NA looks like. It looks like a much larger machine than the current EUV system, and the specs are much higher and so on. Have you already thought what the price of such a machine is? Are we looking at like a EUR 200 million kind of number for High-NA?
Yes. EUR 200 million+.
Okay. Maybe a little bit in the same direction. Do you have people who are willing to do double patterning for EUV, or is that something they want to avoid and would prefer to use High-NA?
Yeah, I think this is a very good question. I think this exactly goes down to the economics of high-end manufacturing. It is all about the cost per leading-edge transistor. Now the question is, can you do that through multiple patterning or double patterning EUV, which would cut the productivity in half? Would you do that with basically a shrink capability with higher productivity to a High-NA tool? The economics that we have calculated on the High-NA tool is clearly preferring a High-NA solution instead of double patterning EUV when you go to the 3-nanometer node. As far as we're concerned, looking at the specs, High-NA is the preferred economical solution.
Right. Just to clarify Mehdi's question, more or less an EUV was 2X the price of DUV. Do you expect High-NA to be about 2X the price of today's EUV?
That's a reasonable assumption.
Yeah. Okay. Great.
The next question is coming from Mr. Adithya Metuku. Please state your company name, followed by your question.
Yeah. It's Adithya from Merrill Lynch. Thanks, guys, for taking my question. I have a quick follow-up. My questions have largely been answered. Well, can you give us some color on the total amount of deferred revenue that you have on the balance sheet related to EUV, that you haven't still recognized in the P&L? Thank you.
I want to say, I have the overall deferred revenue is about EUR 1.2 billion. I'd say EUV is probably somewhere in the EUR 200 million range or so.
Okay. Thank you.
The next question is coming from Mr. Robert Sanders. Please state your company name, followed by your question.
Yeah. Hi, good afternoon. It's Robert Sanders. First question would just be on DRAM. As I understand it, Samsung is exploring DRAM insertion for EUV for performance DRAM only, that you are deferring the decision on commodity DRAM. Is that something you recognize? I'm just wondering, is that what you're factoring into your forecast?
Well, we don't know the exact application of what type of product is going to be used. The only thing that we know it's on this particular mid-teens node, and whether that's used for commodity DRAM or for performance DRAM. I suppose it's going to be performance DRAM, that seems the most logical. I cannot give you a definitive answer on this because you really have to ask Samsung and obviously that you did. We have to believe what our customers say. Probably.
Got it. I'm only asking because performance DRAM is about a third of the capacity in DRAM. The second question would just be on TSMC and other customers. I think TSMC is the only customer that's actually moved EUV tools into the manufacturing line. I was just wondering when you thought Samsung and Intel would have to move tools into the actual line to meet insertion at 7-nanometer. Thanks a lot.
Yeah. I think this is really dependent on how they plan the introduction and the ramp of their nodes. It is really, it's a question that we can't answer for customers. If you ask them and then they give you a timing and one timing looks more aggressive than the other timing, then you can probably draw a conclusion on what they believe the need of their customers is. It's not up to us to draw any conclusions there. It's really driven by how they look at their own roadmap and their customer demand that will determine when they start to ramp up and put tools into production and start ramping capacity.
Thanks a lot.
Yeah. Thanks, Robert. Ladies and gentlemen, we have time for one last question. We're going to squeeze one more. If you are unable to get through on the call and still have questions, feel free to reach out to the IR department. We'll be around for a while. Peter, if we can have the last caller, I'd appreciate it.
Of course, sir. The last question is coming from Mr. François Meunier . Please state your company name, followed by your question.
Thanks for taking the last question. I'm not going to ask why 20 or 21 or 22 next year, because if we roll back in time, probably no one would have believed you would have even shipped more than 10 next year. The question really is about the gross margin, 47.6% with no EUV revenues this year. If my calculations are correct, it's something like 500 basis points higher than last year or something. Can you help us to understand what the bridge is between last year and this year? Is it really around the mix effect with more services revenues and more options and stuff like that? Or is it also because even at the tool level, your margins are going higher?
I think there, Francois, there's two elements. First of all, our revenue in the first quarter was 46% higher this Q1 when compared to last Q1. We had quite a bit of volume effect there. Plus, of course, our mix is developing, right? You see more holistic lithography as a % of revenue. You see HMI with the full quarter in there. You see indeed, as you mentioned, you see a lot of upgrade options that have very, very high margins. It's a combination of all of the above.
Thanks, guys. Well done.
Thanks.
Thank you, Francois. Now on behalf of ASML's Board of Management, I'd like to thank you all for joining us today. Operator, if you could formally conclude the call, I'd appreciate it. Thank you.
You're welcome, sir. Ladies and gentlemen, this concludes the ASML 2017 First Quarter Financial Results Conference Call. Thank you for participating. You may now disconnect.