艾司摩爾 (ASML) 2021 Q1 法說會逐字稿

Thank you for standing by. Welcome to the ASML 2021 First Quarter Financial Results Conference Call on April 21, 2021. (Operator Instructions) I would now like to turn the conference call over to Mr. Skip Miller. Please go ahead, sir.

Thank you, operator. Welcome, everyone. This is Skip Miller, Vice President of Investor Relations at ASML. Joining me today on the call is ASML's CEO, Peter Wennink; and our CFO, Roger Dassen.

The subject of today's call is ASML's 2021 First Quarter Results. The length of this 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 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 this 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 meaning 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 presentations 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, Skip. Welcome, everyone. Thank you for joining us for our Q1 2021 results conference call. And I hope all of you and your families are healthy and safe.

Before we begin the question-and-answer session, Roger and I would like to provide an overview and some commentary on the first quarter as well as provide our view of the coming quarters. Roger will start with a review of our Q1 2021 financial performance, with added comments on our short-term outlook. And I will complete the introduction with some additional comments on the current business environment and our future business outlook. Roger?

Thank you, Peter. Welcome, everyone. I will first review the first quarter financial accomplishments and then provide guidance on the second quarter of 2021.

Net sales came in above guidance at EUR 4.4 billion, primarily due to higher Installed Base business from upgrades. We shipped 9 EUV systems and recognized EUR 1.1 billion revenue from 7 systems this quarter. Due to the delay in one of our customers' road maps, we jointly decided to buy back 2 of their new systems and shipped these to another customer this year. This was accounted for as a revenue reversal in Q1 of 2021.

For the systems shipped in Q4 2020 with a new configuration, we were able to complete site acceptance and recognized revenue this quarter. We also shipped 1 system this quarter without factory acceptance testing, so revenue will be recognized in the subsequent quarter after a customer site acceptance. Again, the net result is 7 EUV revenue systems in Q1.

Net system sales of EUR 3.1 billion was again more weighted towards Logic at 78%, with the remaining 22% from Memory. The strength in Logic drives both deep UV and EUV revenue. The Memory business is mainly driven by DRAM. Installed Base Management sales for the quarter came in at EUR 1.2 billion above guidance due to increased upgrade business as customers pull forward software upgrades that can quickly increase productivity of systems in this high-semiconductor-demand environment.

Gross margin for the quarter was 53.9% and was above guidance due primarily to the additional software upgrade business. On operating expenses, R&D expenses came in at EUR 623 million and SG&A expenses at EUR 168 million, which was slightly above our guidance. Net income in Q1 was EUR 1.3 billion, representing 30.5% of net sales and resulting in an EPS of EUR 3.21.

Turning to the balance sheet. We ended first quarter with cash, cash equivalents and short-term investments at a level of EUR 4.7 billion. Moving to the order book. Q1 net system bookings came in at EUR 4.7 billion, including EUR 2.3 billion for EUV systems and another strong quarter of deep UV demand. Order intake was largely driven by Logic with 76% of bookings, primarily due to EUV order intake, with Memory accounting for the remaining 24%.

With that, I would like to turn to our expectations for the second quarter of 2021. We expect Q2 total net sales to be between EUR 4 billion and EUR 4.1 billion. The directionally lower guidance is primarily due to shipments in the quarter, both EUV and deep UV, that will not receive factory acceptance testing due to customers' desire to bring systems into production as quickly as possible. Therefore, we will recognize revenue in subsequent quarters after completion of acceptance testing at customer sites.

In addition, the Installed Base business is expected to be lower in Q2 versus Q1 as customers pulled forward installation of productivity software upgrades to quickly increase wafer capacity. We expect our Q2 Installed Base Management sales to be around EUR 900 million.

Gross margin for Q2 is expected to be around 49%. The lower gross margin quarter-on-quarter is mainly due to delayed revenue from immersion systems that we plan to ship without factory acceptance testing as well as lower Installed Base Management sales versus Q1.

Expected R&D expenses for Q2 are EUR 650 million. And SG&A is expected to come in at EUR 175 million, reflecting a continued investment in the future growth of the company. In support of our aggressive product road maps and opportunity to fill in some of our high-value product developments, we plan to increase our R&D investments, primarily in EUV, via increased development capacity.

Furthermore, this increase will allow us to compensate for remote work impacts. We don't expect this increase to scale at the same level as our revised revenue increase, with R&D expenses for 2021 around 14% to 15% of sales. We expect SG&A to remain around 4% of sales for 2021. Our estimated 2021 annualized effective tax rate is expected to be between 14% and 15%.

As mentioned last quarter, ASML intends to declare a total dividend with respect to 2020 of EUR 2.75 per ordinary share. This is a 15% increase compared to the 2019 dividend. Recognizing the interim dividend of EUR 1.20 per ordinary share paid in November 2020, this leads to a final dividend proposal to the general meeting of EUR 1.55 per ordinary share. The 2021 Annual General Meeting of Shareholders will take place on April 29, 2021 in Veldhoven.

In Q1 2021, ASML purchased 3.5 million shares under the 2020 through 2022 program for a total amount over EUR 1.6 billion. Our expected free cash flow generation enables the opportunity for continuation of significant share buybacks in the coming quarters, and we expect to complete the execution of our current share buyback program early.

With that, I would like to turn the call back over to Peter.

Thank you, Roger. As Roger highlighted, we had a very strong quarter in both sales and profitability, driven by continued strength in both Logic and Memory, as well as significant demand for upgrades, as customers look to bring additional capacity online as quickly as possible. The additional upgrades consisted primarily of software-based productivity packages.

We are seeing a significant increase in demand from our customers across all market segments on all nodes, mature and advanced, compared to 3 months ago. And we expect another very strong year with demand across our entire product portfolio.

The steeper-than-expected recovery in demand for semiconductors, amplified by the COVID, induced lower investments of the industry in 2020, thus creating significant upside to demand over the past quarter. This more cyclical demand sits on top of the secular growth from the accelerated buildup of the worldwide digital infrastructure and assuring demand not only for advanced and mature logic nodes, but also for memory.

In Logic, customers continue to see strong demand across the broad application space for both advanced node as well as mature nodes. And last quarter, we expected revenue from Logic in 2021 to be up 10% year-on-year. However, we now expect Logic to be up around 30% this year.

In Memory, the applications that are driving the strong Logic demand are also fueling demand for Memory. As we mentioned in earlier calls, the Memory recovery started last year and continues to strengthen as customer plans to increase capacity is driving significant demands for our systems in the second half of the year. Compared to last quarter, where we expected revenue from Memory in 2021 to be up 20% year-on-year, we now expect Memory revenue to be up around 50% this year.

On our Installed Base business, service revenue will continue to scale with the growing installed base and with increasing contribution from EUV services as these systems ramp more and more wafers in volume production. We're also supporting our customers with upgrades to maximize performance of their installed base.

In order to meet the high demand in the current tight chip supply environment, customers are prioritizing software upgrades to quickly increase capacity, as reflected by our hardware upgrade of Memory in Q1. And some hardware upgrades require extended machine time to be installed. And in the current high-demand environment, customers will be less willing to take systems down, which has a dampening impact on the 2021 growth profile of hardware upgrades. We therefore still expect growth on our Installed Base business of around 10% this year, as mentioned last quarter.

On EUV, we continue to see increasing customer confidence in this technology, which is translating to expanding layer counts in Logic and increasing deployment of EUV in Memory at multiple customers, evidenced by a number of customer announcements around increases in their CapEx plans, which will include spending on EUV for advanced nodes.

To support this strong EUV demand, we are working to increase our output capability. At the same time, we are driving our product road map to produce higher productivity machines, which will increase the effective EUV capacity per system and the wafer output capacity of our customers. We plan to transition to the NXE:3600D system in the second half of the year, which will provide customers with a 15% to 20% higher productivity compared to the NXE:3400C systems shipping in the first half of the year.

Limited by the available modules and parts this year, we're still planning for growth of around 30% in EUV revenue this year. With the expanding adoption of EUV at our customers, we see increased demand building in 2022 and beyond. We are improving our manufacturing cycle time and are planning our supply chain for a capacity of around 55 systems next year.

And as a reminder, all of our planned shipments in 2022 will be NXE:3600D systems, with the increased productivity capability. Our strength and outlook on the year relative to last quarter is primarily driven by the demand for deep UV systems.

With increased demand on leading-edge nodes as well as mature nodes running longer and ramping stronger, demand for our immersion and dry systems is stronger than ever. We have put in place plans to increase our deep UV capacity to help meet our customers' increased demands.

In our application business, as demand for scanners continues to increase, we expect a step-up in demand for our YieldStar metrology systems, particularly in Logic. The newly released YieldStar 385 is beginning to ramp across our customer base as well. And with the recovery in Memory, specifically in 3D NAND, we expect a substantial increase in e-beam inspection revenue this year.

For the industry at a high level, we see 3 trends driving considerable growth this year and in the years to come. The first trend, in the shorter term, that's more cyclical or, you could say, a catch-up-driven demand from decisions made in 2022 due to the global pandemic. These shortages were initially evident in the automotive market, but more recently, there are also indications of supply tightness impacting other market segments. We expect this to drive considerable demand for lithography systems this year and into next year.

The second is a secular growth trend, driven by the digital transformation taking place as we become a more connected world across both people and machines. And this transformation was further accelerated over the past year with the increased remote activity and reliance on technology to stay connected.

These secular trends are driven by expanding end-market applications, such as 5G, AI and high-performance computing. These and other mobile distributed applications drive demand for both advanced logic as well as more mature technology required for the services and applications that drive the growth of the digital infrastructure. And along with increased logic demand comes increased memory demand. This, in turn, drives demand across our entire product portfolio.

And the third trend, which we're starting to see now and which we will likely continue to see longer-term, is the desire for more technology sovereignty, which includes semiconductor and silicon-based technology, leading to a geographical decoupling, as different governments put initiatives in place to localize supply chains and become more self-sufficient. This, inevitably, will create some level of inefficiency in the semiconductor supply chain and creates additional equipment demand as more fabs are strategically built across the globe.

If we summarize the growth of the different segments and the trends just discussed, we now expect sales growth towards 30% this year. To achieve this growth, we are ramping up our capacity to support customer demand, resulting in a stronger second half.

With the higher revenue and increased mix of deep UV and upgrades, we now expect gross margins to be between 51% and 52% this year. With the industry as a whole, the long-term demand drivers only increase our confidence in our future growth outlook towards 2025, and we plan to provide an update on our 2025 scenarios at our Investor Day in September.

And with that, we'd be happy to take your questions.

All right. Thank you, Peter and Roger. The operator will instruct you momentarily on the protocol for the Q&A session. Beforehand, I would like to ask you kindly limit yourself to 1 question with 1 short follow-up, if necessary. This will allow us to get to as many callers as possible. Now operator, could we have your final instructions and then the first question, please?

(Operator Instructions) And our first question comes from the line of Mehdi Hosseini of SIG.

Yes. I want to follow-up on the EUV revenue target for this year. I understand that you're keeping it unchanged at 30%. But in case you're able to improve the supply chain and availability of subcomponent, could there be upside to the shipment, even though you may not be able to recognize the revenue? And I have a follow-up.

Yes, Mehdi. You just used 2 very important words, in case, which is the problem because we said it last quarter, these are very long-lead-time items. I mean, as a lens, as a manufacturing cycle time or, yes, I would say, lead them over 12 months. So if we have an audit, the lens is not going to be there.

So I think when we look at capacity increase, we really, really need to look at next year. I think it's virtually impossible to get more out of 2021 as we have planned today. So really, it's going to be -- that's why we also mentioned that working with our supply chain and looking at the demand next year, we are now planning, and we haven't received full confirmation yet of the supply chain on it. But our focus is on the 55 systems next year.

Bearing in mind also that, I think we said it before, customers are buying systems, but effectively, they're buying wafer capacity. And with next year, only 3600Ds, which already provides you with a 50% to 20% productivity increase. Well, if you then multiply, let's take the average of 15% to 20%, 17.5% or, let's say, 18% extra productivity increase for your system, then the 55 systems actually translate into 65 systems, with 3400C productivity. So it's -- this is the way that you need to look at it.

Great. Thanks for the detail. And a quick follow-up. As I think about beyond 2021 and your operating margin, could there be a scenario where your key customers would share some of the development costs associated with High-NA, similar to what happened to EUV almost a decade ago, so that, that could also help drive overall operating margin to above 33%?

Well, I think -- I don't think it's likely because it's not necessary. I mean when we go back to almost a decade ago, our R&D expense at that time was EUR 600 million per year. And then when you looked at the size of the EUV program at that time and still you need to consider that the entire Low-NA EUV program will have cost more if we don't end it than the High-NA program.

So if you have a look at it then, the relative effort to bring EUV to life was so large that it would have put a significant financial strain on the P&L of ASML which, of course, we could not afford, which is not the case today. And I think, if you think about operating margins, it is more a matter of maturity of EUV, which will, of course, lead to better-performing tools with higher productivity, with higher value. EUV service revenue, that will grow going forward. And I think those are the most important drivers for operating margin. Roger?

Yes, I agree. I mean the gross margin components, I think, Peter, you just referenced them why -- it's the EUV ASP. And as a result of that, the EUV systems gross margin and the EUV service gross margin, those are the major drivers on the gross margin side.

And I think, also, in the introduction, we've been pretty clear on what the ambition is for OpEx, right? So 14% to 15% for R&D and 4% for SG&A. And our longer-term ambition, at least, on the R&D side is to try and model that back to around 13% over time.

I think that's what we stated, this R&D, yes?

That's in R&D. And those are the major drivers, I would say, to get the operating margin further up.

Yes.

And our next question comes from the line of Pierre Ferragu of New Street Research.

Can you hear me well?

Very well.

Great. So thanks for the color on how you increase capacity for this year and next year. That's very clear. My question would be more over 2, 3 years. If I recall correctly, like about my visit in winter in Veldhoven, you guys have set out to be able to do 60 EUV tools per year then 20 High-NA EUV tools. And when I took a shot at looking at how much would be needed over the next 3, 4 years, I felt like this was actually a bit limited and that you might actually need more capacity.

And so I would love to hear your take on that. In the longer term, do you think it will make sense to think about increasing your overall capacity? And if that's the case, what kind of lead times are we talking that for that kind of endeavor? And I have a quick follow-up.

Yes. It's a very good question, Pierre. And I think it's a long-term strategic question. Let me try to answer it this way. When I talked about the 3 trends that the company will have to face in terms of demand cycles, that's, of course, the shorter term, which, you could say, it's a catch-up of the lower level of investments we saw last year because of the pandemic, which is -- which are catching up this year and I think throughout next year or, I think, as a part of next year.

Then as the secular growth turn, which we have underestimated, I think you can go back to the Capital Markets Day in 2016 and '18, I don't think we, at that moment in time, anticipated the strength of the deep UV market that we see today. And that's the third trend, which is basically the driver of technological sovereignty by different countries and different governments.

Those are all -- at this moment in time, those will be the drivers. But it's also difficult to really get a very clear view on this. So what your question is, taking that into consideration, should you guys be looking at increasing your capacity beyond the 60, the magical 60, that we always mentioned? Good question. And I think this is actually the work that we're doing today, yes?

We're looking at this, and why do we need to do this? So now the lead time -- now if you have to build a complete new factory, let's say you need to build a new optics factory, then the lead time is 2 to 3 years, yes? Because you need to build the factory, you need to procure the machines or then -- by the way, when it comes to optics manufacturing, you have to build those machines yourself because they're not available. So that will be longer.

But a lot of ways to increase capacity and that's through cycle time reduction, that's through process optimization. And so it's this complexity of measures that we can take to see how we can drive the total capacity up, which, by the way, all have lead times that are beyond the 55 units that we're planning for next year, which doesn't mean that it couldn't be higher in '23 and '24. But that's the work that we need to do today, which, by the way, we are doing, yes?

That makes sense. And a quick follow-up is very much related to that. You mentioned like the geopolitical situation. When I look at the overall value chain and how much it depends on what's happening in Veldhoven, on the single site where you assemble your tools, which is almost scary, it's a weak point for the global value chain and probably a place where your clients and governments are getting maybe a bit nervous about that. So do you have conversations with people around you about diversifying your size and creating supply that would not only be -- like, would be less dependent on the Netherlands?

Yes. I think the -- that's the -- the situation of today is not very different than the situation 5 years ago. You could also -- you can only argue that the realization of people is now different, but has never changed. It's what it is.

But the same can be said for our own customers that are basically concentrated in one particular area, or some of our peers that -- with some of their production facilities only make a couple of hundred tools per year in 1 or 2 single sites. It is the concentration of the semiconductor industry in different geographical areas that actually now starts to make governments think it has never been an issue, yes? It's only becoming an issue when this -- an almost seamless ecosystem that has been built across many, many borders are now being -- that ecosystem is now being threatened, I would say, almost by blockages of that seamlessness, yes? And then you get an issue, yes?

So -- but I think it's nothing different. It's always been this way, and especially over the last 5 years, I would say. There's high concentration of leading-edge technology across the value chain. But given the geopolitical situation, people are more aware of, it and they start now thinking of self -- of levels of self-sufficiency that, a couple of years ago, nobody thought of.

And our next question comes from the line of Joe Quatrochi of Wells Fargo.

Was wondering if you could help us on the updated 2021 guidance. And I think, last quarter, you had talked about there being potential upside from domestic China as we hadn't seen any change in the geopolitical situation or export controls. So just curious, can you help us understand, is that now included -- that upside included? And if so, how much is that?

Yes. I think it's fair to say that, that upside is now included in the numbers that we've given. So it's clear, and Peter went through that also in the presentation at the beginning of this call, that we've revised upwards, if you like, the outlook for the full year. And further down into the year, we've now said this is how we look at the full year, of course, still expecting that the regulatory situation remains a little bit the way it is today. And therefore, what previously was upside -- and the way we talked about is this upside has now really been included in that number.

Remember, last time, the expectation that we gave was that could be about EUR 600 million upside coming from China. And that number is now included in the, let's say, close to EUR 18 billion that we've now talked about.

Yes. And if you then really want to think about the risk profile, you could say, well, when the regulatory situation changes and you could not ship to China, when you look at the market situation today, then -- which is quite different than 3 months ago, then we would ship those systems somewhere else.

Great. And then just as a follow-up, I just want to make sure I understand. The 2 systems that you repurchased from a customer, those were on the EUV side, and then you plan to ship those this year. And I guess, is that -- if that's correct, is that embedded in the 30% EUV revenue growth?

Yes, it is. Yes, it is. Yes.

Yes. So we need...

So you have a reversal in this quarter, and you will see it come back in the subsequent quarters. For the full year, it's neutral price.

Yes. It's a wash.

It's minus 2 and plus 2, so it's a wash.

And our next question comes from the line of Aleksander Peterc of Societe Generale.

So you gave us a pretty clear outlook on EUV for '22. I just wondered about the longevity of this stronger cycle. Did you see the management in dry? Because that's where all of the upgrade today is coming through for '21. So I'm just wondering, is all of that then carrying into '22 as well? And then, just very quickly, if you could also provide a quick comment on ASPs? They were again very strong in this quarter, so if you could give the puts and takes there.

I'll do the ASP -- well, Roger will do the ASPs. And your question on the...

Sustainability.

On the sustainability, yes, and especially deep UV and apps. I think, as I said it on the -- as an answer on an earlier question, when we look at how we see the deep UV market and the aftermarket today as compared to 3 years ago, about 2018, I think we have a different view. And that's driven by the fact that, of course, our entire deep UV portfolio, which is immersion and dry, we have -- we cannot fulfill the demand of our customers all the time.

And that has to do with the fact that our analysis shows that with the combination of, let's say, advanced sensing technology, 5G, the ability to process all the data through high-performance compute. And that, in a distributed fashion, basically leading-edge compute also goes to the edge. I call that distributed systems. And the distributed system is, for instance, a car, but it's also one of our machines in the field.

And then, increasingly, requires collection of data, transport of data, processing of data, not only to the most advanced high-performance compute but as part of the distributed system. And that system inevitably includes mature technology, which could be image sensors, power ICs, MEMS, analog solutions. It's the whole thing. And that means, if we would have a high line system available today, it will be sold yesterday.

So it's everywhere. And that has to do with the proliferation of chip technology and the distributed computing and the distributed systems that we are seeing. That is something that will not go away in our minds. This is why I said also in my prepared remarks that we are also planning and looking into what is the level of our deep UV capacity increase that we need, both in mature and in emerging. And it's going to be double-digit increase as from where we are today.

Now how much can we stretch that? Or how much is needed? That's exactly what we're doing with EUV. We need to line up with the supply chain. And we need to go deeper into what it takes in terms of capacity lead time. Do they need to build square meters? Do they need to hire people? Is it going to be possible through cycle time reductions or through process optimization?

This is the work that we're doing today. Because remember, 5 months ago, we were looking at a completely different world, yes? So this is a ramp-up, but we do believe that this has a long runway.

Yes. On the ASP for EUV, it's well-noted, you're absolutely right. So this quarter, in Q1, we're looking at an ASP for EUV of about EUR 160 million. Part of that is configuration. We've talked about that in the past. If you look at the last 3 quarters, we've consistently been looking at an ASP of around EUR 145 million. And this time, it's a little bit up for another reason that I'll come to in a moment.

So in fact, what you see is that -- in fact, in the last year, we've been positively surprised by the options on the 2, the options that were ordered on the 2. And as a result of that, the configuration was richer than anticipated. And as a result of that, you fairly consistently see now that the ASP ranges above the EUR 130 million that we talked about in the past.

And it's fairly consistent. We've seen EUR 145 million or up. The reason that is extra of this, this year, there's a few accounting things. We've talked about accounting for VPAs in the past. So there's a few accounting reasons why it's a wide higher, it's a wide EUR 160 million rather than EUR 145 million.

But we're out of the -- just looking at the configuration that we've seen on the tools in the past couple of quarters, the EUR 145 million anchor points, EUR 40 million ASP for EUV is probably the right way to go. And I'd say that, recognizing that this is about the last quarter that we're really looking at a large number of fees in the revenue.

And our next question comes from the line of David Mulholland of UBS.

Just coming back on the EUV capacity increase and shifting to 55 tool capacity for next year. Obviously, there were some changes made last year that ended up limiting what potentially should have been higher shipments this year, and some of that probably kicks in next year. But in your discussions with customers, Peter, just -- obviously, there's a very, very strong cyclical tailwind to the industry today that's driving appetite for capacity increase.

But EUV is on a very, very long time horizon for customers 18 months rather than 6 months planning. How do you feel like their assumptions have changed that's driving that upside? Is it primarily because of their confidence and penetration of EUV in higher layers? Or is there an element of actually just building bigger nodes now that's driving the need for more capacity at this point?

I think, David, it's both. I think we're seeing -- I only -- what I refer to as a comment that was made by the CEO of one of our large customers. That actually manage and know we're going to double the number of EUV layers on our next node. And I think that's a trend that we are seeing, simply because the advantages of EUV are not only, you could say, the pure economic cost per layer because you can eliminate multiple patterning, but it's also the electrical characteristics and the simplicity of the process. Lower risk, that's one of them, yes? So there's many other side effects that actually lead people to go for EUV.

So yes, it's higher layer counts, that's what we're seeing, which is true for Memory and for Logic. But also, when we talk to customers, we all talk about bigger nodes. It has to do with this secular trend, trend #2 that I talked about. And that number -- and trend #3, the technological sovereignty, will just be a layer on top, which will not go away, which will take time. It takes 2 to 3 years to build a semiconductor factory, and we put it into action.

So -- but I think this is -- so it is bigger nodes. It is strategic investments. It's High-NA accounts. And this is what I -- as an answer to a previous question, we all need to take this into consideration. Some of it we already saw coming, but what does it mean for our capacity needs beyond 2022, which, by the way, some of that capacity, if that would be needed, would be the result of, like I said, process optimization, cycle time improvement, different work schedules and potentially square meters, that could be. So -- but it's all of the above.

And just one quick follow-up. Obviously, from a financial perspective, at some point in the near future, you need to start building those tools through the supply chain, given the lead times. Are you requiring customers to place deposits and make the financial commitments that you've been starting to make before any of those tools get built? Or is there an element still this year where you're going ahead and starting production before you've had firm deposit-associated orders from the customer?

Yes, David, the policy hasn't changed there. So that means that when the customers put in a PO, a down payment is required. And I think we mentioned in the past that the amount of the PO is a little bit dependent on the moment where the order is being placed. So if the order is being placed nicely in line with the lead time, then the down payment is lower than one that's placed a little bit later than that.

But you also know that the PO in and by itself is a bit of a nonevent for us because we work so closely with the customer that we sort of understand what they're doing, what they want and when they need it. So the PO is a bit of a nonevent for us. But yes, the short answer is absolutely, there will be down payments to place at the moment of the PO.

And our next question comes from the line of C.J. Muse at Evercore ISI.

I guess, first question, Peter, can you speak to the sustainability of Memory? I think the headline above 50% might be a bit concerning. However, if I think about EUV in there as well as the strong pickup in EUV, in voltage and contrast imaging, it certainly looks like Memory, ex those 2 things, is tracking more like up 20%, 25%. So I would love to hear your thoughts on the moving parts there, and how to think about sustainability into '22?

Yes. I think, for this year, it's not a surprise. I mean, you just need to go back to our conference calls in the last 2 quarters, where we actually saw a recovery of especially DRAM starting. And I think I gave some color 3 months ago where I basically said, yes, we had a decent shipment core in Q4 of last year, which basically goes into production in Q1.

But if you look at a bit growth of 20% this year, then our calculations show that we very quickly -- at the next output capacity of our DRAM customers in this year, where -- and we would reach that pretty quick, so we would need more capacity addition. That's exactly what we have seen. So I think it's just the beginning. These things -- so yes, I think, yes, it will move into next year.

Now you've been around also long. So that -- in the Memory business, yes, there is more tendency to have, from time to time, some overcapacity and some undercapacity. But how long that will last, I don't know. But still, this is some -- [there was a BM above] telling me, say, this secular trend, when there's such a high demand in logic, all this stuff doesn't only work with only logic. So it also needs memory, yes?

So it's going to be quite interesting to see how quickly the Memory capacity will be added throughout this year, early next year, and what the underlying bit growth percentages will be. Now I think, having said that, you notice also Memory is more cyclical than Logic, but nothing that indicated to me, at this moment in time, that we are looking at building an overcapacity. There's nothing I can see at this moment in time. We're just starting.

Very helpful. And then just a follow-on question on the DUV side of things. I think you guys have talked about growing your non-EUV tool business by about 40%. So that's roughly $7.7 billion this year. Is there a way to frame how much capacity you're planning to add on the non-EUV side relative to that $7.7 billion for us to kind of gauge what you're capable above into '22 and beyond?

Yes. As I think, we -- if we should calculate it the way that we would calculate about, for example, we would just be over $8 billion, yes?

$8.2 billion.

About $8.2 billion. And that's really stretching it, in terms of our production capacity this year. If it would have more capacity, we could probably do more. Okay. Likely, we would be able to do more. So your question really is answered in a sense that I already said in my prepared remarks, we're looking at increasing our deep UV capacity, both dry, mature and immersion. And that's not going to be 5%.

I mean it's going to be double digit. But that will take some further analysis of what needs to be done. Is it through optimization, yes? Things like process optimization and added workforce or cycle time reduction. Or has it to do with square meter capacity for like factories? That is something that we'll -- as under consideration today, and we're looking into that together with our supply chain. So not so much for us, that is really in the supply chain, yes?

So the reason why I say this is because we believe that we need extra capacity, yes? And it has to do with the underlying trends, the trends that I talked about earlier. So yes, the 40% makes me think to just over $8 billion, yes? Where you have $7.7 billion, so you're a bit more conservative. But this is -- directionally, it is correct, but this is the answer.

And our next question comes from the line of Andrew Gardiner at Barclays.

Another one, Peter, on the sort of capacity planning issues that you guys have just been talking about. I'm wondering, given the changes that we've seen from a number of your lead customers or at least the public announcements from your lead customers over the last 3 or 4 months and the fact that you are now having to rethink the need for additional capacity expansion, whether your site or within the supply chain, are they now giving -- and I presume yes, they're wanting it sooner rather than later.

It always seems to be the way. And yet you've got fairly extensive lead times, particularly so for EUV, perhaps slightly less so for DUV in terms of building out this capacity. Are you getting better visibility today in terms of what those customers are going to need looking out over the next 3, 4, 5 years in order to give you that confidence in making such a decision?

Yes. It's a good question. I mean, and also, you've been around also. I think all these people have been around for a long time. So that -- yes, when we underinvest, there's also going to be a knee-jerk reaction, like it always happens. So this is more the shorter-term correction of a misplanning, yes? That's happening today.

That's what I call -- it's the trend #1, yes? But clearly, customers understanding that our capacity lead times and our manufacturing lead times are wrong. So they do share with us a longer-term vision. Now I'm not 5 years out, that's a bit too much to ask. But certainly 2, 3 years out, yes?

So I think, I would say, for 2022, 2023, we definitely have the discussions about what they need in terms of expansion plans, in terms of that they have plans, in terms of what they see. When we talk about Logic, the anticipated capacity ramp in terms of the tape-outs that they have on the shelves so that they see it coming on and 5 and then 3.

So it is not just some pie-in-the-sky thinking. It is this real underlying discussion that they have with their customers. And their design requirements that they get from an increasing customer base, which has to do with high-power compute there with AI and with -- with all the applications that we talked about because of the whole digital transition.

So yes, there's more visibility to the point that the planning visibility is, I would say, deeper. And this is also where we base our request to the supply chain on. It's a vet on that discussion. And which is going to be -- it's a costly discussion because you not add capacity for nothing. So this is -- and it is also true for our customers. It's going to be costly, but they've been very explicit to the outside world and to the media about what their plans are.

It's thought through based on what they see for their customers and the customers of their customers tell them. So that's why I talk about these 3 trends. And this is -- this, basically, is the base for the visibility, but how much is in the end will be or we will need. There's always a risk of underestimation, but also risk of overestimation.

And that's what we need to take into consideration when we say what's wise. And then, like I said, this is just something of the last 3 or 4 months where it starts to accelerate. It gives us a bit of time in our supplies, particularly, to figure out what is with them.

Thank you, Peter, for that insight. Even if I think you did call me old at the outset there, I still appreciate the color.

It's all relative. And from my perspective, you are very young.

Our next question comes from the line of Krish Sankar of Cowen and Co.

I've also been around quite a bit, so let me ask you 2 quick questions. Clearly, a nice jump in EUV bookings in the March quarter. Can you give some color on how the EUV bookings are trending for the current quarter? And what is your EUV backlog in terms of units? And then I have a follow-up.

Roger?

Yes, I think I could take it. So in terms of order book, the current backlog is at EUR 7.4 billion for EUV. So that's pretty strong and obviously also driven by significant intake in the last quarter.

I'm not going to comment specifically on this quarter, but in general, just listening to the commentary that we made on our expectations for next year, it's pretty clear that, also for the next quarters, we do expect a healthy order intake for EUV, right, because at this stage, EUR 7.4 billion. And if you do the math, what you think system business is going to be next year, if you translate that at the 55 capacity that we've been talking about, it's pretty clear that we're still looking at a significant order intake. Expect this for the next -- for this quarter and for the next quarters.

Got it. Well, that's very helpful, Roger. And then just as a follow-up. Peter, you -- I know you gave some color on like the Memory outlook for this year. Can you just give little more insight in terms of how to think between DRAM and NAND in the context of a memory growth of 50% for this year?

Yes. I think the -- I think we're also -- it's an extension of what we said in the previous 2 quarters. It is primarily driven by DRAM, but we also see that 3D NAND will follow. I mean, what we see is basically following the utilization of our tools in the field. That will follow. But it is primarily driven this quarter, next quarter, and I think the majority of this year by DRAM.

And our next question comes from the line of Alex Duval at Goldman Sachs.

Congratulations on the results. Just have a quick question on EUV gross margins. I mean the expected trajectory there, and so there's a small sort off the end this year to be at good gross margin levels. I wonder if you could just talk through the drivers of improvement as we go towards the end of the year. Then you referenced, related to this, that you're seeing higher ASPs related to a particular area the CMO. I was just wondering, to the extent that we see an expected ASPs of a D model could also be a gross margin with -- of EUV.

Yes, Alex. At the beginning, you were a little bit tough to understand, but I think your question really was, how do you expect the gross margin to further develop on EUV. And as we mentioned before, we do expect in the second half with the introduction of the D model, we expect EUV system gross margin to be at the level of the corporate gross margin. So -- and I can confirm that that's exactly what's going on.

You are right that the -- that we did see a little bit of a reset on the ASP for the C model, right? So we've been talking about 130. But as I mentioned, the configuration quarter after quarter turned out to be richer than we previously anticipated. And I think it's fair to say that the 10% to 15% uptick that we've been talking about on the -- for the D model, in comparison to the C model, still hopes right? So if you take the 145-ish basis for the C model, then I think a low, low TAM increase over that in terms of ASP, I think, is probably what you should be looking at for the ASP calculation.

And of course, to the extent that, that is better, that, of course, is also helpful a little bit on the gross margin side as well.

Yes. Although the D has also somewhat higher cost.

Sure. Yes.

So that -- it's 100% to the bottom line, that we have some higher cost on the optics and also other parts. Correct, but it will help.

And our next question comes from the line of Amit Harchandani of Citi.

I'm Amit Harchandani from Citi. My first question is on the topic of deep UV. If I try to collate all the data points you've shared with us, is it fair for me to say you would expect deep UV sales to be up year-on-year in 2022, given that you are just getting started in memory and cyclicality trend could persist into 2022? And as a follow-up to that, if I may. You've talked about building capacity for deep UV. You are stretched at EUR 8 billion. And you talked about double digit, I think, earlier in the call. Is it fair to say that you would expect this EUR 8 billion to be exceeded, or basically even a higher contribution from deep UV? If not in 2022, then certainly beyond 2020?

Yes. You've listened carefully, Amit, and that is true. Everything you just said, do we expect UV sales to be up year-on-year. I don't know about the year that you are imagining. But the fact that we are looking into increasing our deep UV capacity means that we believe deep UV sales will be up and in what year and is year-on-year. But I think, medium to long-term, we have, indeed, reassessed the need for deep UV capacity going forward.

And when you talk about capacity increase, you don't do this for 1 year. You do this because you feel medium to long-term that is needed. So yes, and it also means that if we do that, then -- and we say we'll be around EUR 8 billion this year on deep UV. Yes, that we would expect that, that would increase because otherwise, you don't need to build capacity because we could do with what we have today.

Okay. I just wanted to clarify that. And secondly, very quickly on geopolitics. We've heard a lot of noise. Based on what you have said today, it seems like you expect the whole geopolitical dynamic to be a net positive. It may be some downside risk to China potentially being offset by potential upside in Europe and the U.S. Is that a fair assessment? Or more broadly, what are the latest -- what is the latest, in your view, versus what you shared with us end of January?

Yes. I think Roger said it, is that I think we are now where we are because we're through Q1, we're in Q2 now, we have included that EUR 600 million upside into the number that we gave you. However, we don't control the geopolitical situation and the lawmaking. But given the market situation where we are today, if some of that EUR 600 million would not be there because of geopolitical roadblocks, then the demand, actually, will ship those systems somewhere else.

And if you then look at how that affects shipments to China versus the longer-term geopolitical impact. It's just a matter of timing, yes? That's -- the first one could be short-term this year, the others are definitely long-term because when you -- they all -- it is all about where do we build a new fab. Well, it takes 2 to 3 years. So when you really look at adding capacity at the time frame, 24, 25. Is that capacity starts coming to the market. So it's all -- that's all matter of different timing, a different time perspective.

All right. We have time for one last quick question. So if you were unable to get through on this call and still have questions, please feel free to contact ASML Investor Relations department with your question. Now operator, may we have the last caller, please.

That is from the line of Didier Scemama of Bank of America.

I think the most important question was just asked and answered, so thank you. I wanted to ask you question about Installed Base Management for EUV, Peter, if you would like to maybe share with us what do you think the sort of revenues could be for EUV installed base this year? And how you're thinking about the slope of growth over the coming 5 years, those sort of things. So we have a feel for the change in the mix towards the sort of more recurring revenues in Installed Base, that would be great.

Yes. So on the more recurring side, I think what we've said in the past is that the way to model that for EUV is to take the ASP of a tool and take about 6% of that. And that's when the tool is up and running at it's envisaged capacity, then you would have 6% of that ASP as the recurring revenue for service and maintenance for that tool.

So that excludes upgrades, but just for the regular sort of a maintenance. 6% of the ASP is what you would see there. So again, the assumption, the tool needs to be up and running at the envisaged capacity, and it should be out of warranty. As long as the tool is in warranty, the number is a bit lower. So that's really the way to model it.

As it comes to upgrades, that's a little bit more difficult to do, right? Because upgrades is lumpy, upgrades is dependent on what the customer wants in terms of productivity gains, what they want to make available to us in terms of the machine time. So that's a little bit harder to predict. But at least, on the regular service and maintenance side, this is the way to model the 6% of the ASP once the tool is up and running and out of warranty.

All right. Thank you. Before we sign off, I'd like to remind you that our Investor Day will be September 29, 2021. The event is currently planned to be held in London. We moved the date from June and hope we can have a face-to-face meeting at the time, but of course, this will depend on the progress against the virus.

More details will follow in due time, and we do hope you'll be able to join us. Now on behalf of ASML, 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.

Thank you. This concludes the ASML 2021 First Quarter Financial Results Conference Call. Thank you for participating. You may now disconnect.