ASM International N.V.

Victor Bareno - Head, IR Chuck del Prado - President & CEO Peter van Bommel - CFO

Peter Olofsen - Kepler Chevreux Chetan Udeshi - JPMorgan Tammy Qiu - Berenberg Bank Philip Scholte - Kempen

Thank you, Leona. ASM issued its 2015 second quarter results last evening.

For those of you who have not seen the press release, it, along with the latest investor presentation, is accessible on our website, ASM.com. We remind you that this conference call may contain information relating to ASM's future business or results in addition to historical information.

These forward-looking statements involve risks and uncertainties that could cause actual results to vary materially from those expressed or implied in such statements. These include, without limitation, statements relating to revenues, margins, cost reduction programs, liquidity, breakeven levels, strategies and economic conditions.

Please refer to ASM's press releases and filings with the U.S. Securities and Exchange Commission on Form 20F and 6K for more information on the risk factors that could affect results.

All forward-looking statements are based on the information as of today, July 30, 2015 and the company assumes no obligation to update these statements. I will now turn the call over to Chuck del Prado, President and CEO.

Chuck?

Thank you, Victor and thanks to everyone for attending our second quarter 2015 results conference call -- and for your continuing interest in the company. After a review of operations, Peter van Bommel, our CFO, will join me in answering your questions, of course.

So let's start with a review of our second quarter financial results. Performance was again solid in the second quarter with revenue, gross margins and operating profit at the highest level ever for our front end operations.

Revenue in the second quarter increased to a new record, up 24% from the first quarter and up 35% compared to the second quarter last year. At €201 million revenue in the quarter was slightly above the top end of the range of €180 million to €200 million that we had guided for.

Our ALD businesses again it drove the increase. By customer segment, the revenue stream was led by memory, particularly DRAM, followed by foundry and then logic.

The gross margin increased to 45.1%, also a new record high and up from 43.3% in the first quarter of 2015 and 42.3% in the second quarter of last year. The increase was driven by a relatively strong product mix and further efficiency improvements.

SG&A expenses rose by 16% compared to the previous quarter and by 24% year on year. However, as a percentage of revenue, SG&A dropped to 12% compared to 13% in both the first quarter of 2015 and the year ago period.

R&D expenses increased by 15% sequentially and by 52% compared to the second quarter last year. As a percentage of sales, R&D amounted to 12% in the quarter, down from 13% in the first quarter and up from 10% in the year ago period.

Apart from a currency impact the increase in R&D was primarily driven by requests from customers related to new opportunities, similar as in previous quarters. The second quarter results also included a relatively small restructuring charge of €500,000.

Operating results increased strongly to €42 million in the second quarter with an operating margin of 21%. This compares to operating income of €29 million in the first quarter and €27 million in the second quarter of last year.

Financing result in the quarter was €12 million negative and primarily consisted of a translation loss of a similar size. In the first quarter of this year the financing result still included a translation gain of €28 million.

As a reminder, a substantial part of ASMI's cash balance is held in U.S. dollar.

The translation loss in the second quarter mainly reflected the depreciation of the U.S. dollar at the end of this period as compared to the end of March.

Results from investments which reflects our share of approximately 40% of the net earnings from ASMPT, improved to €21 million for the quarter, up from €13 million in the first quarter and €16 million in the year ago period. These figures exclude the ongoing amortization charge which amounted to almost €7 million in the second quarter.

For the full year this charge is projected at €27 million. ASMPT sales increased in the second quarter to HKD3.8 billion, up 22% from the first quarter of this year and up 9% compared to the second quarter of last year.

ASMPT reported bookings of $494 million for the second quarter, an increase of 6% compared to the first quarter, but a decrease of 22% year on year. This decrease was mainly caused by a sizable order in Q2 of last year from one particular customer.

Now turning back to ASMI's consolidated operations. ASMI's net earnings on a normalized basis amounted to €47 million in the second quarter, down from €66 million in the first quarter and up from €40 million in the second quarter of last year.

The €40 million drop in the translation results from Q1 to Q2 as mentioned is the main reason for the sequential decrease in net earnings. Without translation results the net earnings on a normalized basis would have increased from €39 million to €59 million.

Our new orders in the second quarter continue to be solid at a level of €167 million, up 5% from the first quarter and up 39% year on year. And was in the expected range of between €160 million and €180 million.

Similar to revenue orders were mainly driven by Atomic Layer Deposition. In terms of customer segment, bookings were led by the memory sector followed by foundry and then logic.

Within memory bookings were spread among multiple customers in the DRAM segment and to a lesser extent NAND flash. Let's now look at the balance sheet and cash flow.

The cash position decreased to €363 million at the end of June compared to €423 million at the end of March. This decrease was primarily the result of the cash spent on the share buyback and also dividend paid to ASMI shareholders and negative currency effects partly offset by the dividend received from ASMPT.

On an absolute basis net working capital rose to €149 million at the end of June, up from €123 million at the end of the first quarter. This increase was mainly due to an increase in the accounts receivables which in turn was caused by the higher revenue level.

The number of outstanding days of working capital measured against quarterly sales, decreased slightly from 68 days at the end of March to 67 days at the end of the second quarter. We generated €11 million in operating cash flow, down from €36 million in the prior quarter.

On the positive side, operating profitability increased substantially, but this was largely offset by the earlier mentioned increase in working capital. During the second quarter we completed the €100 million share buyback program.

We spent the remaining part of €40 million to repurchase approximately 900,000 of our own shares. Since we announced the program last October we have bought back a total of nearly 2.6 million shares at an average price of €38.55 per share.

The number of outstanding basic shares decreased to approximately 61.9 million at the end of June which compares to 62.6 million shares at the end of March and 63.7 million shares at the end of the third quarter of last year. Just before -- and that was just before we started the buyback program.

In addition, we paid €33 million in dividends during the quarter after the shareholders' meeting in May approved our proposal for a 20% increase in dividend per share. Dividends received from ASMPT during the quarter amounted to €25 million.

So we announced today our intention to delist our ordinary shares in the U.S. from NASDAQ.

We have taken this decision because of the low and declining trading volume. We plan to have these shares traded on the OTC market in the U.S..

After delisting from NASDAQ, ASMI will continue to maintain the level of disclosure expected by the international financial markets. Let's now look a little closer at our ALD business.

The strong performance in the second quarter again shows that ALD is a solid growth driver for our company. Today ALD has become a critical and enabling technology for the manufacture of virtually all leading edge logic and memory devices.

ALD supports our customers to stay on Moore's Law and to transition to the next technology nodes. In the memory sector our ALD solutions for multiple patterning have supported the further scaling of DRAM and NAND flash devices over the last several years.

For logic devices our ALD high-k solutions have now been extended for four device generations. As highlighted during earlier calls, we have steadily expanded our customer base in the past one to two years beyond the leading device manufacturers.

We now supply our ALD products -- our ALD solutions to all of the top 10 CapEx vendors in our industry. The strong increase in our ALD products revenue in the second quarter was supported by a sizable contribution of several of these newer customers.

With the advantage of more than 50 years of developing ALD, improving our technologies and building customer relationships we're confident we have a strong ALD offering in place. We remain sharply focused on further expansion of our pipeline of new ALD applications.

An example of an innovative application that we highlighted at our technology seminar earlier this month at SEMICON West is the use of PEALD doped oxides for fin doping in the 14 nanometer node. This is just one example where ALD technology supports our customers in addressing the challenges of ever smaller geometries in 3D device architectures.

We expect that the number of ALD applications will continue to increase with the transition to the next technology nodes. In memory we expect the spacer defined multiple patterning where we have a strong position with our ALD solutions will be a strong driver of ALD market growth.

Spacer-based multiple patterning which has so far mainly been used in memory segment of the market, is also on the roadmap of customers in The Foundry and logic sector. We believe we're well placed to address these growing patterning opportunities in the logic and foundry segment as well.

In addition to the patterning opportunities in logic and foundry, you expect that successive generations of FinFET devices will require steady increase in a number of ALD based process steps. As a market leader we believe our company has strong opportunities to benefit from continued solid growth in the ALD market in the next few years.

So looking at the market environment, momentum in a number of key semiconductor end markets has recently weakened, evidenced by reduced growth expectations for the smartphone and PC markets, by increasing inventories in the semiconductor supply-chain and by a weaker pricing environment in the DRAM sector. Market watchers such as Gartner and VLSI have accordingly taken a somewhat more cautious view, projecting the wafer fab equipment market in 2015 to be somewhere in between flat and a low-single-digit growth.

Memory spending in the second half of this year in DRAM on the 20 nanometer node is expected to be lower compared to the first half of the year, while spending in NAND flash is expected to stay at the level of the first half of the year with the share of 3D NAND compared to planar NAND likely to increase towards the later part of the year. In the logic foundry sector equipment spending in the second half of this year is expected to be at a lower level than the first half and spending is expected to primarily focus on development tools for 10 nanometer.

So let's now look at the -- yes, the detailed outlook we gave as part of our press release overnight. For Q3 we expect sales to be between €155 million and €170 million on a currency comparable level.

And the Q3 order intake on a currency comparable level is expected to be in a range of between €130 million and €150 million. Beyond Q3 the visibility on timing of next technology nodes insertions of new equipment capacity is still limited.

At this moment we're happy to answer any questions that you may have.

A couple of questions starting with memory. I think in the Q1 call you talked about 3D NAND where you were seeing several potential applications for ALD.

It seems that some of the node makers are going to ramp 48 layer 3D NAND going into next year and also some of your peers seem to have some attraction there going into the September quarter. When will we start to see this in the ASM numbers?

Is there already something in your Q3 sales and order guidance? Or is it something we will start to see going into next year?

Okay, yes, as we said before, compared to planar NAND current generations or 3D NAND are still relatively light on ALD in our view. But we also see that the need for ALD is likely to increase indeed in future generations of 3D NAND.

And we have been very, very focused on increasing our presence with those customers based on the much better position that we have built in memory over the last five years through the expansion of our existing ALD applications. And as a result of that, Peter, our belief is that, based on those engagements, that we expect an increase of our revenue for 3D NAND applications going into next year.

But we don't expect 3D NAND to be a significant contributor to the P&L in the second half of this year.

Okay, that is clear. And then maybe on the announcement earlier this week of a new memory technology by Micron and Intel, is that a technology with specific process steps that may require ALD?

Yes, you clearly follow the trends very well in the industry. And of course, yes, we have to be aware of these trends also.

And we of course are -- our customers don't allow us to talk about specific programs with them. But the test way to answer the question as a result is that if we look at next-generation memory developments, such as resistive memory and other technologies, the introduction of new materials and very often also deposition technologies play a very important role because as a result of those new materials ALD can be an enabling technology to bring those materials into play.

And those are exactly the developments that we're focusing on in our engagement with customers. That is I think the best way to answer it, Peter, without going into details on our specific engagements with customers.

And then my final question on the logic foundry side of the business. Where do we stand in terms of equipment selection for the 10 nanometer nodes?

Have these selections already been made or is it still -- are the evaluations still going on?

Yes. Well, in principle it is clear that if you look at the 10 nanometer node investments they -- for logic and foundry, we don't expect those to hit the P&L this year.

Looking at the timing of those customers at this moment in time we expect those to mostly hit a high-volume, I'm talking about high-volume investments, in the course of 2016 for 10 nanometer for logic and foundry. But at the same time, although that is the case, a lot of those customers are already pretty far in their selection because they are in a serious -- pretty far in their -- let's say in their development and integration process of the devices that should hit the market -- that they should start building an high-volume in the course of 2016.

Yes, so we think we have a pretty good view of what our position is in preparing for -- in those customers preparing for that node. And as a result of that in general our view is that we believe that the number of applications that we're engaged in with those customers for 10 compared to 16, 14 we believe will significantly increase.

The number of applications you are involved in, that will increase?

Yes, compared to 16, 14.

And to what extent will you have to share these applications with other equipment vendors where you may have still been the single supplier at 14, 16?

Could you repeat that one more time?

Yes, the question is as the number of applications is going up at the same time we see companies like [indiscernible] also looking to get a bigger share in the ALD market. So if you take the higher number of applications and market share developments, what then the net result will be overall opportunity be bigger than it was at the 14, 16 nanometer?

Well, of course, again, I -- we have said that we -- when I talk about the ALD market, taking 2014 as a baseline, that in the three to four years we expect that market to double from $600 million to over $1.2 billion. At the same time, so -- and we also said that competition is starting to focus more and more on those markets.

So it is logical that competition also here and there will take -- will increase their piece of the pie. But since that market is going to double we believe that we're in a great position to still outgrow the industry in the coming years through our leading position in ALD at this moment in time.

So, I think competitors may go through an improved position, but we're eager to maintain our position as a leader in this industry. Yes.

So I trust that answers your question.

First one would be regarding the oil trend. You mentioned last quarter that you were expecting a [indiscernible] order in Q3 and I guess that is what your guidance shows.

But that for Q4 or Q1 you see a rebound coming from logic foundries. From what you just said about 10 nanometer [indiscernible] 2016 story.

Flash NAND is not expected to continue so much in H2 and DRAM is expected to go down. So your first take on Q4 order intake, is it that it should remain at around a flattish level compared to Q3?

I'm not going to give a more specific indication on Q4 because then we should have done that in the press release also. The only thing -- yes, the only color we gave is -- with regard to specific numbers is on Q3.

And for the rest the only guidance we gave is that we in general in the industry expect the spending in memory, in DRAM especially, to be lower in the second half than in the first half. And in flash, for it to stay at the same level as the first half, that is what we expect.

But with more of a shift from planar to V-NAND. And we also said that our penetration in V-NAND, any let's say increase in revenue there will not show before next year.

So I think you should combine those data points yourself. And that combined with logic and foundry where we expect that, yes, the investments in 10 for high-volume manufacturing will only start to happen in the course of 2016.

And the exact timing there I think all the equipment makers are anxiously looking for a more detailed forecast from our customers. But for us the key thing is besides the timing on whether it is quarter X or quarter Y, the key thing for us is that we're well-positioned.

That we're well-positioned for the next node in DRAM, in logic and foundry and that we gradually increase our let's say presence in V-NAND. And from that point of view we believe that we have been taking the right actions and that we can build on that going into 2016 and 2017.

Also you mentioned that the weight of Asia is going and new customers notably are taking more and more space. Could you give like [indiscernible] fleet of yourself and also the weight of your top three customer in H1 of this year?

I don't know how to answer your question in the best way. But from the top of my head again, we try in our dialogue with our investors, we always try to show the -- what the top three of our revenue, our top three -- what the percentage of total revenue is of our top three, what the total percentage of revenue is of our top 10 as part of our total revenue.

And we also indicated that our top three pretty much aligns with of the top three in the industry and our top 10 also pretty well aligns with the top 10 in our industry. And I think that is something that we achieved since 2012 for the top three that are -- the spending in -- the percentage of spending of total revenue of our top three is much better in line with the spending of the top three in the industry as a whole.

And that since 2014 also the representation of the remainder of the top 10 is much more in line with the industry than ever before. So I think that indicates how well, especially with our ALD product lines, we have been penetrating the complete top 10 in terms of rank in terms of CapEx spending.

And based on that you should be able to derive the geographical spending.

And one last question on my side. It was regarding your margin, notably your gross margin which is pretty high in Q2.

And you mentioned some product mix positive impact. With them going down in H2 what should we think about the gross margin evolution?

And also regarding OpEx, in absolute terms it has increased [indiscernible] it's a bit below. What also should we input or what do you see for remaining of the [indiscernible]?

I can only repeat what I have said in previous calls. I mean when you look to the gross margin, then we have always said that in an upturn we expect low- to mid-40s.

Now we were happy to show now after being on a 43% level for quite some quarters that we indeed are able to grow maybe to the mid-40s so the 45%. That of course is what we have written down also in our press announcement by a strong mix in the quarter.

A mix, as you might know from previous calls, that can change within quarters. We see also that gradually the improvements are coming in from all the logistics, the supply chain improvements that we have done so further efficiency was also visible in the second quarter.

So that is the color I think that I can give about the gross margin. So still sticking to what we always have said, low to mid 40%s, a little bit dependent on the exact mix and the actions that we do on the efficiency side.

When you look to the costs we're spending -- OpEx is strongly impacted. So by currency movements, so nearly half of the OpEx spendings in Q2 over Q2 last year, the increase on that is caused by currencies.

We're spending more on R&D, that is also to facilitate what Chuck mentioned earlier. The inroads that we're making with more applications, more customers which is eating up more of our R&D spending.

So we perceive that those are investments in our future. And in SG&A we have seen some costs, it is a combination of two things, it is a combination of a few one-offs that we had in Q2.

And on the other hand we have some slightly higher cost going into 2015 which is also in line with early indications that we always have given. We expect SG&A to increase a little bit over time.

But it is relatively stable within the [indiscernible]. The only reason for the delta between Q1 and Q2 this year was given by the fact that we introduced or that we changed the actual changes in our long-term incentive program in the company.

And that meant that that for one quarter we had [indiscernible] for those costs. So that was the biggest change from quarter to quarter plus the one-offs that I mentioned.

So I think that gives you -- should give you sufficient color for making the calculation by yourself.

I had two questions. First was you have answered previously that you expect competition to take some share in midterm, but I just wanted to hear your thoughts on new tool announcement from AMET where they are claiming to have structurally different [indiscernible] architecture for ALD than conventional tools.

Any thoughts there? And secondly, any feeling you could give in terms of what your customers are thinking overall in terms of their CapEx spending plan for their Q4 or early part of next year?

Not quantitatively but in terms of qualitative comments? Thank you.

Okay, okay, CapEx Q4 Q1. Okay, so I will start with the polite question that you asked.

I would like to say without going to do a design review of individual competitive tools which this is not the forum for, of course. I think our answer to the competitive landscape did not change based on the announcement that they made.

It would also be a little concerning if we would only learn of this tool as a result of that announcement in May because this tool has been in the field for one, two years already. And only the public announcement was done in May.

So of course we try to watch what competitors are doing continuously. And so the tool was -- the introduction of the tool was not new to us, but we don't under estimate any of our competitors and this is just part, Chetan, of yes of our statement that we made earlier, that we have seen activity of competitors stepping up over the last couple of years.

And that as a result of that in line with our expectations and which is logical that as a result of this market to increase to double in the next couple of years that in selected applications, selected parts of the market, some of these competitors will step up and here and there make some inroads. And again we're trying to anticipate these steps from competitors as well as we can and we're focused to maintain a leadership position in ALD and we believe in the strengths of our ALD offerings and that has not changed in the last two months.

Okay, then on the CapEx. Yes, you were asking for a little bit more color on Q4-Q1.

Yes, we cannot add much more than what we already shared earlier in the call. The only thing I can repeat is that for DRAM -- so far with the visibility we have now we expect DRAM to, at least in the 20 nanometer -- in the current node that spending will be lower in the second half compared to the first half.

And that it is still unclear how quickly the industry will really step up for significant investments in DRAM for let's say the 1X nodes. That is still unclear.

And most of the time those decisions are being made pretty short in advance and then equipment makers are expected to anticipate very, very quickly. And the only thing that we can do is that we have the right solutions in place for them to choose us.

In flash we talked about earlier there is a transition ongoing from planner to V-NAND, some initial investments are being done already in the second half of this year in Q3 and Q4. But as we said for us as a company we don't expect to show positive impact of that until we're in the next calendar year.

And for logic and foundry, yes, the key thing there is when do the investments and the bookings really kick in for the 10 nanometer node. And yes, visibility we have now is the very likely that will happen in the course of 2016 and not in the remainder of this year and then the question is how quickly in 2016 will that happen.

And again, there we're ready when our customers are ready and we anticipate that in logic foundry -- we will be able to increase the number of applications, ALD applications that we're engaged in compared to 14, 16. We believe that we're in a very, very good position there.

So it is just a matter of timing, indeed.

My first one is on the [indiscernible], you have mentioned that logic and foundry is heavily dependent on what the timing for 10 nanometer kick-in. So does that mean 14 and 16 has already been done from an equipment perspective, at least from ALD perspective?

And also regarding the tool design. From your experience what the factor or specification they are both -- about ALD?

Is that throughput because we know that is actual the bottleneck for ALD or is that [indiscernible] more about a sufficient level of ALD without thinking about the throughput? And last one is that what is your thoughts on further share buyback.

Thank you.

Okay, Tammy, thank you for your questions. Yes 16, 14, yes, we do see some -- there is some business represented for 16, 14 in the second half for logic foundry.

But it is not significant. It is like we said before, memory has so far been a stronger driver.

Has been a stronger driver in the -- yes, it's been the strongest driver in the first half or -- in the later part of 2014 and basically also in the first half of 2015 memory has been the leading driver for us although foundry picked up somewhat in Q1, bookings Q1 and some deliveries going into Q2. That it has been modest compared to the contribution of memory.

We don't expect that to change significantly in -- we don't expect that to change in a meaningful way in the second half of the year. Then the ALD specs which specs are important.

Yes, it really -- and that is not to way to get away with your question in an easy way. But all specs are important.

Of course technical specs are important and throughput specs are also important because technical specs are important when you go to smaller geometries or to more challenging new layers which existing deposition technologies like CBD, PBD cannot accommodate anymore. But at the same time if the price tag to it is too high for a customer then they may try in whatever way to extend the existing deposition technology.

So in that respect throughput cost of ownership or all aspects of cost of ownership are important. So you cannot ignore any of those aspects.

At the same time of course when some ALD applications get a little bit more mature, then of course cost becomes even more important than when an application is initially inserted. So that is about the color I can give you.

So we cannot afford in the development of our equipment and also in the CIP of our equipment and the continuous improvement programs on our equipment to ignore any of those aspects. Okay, so that is on all of that part.

And then Peter can talk to --.

I can, yes. And the question of course let me broaden a little bit because the share buyback -- we just finished the share buyback program of the 100 million as you have seen in our announcements.

We always have stated that excess cash and we have given indication that the excess cash should be an amount substantially above the €300 million and should be there for longer periods of time as we expect. That we will use excess cash one way or another for the benefits of our shareholders.

We have €360 million approximately of cash at the end of June. So we will review periodically what we have done in the past periods.

If this would lead on a certain moment to a substantial excess cash amount and then we will decide which way that we're going to use that for the benefits of our shareholders.

I have a follow up question on the potential threat to your market share in ALD maybe. By how much are you already specced in on the potential 10 nanometer process?

Or in other words, what is the threat of your competitors actually eating into your market share at 10 nanometer or is actually the threat more even below that node? And my second question is, are you willing to maybe specify a little bit more the breakdown in ALD between what I call your two main applications as in the metal gate and the specialty defined double patterning versus the other applications?

Is that already a sizable business within the ALD portfolio or is it still small?

Okay, to start with, Philip, with your last question. I think currently patterning and the FinFET high-k metal gate area, those combined are in the lead compared to the other applications at this moment in time.

But, yes, so that is the answer to that. And we just don't want to provide more color on that to the market as a whole.

Then, yes, your question on 10 nanometer. I don't know how much more color I can give than we gave in earlier questions which were basically similar to your question.

Because in my perception the same question was asked earlier in the call. Again, the 10 nanometer logic foundry insertion is currently expected by the market to happen in the course of 2016.

Yes, so customers are expected to be pretty far along in let's say developing their device structure and associated process structure and required equipment to do so. But it does not mean that they may make some changes in the last phase of their development.

But I don't expect them to change let's say 50% or even 40% of their current path in the last six months of their development program. And based on that we make our own assessment of where we think we're in terms of our position in preparing for those customers to ramp.

That is the best way to answer your question, Philip.

The question is on your SAM expansion which you expect over the next three to four years. When we think about the logic and foundry market, that covers up to 2018, your three to four year guide.

So when equipment suppliers are expecting 7 nanometer investment to start for the logic and foundry guys which as I understand has a step up in spacer defined multi-patterning. Can you give us a guide as to how much of your incremental $600 million in addressable market is as a proportion for multi-patterning applications versus other applications you are talking about?

Yes, we have not -- I understand that you would like us to provide more color and we're continuously investigating ourselves how can we guide you or our audience in a better way going forward. And we will take this input also, this question also to see how we can improve our overviews in the future.

The only thing we can say at this moment in time that it is a very significant contributor to the doubling of the market from 2014 to three or four years out to $1.2 billion and beyond. It is a very significant contributor -- the patterning market.

But we're still -- in that projection it is partly because, yes, we don't want to share all the details immediately, but it is partly also that we're learning as we speak based on customer feedbacks about new applications that our customers are introducing and then trying ourselves to size those opportunities. So we're also learning as we speak with you and others on this call.

Just one more on 3D NAND. What point layer count does your addressable market for ALD and 3D NAND get to similar levels as you saw in planar NAND?

Are we looking at higher than 48 layers or is 48 layers pretty similar to what you are seeing in planar?

Where there are two elements to the question. One is the patterning part, the patterning part which is now of course a very important component for -- in planar NAND for us.

That part is going to reduce significantly going from planar to 3D NAND. And so what we're mainly working on now with our customers are completely different applications, completely different applications that in -- that are in going from 32 to higher, bigger stacks will push our customers to use more ALD layers or other applications beyond the ALD that we could serve the market on.

But that's what we're looking at in our dialogue with our customers. And trying to learn how we can grow our share of wallet in that part of the market.

And just finally, I guess when you look at 10 nanometer for the foundry and logic guys, do you see that as a bigger opportunity than what you have seen at 16 and 14 nanometer? Because when I look at 16 and 14 there's been a pretty sharp and pretty quick transition, I mean two quarters probably orders.

So I am just wondering how we see the 10 nanometer progression and maybe the length of that.

Yes, I think it is a very good question. The short answer is the step from 16, 14 to 10 we view as a bigger opportunity than the step from 20 to 16.

Okay just on that then. So was 16, 14 -- has that played out how you are expecting?

Or were you expecting that to be a longer --?

In terms of number of applications there were no surprises at all. In terms of volume -- well, in terms of volume I think -- I cannot speak for the whole industry but I think the industry as a whole might have expected initially -- a year ago might have initially expected that to be maybe a slight slightly better, slightly better.

But there were not too big surprises there. But we always viewed already a year ago we know that the step from even before 14, 16 started we knew that the step towards 10, we believed that to be a bigger opportunity for us than the step from 20 to 16, 14.

Actually a couple of follow-ups. Coming back on the discussion on DRAM where you said that it is still unclear when spending only 1X now may start.

But given that 20 nanometer has a cost benefit over the older node, one would assume the DRAM makers to continue these conversions. So what do you expect?

Where do you think your customers are in terms of converting their capacity to 20 nanometer? It seems there are still a lot of upside there.

So will be interesting to hear your thoughts on that. And then maybe looking at 2016, if we were to see a shift in the mix from PEALD for specially defined dual patterning to new ALD applications for 3D NAND, could that potentially affect your gross margin at least temporarily as there will be a higher share of new clients, new applications and you may have to go through some learning curve effect?

Okay, so initially on DRAM 20, the only thing we can say is that the cycles of DRAM really have been between different nodes have been pretty consistent so far, the timing between the different nodes and we don't see any indication why this time that would be different. And for the second half of the year our current visibility is that the investments in DRAM for 2X are expected to be lower than for the first half.

And that is the only thing we can say, Peter, about it. There is not much more to add.

So I think significant meaningful investments in DRAM should come from investments in the next DRAM node, that is most -- the most likely scenario with our current visibility. Then on the PEALD gross margins.

Yes, let me give that some color. What we have indicated in earlier cases is that we expect low- to mid-40s and an upturn.

That takes into account also the [indiscernible] tools. I tried to give every quarter as well in our press announcement as in these calls some color on what is going on with the mix.

So it could lead on a certain moment when there would be a lot of new tools and a lot of new customers. It could put some pressure on the gross margin.

But that you have seen from the previous period, I think from previous periods. I think we're rather fine with confirming what we have said earlier, that we -- even when that will be the case, that we would stick to the low- to mid-40s gross margin guidance that we have given in earlier times.

Maybe one final question then. At the SEMICON West presentations of some of your peers and competitors they were talking about growth opportunities in services.

Do you consider that to be a growth opportunity for ASM as well or will it just grow with the installed base?

Indeed if you grow your installed base then our services grows. We don't -- we would not rely on a position now to present services as a meaningful other growth engine beyond the growth of the installed base at this moment in time.

We're not ready to make similar statements.

But do you think that may put you at a disadvantage compared with some of your competitors also in selling equipment?

I don't think so.

Okay, thank you very much for the lively discussion today with all of you. Peter and I, we trust that we have provided you some more color on the status of the company at this moment in time.

And thank you very much for your continued interest in the company and please feel free to follow up with us through Victor or the team here in case you have any follow-up questions. Thank you again and have a good day.

Bye, bye.