Cymer Q3 1998 Conference Call, Part 2 of 3 - Q & A
Jay Deana, Morgan Stanley
Deana: Hi Bill. I didn't quite catch all the numbers that you gave up front. You said something about laser unit declines and litho shipments and lithos installed or something. Can you go through those numbers again?
Angus: Sure Jay. The lithography unit shipments for the quarter decreased 25% from the 2nd quarter. And the install rate at the chipmakers decreased 40%.
Deana: Ok, and that's litho shipments from all suppliers, right?
Angus: Yes.
Deana: Ok. And then you gave two different numbers for average selling prices. One hedged and one not hedged. Can you clarify that again?
Angus: Yeah. The easiest way to look at - since we do hedge our contracts - is to take out the effect of, from revenue, of the currency fluctuations, since they are hedged, and look at them on a pure basis. And so therefore, on an adjusted basis, eliminating that, ASPs have risen 13% year-to-date from the beginning of the year.
Deana: And what were they adjusted again in the quarter?
Angus: Adjusted in the quarter at the end of Q3, 469,000 was the average ASP, as opposed in Q2 of 433.
Deana: Ok. Now, Bob, in terms of your comments, did you imply there that you thought that 99 would a show sequential quarterly increases starting in the 1st quarter?
Akins: No, and I didn't say sequential, Jay. I wanted to stress the volatility. We have reasons to believe that principally due to product mix changes, that the adoption of our 5010 model specifically, and perhaps some Orion sales beneath that, to provide some upticks in our quarterly revenue going forward. But, it's too difficult at this particular point in time to say that that's going to be a sequential event or not …
Angus: Or when exactly it might even kick in.
Akins: Exactly.
Deana: So just so I'm totally clear, are you referring to at some point you have a quarter that's higher than the one before, or are you talking about year-to-year growth rates here?
Akins: You mean a quarter which is higher than the year before, or just the quarter before that?
Deana: Yeah, I'm just trying to clarify exactly what you said in that opening comment.
Akins: I'm certainly not ruling out the possibility of increases on a quarter-to-quarter basis. That may happen, but again the guidance we are trying to give here is that the potential for revenue increase is certainly there, principally from a product mix standpoint. Not as an indication that the industry has rebounded and is demanding large numbers of these tools, but because they are adopting higher performance, higher ASP products than (unintelligible), and that rebound, that increase we see happening in 1999, but at this particular point in time, it's not possible to talk about a particular starting point on a quarterly basis. We are seeing increasing of orders as we speak. So, as we said in the conference call.
Deana: Right. And that increase in orders is based on the trough in the 3rd quarter, which was pretty low, right?
Akins: That's correct.
Angus: Yeah, it's up compared to what?
Akins: It is an uptick in orders, and because of the percentage of our product mix - we're seeing a relatively aggressive transition from this 5010 as I indicated, up to 50% of our KrFl products shipped this quarter, versus 20% last quarter. So it's a combination of those two events.
Deana: Right. And is that units or revenues you're referring to when you say 50% vs. 20%?
Akins: Units, Jay. That would translate to ASPs even higher than that from a revenue standpoint.
Deana: Right. Thank you.
Robert Mayer of DLJ
Mayer: Good afternoon. Could you give us an update as to competitive position vis-à-vis Lambda Physic, Komatsu, others? I asked at a recent industry trade conference, and no one else had suggested that there were any other production units in the field. Could you give us your view and your update?
Akins: Yes, Robert. There's not a tremendous of new or significant information for us to report at this time on the competition. Things remain about the same as we last discussed in the last quarterly conference call. I think that there are a few things of note, the first is that by our own best tracking, we feel that the competition, principally Komatsu and Lambda Physic, between them probably have about something about 10 or less than 10 lasers that are at chipmakers. Those are primarily for evaluation, but some that can be used in production are seeing early production use. Those fights are principally being driven by chipmakers who simply have a two-source policy and must drive to evaluate other lasers than Cymer. So those are sales opportunities that really Cymer can't access because of the policies of those companies. Since our last talk, there has been an ArFl international conference held in Austria, focusing on ArFl and its development for product production in the years to come. We attended that conference, presented a few papers. We're talking about ArFl levels of performance in power and pulse repetition rate, which are aggressive. We have one-kilohertz ArFl technology today that we ship. We are working on two-kilohertz and three-kilohertz as we announced in that conference. I think that our competition is still working to achieve adequate one-kilohertz performance there, so we certainly are extending our R&D lead in that new technology area going forward. Also for completeness, I should point out that at the conference Ushio also announced that they were developing an ArFl excimer laser. Ushio of course being a Japanese company that has been supplying (sounds like 'murky balls') to the lithography industry. They decided to enter this marketplace as well. Information there is very preliminary. They have some laser specifications to announce. Obviously they have a lot of work to do to get to the point where they would be at the same point of development of either a Komatsu or a Lambda Physic.
Mayer: Ok. One other question. Could you tell us if you've seen any market share shift or in your view have sensed any market share shift given perhaps more aggressive price tactics by Japanese stepper manufacturers as the market contracts, and do you see any reversal of that if there was that given yen-dollar trends? Where do you see market shares coming and going from?
Akins: I think we can't make any comments on that issue at this particular point in time, Robert. You're probably as aware as we are of the activities going on there. I think it would be inappropriate for us to make some comments about that, especially when based upon us interpreting what our customers may or may not be doing relative to their position at chipmakers. I'll just reiterate as Bill had referred to in the conference call earlier, that ASML now accounts for - at least in Q3 - for 40% of our sales…
Angus: No, that's year to date.
Akins: I'm sorry, year to date, for the first nine months of the year. Nikon 28%, Canon 21%, and SVGL 6%. And I think that it's most responsible if we just continue to represent that fraction of those customers volume, Cymer volume.
Mayer: Ok, thank you.
Bret Hodess, Montgomery Securities
Hodess: Good afternoon. A couple of questions. Can you give us an update on the transition to scanners, with the 5010 and the 6000 targeted to performance boosts I guess on those products? What do you see in terms of the different your different customers being ready to ship there and what do you see in terms of the end customers moving toward those?
Akins: Bret, we still remain very bullish on scanners. You're very well attuned to the advantages of using scanners over steppers as these fields get very large combined with the higher resolution involved. And I think that some of the original design concepts for scanners to be able to maintain not only high resolution, but also larger depth of focus and process latitude because of the real-time on-the-fly adjusting they can do. It's certainly turning out to be the case in practice. So these tools are rapidly emerging to be the upcoming tool of choice. The difficulty of course has always been that it's a much more complex system, requires much higher stage speed, requires much more performance from the laser, etc., etc. And those issues, though difficult, are as you can imagine the major subject of R&D and new development for our customers on their new programs. Also, the way we see things playing out, the 5010, which I had referred to as the laser which also offers good pulse energy stability for the first generation of scanners, is the tool which is envisioned to ship or is currently shipping on the most advanced scanners that are shipping today, and in the short term. The increased emphasis on the 5010 is in part driven by that. The 6000 series, the old Orion as we had code-named it in the company, is a product which operates at much higher levels of performance, but will be attached to a completely new generation of scanner. Again, of much higher performance and throughput. So we are shipping Orions at the end of this year and next year. We expect to ship those in relatively small quantity, and for the majority of those to be used for the whole integration process, of a new laser meets a new scanner, to deal with the optical, electrical, software interface issues between the two. Similar to what we saw some years ago as we shipped the first 5000 series lasers. So the 6000 will see increasing potential late next year and into the year 2000, and will be the principal light source for the 2nd generation of scanner, which will be introduced sometime mid to late 1999. Also of course, the thing that all of us realize that as the industry pushes to 0.18 micron, that becomes more doable in scanner format rather than stepper for the purpose of maximizing manufacturing latitude. So that is a major factor also driving the conversion to scanners. And I think that if not just for achieving the resolution with depth of focus, but also for the accuracy of overlay and alignment across the whole field needed at 0.18 micron geometries.
Hodess: Bob, do you see in the slowdown in the industry, do you see your customers, is it slowing down or speeding up do you think their move to the scanners?
Akins: I think that there's, driven by chipmakers there's a increased, uh I think that the number of scanners shipping over the next 12 months or so will definitely be seeing an increase. And some of that is being driven of course by chipmaker demand. Again, as chipmakers move to 0.18 micron, and their need for higher accuracy of alignment and overlay, they in turn are driving the stepper companies to increase their focus on the new scanner technology.
Hodess: So you, just to, your comment about starting to feel that there'll be an uptick in revenues on a quarterly basis next year, is that fairly tied do you think to the adoption of the scanners and the fact that the chip companies will basically start to have to proliferate some of this technology to get to .18?
Akins: Yeah there certainly is a component of that, Bret. Because again with the first generation of scanners being powered by the 5010, that plays right into what we had described as this demand for shipping to a new product in Cymer, of higher performance. So the demand, the shifting to scanner demand is also a shifting to 5010 demand, and then ultimately to 6000 series demand. Now that's not the only demand. There can be reasons that the 5010's performance can be fruitful for steppers as well, because one of the improvements in the 5010 is reduced bandwidth, which is a big help for enabling that the resolution of the high numerical aperture steppers as well. But certainly that's one of the components.
Hodess: And a final question. ON the recent uptick in orders that you're seeing, or that you know, or stop dropping in orders, or however you want to characterize it, is it on the unit side or is it picking up a bit on the spares side as well? You know, indicating that perhaps people are starting to use the lasers a bit more?
Akins: To begin with, the pickup is in part due to, again, the shift to the new products and to the higher ASP products. It's also indicative of the fact that, I mean, everyone is reading about the chipmakers pushing extremely aggressively to the small dimensions, not just for the purposes of doing so, but for reducing their costs as they move to more and more aggressive geometry shrinks. So there's a number of pieces in this channel. On the spare parts side, we have, the data goes up and down on a month-to-month basis, as far as the usage and utilization of these tools. Our overall trend is that the utilization of tools is continuing to increase, but not at an extremely rapid rate. We saw major increases earlier in the year and last year, now the increase is much more slow. That's indicative of the fact that there were huge increases in utilization earlier in the year, and that's going off now as these tools see a higher and higher uptime in utilization.
Hodess: And that's why in the 4th quarter the slower utilization growth rate in the 4th quarter? You think the spares and service business might drop off a little bit?
Akins: That's right. Your question was specific to the utilization of spare parts for the purposes, uh tied to the utilization of the tools. Also remember that earlier in the year, as we had discussed, some of the shipments of spare parts went to filling the inventory channels for our direct customers as well. That's kind of a one-time event, as you fill those channels with 5000 series type parts. So that was artificially inflating the instantaneous run rate of our sale of spare parts as well.
Hodess: Good. Thank you.
Leonard Sanders, Needham and Company
Sanders: I wanted to follow up on the same thought. In the beginning of your conference call, you talked about reducing the cost of operating you lasers significantly. Can you go over that in a little more detail, and is that another reason why your service and spares business might be down? And could you also comment on how that might affect your inventory - there's some comments in your press release about potential inventory writedowns.
Akins: Ok, the cost of operation, what we really wanted to announce today was more or less successful implementation of a particular phase of what has been an ongoing program here in Cymer to study the cost of operation of our lasers, and to make either technology developments or engineering improvements to principally increase the lifetime of the key and expensive modules in the laser. And those programs have certainly yielded very successful results impacting most of the key 4 core technology modules in the laser, the discharge chamber, solidstate pulse power, the (unintelligible) package, and the wave meter. I don't want to say too much about exactly what the technology breakthroughs are, because we are in a situation where we have a number of new patents that have been filed in this area. I'll just that in general, as you can imagine, these improvements lie in the area of optical enhancements, electrical enhancements, and some specific improvements especially to the discharge chamber, the discharge chamber being of course the single most expensive module in the laser. The improvements manifest themselves in such a way as I've described, that a 5000 series laser that were to be upgraded with these new longer lifetime components would not see a change in performance specification, but in lifetimes of the core modules, the key consumables. And it can be a very significant increase in lifetimes, resulting in a 45% reduction of the cost of operation. Which of course is a significant amount of money and of significant interest, both to our direct customers and especially chipmakers. So we are currently in discussion with our direct customers and in some cases with some chipmakers about what those, what such a upgrade could do for their cost of operation, and it's receiving a very significant amount of interest. The inventories - uh what was your question on inventories specifically, was it spare parts inventory, was it…
Sanders: It's inventory in general…
Akins: Ok, well I think that our inventory on spare parts is going to continue to be used. We're going to have significant ongoing demand for 5000-related inventory of spare parts. The, in Cymer we're seeing a very rapid transition, where our direct customers and chipmakers want the 5010 new laser, as opposed to the 5000-series lasers. Imagine the 5010 takes advantage of these cost of operation improvements inherently. But also combines it with higher levels of performance specifications. So part of what we discussed here today in the loss for the 4th quarter is coming to terms with some of the inventory writedowns that may be necessary, to impact uh to implement an either the change in the 5000 series to a different level of configuration of lower cost of operation and/or the shift from the 5000 series laser to the 5010 laser. The attractiveness of these new products is more significant than we had anticipated, and therefore the conversion, uh the demand in these new products is occurring faster than we had envisioned. In the 3rd quarter in fact, we've had significant change of configuration of products that we ship due to this changing demand. So it's a very rapidly-occurring event, and we're still in the process of analyzing what that will mean for us, but obviously we're providing this guidance to the street about the losses for next quarter to include the impact of some of those rapid changes.
Sanders: Ok, thanks.
David Wu, ABM Emerault(?)
Wu: Number one, on the technology platform, a number of the major logic as well as DRAM manufacturers are talking about a big jump into .18 micron process early- to mid-1999. I was wondering from a technology standpoint, if they're going to go to steppers, would a natural inclination be going to the second generation steppers with the 6000, or would it be more like the 5010? Second point is about, could you talk a little bit about the levels of inventory in the hands of your EOMs? The direct customers. How much of that inventory is still out there, and have there been any reductions so that you see more sell-through for the 5010, and as far as your spare parts inventory is concerned, is, how should I look at your inventory at the end of September? How big a percentage of that inventory is actually spare parts that might be subject to writedowns. Lastly, just in case that the worst ever happens, on the 43 million dollar breakeven level, if you have to do it one more time, where could you possibly cut?
Angus: Let me take the inventory valuation issue. If we knew specifically, and could quantify at this point in time, the specific amount of exposure we have, we would have had to have taken those charges. The point is, due to the lack of visibility usage rates that we have going into 1999, we are still trying to assess that, plus trying to get a final determination of what our product mix is. We know there are questions there. There is a probability that we are going to have to have some adjustment on the inventory. And we've tried to swag as best we can, and that's all it can be. If could have definitively have defined it, we would most appropriately taken the charges. So that's the only way that I can really answer the inventory valuation issue. Bob, do you want to follow up on the other aspects?
Akins: Let's start with your question on the platform. We certainly agree with your statement about various logic and DRAM manufacturers. (Unintelliglble) to push forward with scanner adoption mid-1999. And while not being, I can't be overly candid on who's doing what, but I will say that the 5010 model will constitute the majority of the, as the light source on those scanner tools, on that particular generation for 1999. For one customer with a high-speed scanner, uh uh stepper capable of making exposures at higher speed in shorter periods of time, may incorporate the new 6000 series laser. Let me just say as a matter of background, you may remember that we've discussed in various past conference calls and conferences that the 6000 was originally conceived as a laser only for scanners. The idea being that the throughput, that the higher power and rep rate would improve the throughput of scanners significantly, and make an attractive laser for that, for second-generation scanners. But since that time, at least one customer has also pointed out that that laser, with its higher average power could make exposures in a DUV stepper in a shorter period of time, and see a significant increase in throughput even for steppers. So, we may see a dual use for high-end steppers as well as scanners. On the inventory at OEMs question, we estimate that the inventory at our OEMs is approximately 360 lasers. Roughly 60 of those are lasers used for tooling purposes, leaving the actual KrFl working inventory at about 300 or so. And then lastly, on the 42 million dollar fund rate…
Angus: Let me talk to that. We are currently in the middle of our 1999 planning process, as you might well imagine, and at this point in time we have a wide range of scenarios as far as the top line are concerned, going into 1999. Obviously, as part of this process, we will have to put a stake in the ground and draw some assumptions. It would inappropriate for me to at this point guess where they will be and what the ultimate prioritization on our spending would be until the planning process has run its course. But, part and parcel with that is the overall process improvement strategy that we have had. Remember, we have ramped up very very quickly as we came out of the box here in 1996 and 97. A lot of inefficiencies were built in and we have been trying to work systematically to reduce those. And we'll continue to do so as we go forward. However, the one thing to keep in mind is that Cymer has consistently never made decision that we felt in any way would sacrifice our future technology leadership position.
Wu: If you have 300 KrFl lasers that are still in the OEM channels, what is a more normal level of inventory?
Angus: That's going to depend on, to a certain extent, on how our direct customers view their business, and what the lead time or integration time is that they require. There had been a tendency at some of the integrators to reduce that lead time down. At others, it's still rather extensive. So, it's a little hard to say, but I think that we would be more comfortable if that number were down in the area of about 200 units at this point in time, rather than where it is today. |
