Transcript of Conference Call Part 2 - Question and Answer Session
Brett Hodess, Montgomery Securities
Hodess: Bob, your comment that more systems were installed this quarter than in the previous six months: can you give us an idea how many of the lasers you think you've shipped since the ramp-up has begun that are actually at customers' at this time, uh in operation?
Akins: Uh, yes Brett. About, uh as of the end of the third quarter, approximately 160 of our lasers have been delivered for systems that are bound for production.
Hodess: So a little over half of the shipments thus far since last Fall when you started to ramp-up?
Akins: I'd have to look at those exact numbers to give you a proximal [may be 'approximate', that would make more sense, but it sounded like 'proximal'] basis. I think it's probably less than a half.
Hodess: Ok
Akins: But the number is certainly growing.
Hodess: Ok, so you're just getting to the point where you can really, uh, really get a lot of feedback, actually. In terms of high volume in the field?
Akins: Yes, although I'll point out that still as we speak, to the best of our knowledge Samsung is still unique in its really heavy usage of these tools for full production. But we certainly expect in Q1 and Q2 of next year to see many more of these systems to be used in high-duty-cycle usage for full production.
Hodess: Ok. Then my second question has to do with now that you're running at a high-volume run rate, and, have your main infrastructure in place for manufacturing and have a lot of the big five issues if not under control understood, at least, at this point, can you give us some feel for what might happen with the gross margins over the next few quarters?
Angus: Yeah Brett, this is Bill. Yeah, we are expecting once we get out of this year, in particular, and start getting into next year, we are looking for a gradual steady improvement in gross margins, as we can start to become much more efficient in our manufacturing operations. I still don't want, uh, I'm a little hesitant in characterizing fourth quarter at this point because I'm sure there's still some cleanup from the move and everything else of getting settled in the new facility that could potentially still impact us here in the fourth quarter. But definitely during 1998 we're expecting steady improvement there towards our targets.
Hodess: The final question I had has to do with the gas chambers, the refill of the uh, you know, the recharge and refill of the gas chambers. Are you starting to see that ramp-up now that you've got significant unit volumes in the field, and how does that business look?
Akins: No, actually, again, the lasers are being used in some aggressive duty cycle, but not compared to when all the chipmakers are in full production. I'll also point out that as part of the big five improvement program, chambers are also being attacked in that program so that a lot of those exchanges that we might otherwise see in a different way are being rolled up as part of the big five program. So I think that we'll really see a significant impact of that on our sales of chambers, Brett, starting at the end of the second uh, in the first quarter or in the second quarter of next year.
Hodess: Thank you.
Akins: Thank you.
Robert Mayer [My guess at spelling. Pronounced 'mayor'] of DLJ
Mayer: Yes, hi, I have a few questions also. Congratulations on the nice quarter.
Akins and Angus: Thank you Robert.
Mayer: You had suggested that over the next two quarters you see continued sequential increases. Do you have any initial take on the two quarters beyond that?
Angus: No, uh, Robert, because we're basically managing our customers to a six-month backlog [baby voice in background from somewhere]. We do have some visibility past that point, but are reluctant to characterize it. So far we don't see anything that leads us to believe that would be anything but, but we're just reluctant to commit ourselves at this point, given the fact we don't have the orders.
Mayer: Ok. Another question. On the recent conference call, Jim Bagly of Lamb had suggested that some customers had suggested that some customers were pulling in orders to more aggressively do die shrinks on 16 meg DRAM and beyond. Are you seeing any changes in demand from your customers, are your customers indicating any pull-ins or push-outs or other alterations of schedule and any impact on your customers' ability, uh technical issues that might impact their ability to pull-in or push-out?
Akins: Actually Robert, we, uh I had a conversation with a customer this morning that described the same phenomenon, that basically because of the lack of making money in the 16 megabit or even initial versions of 64 megabytes [that's what he said. Think he meant 64 megabits] that there was a race on to the supershrink versions of the 64 megabit, which was accelerating the whole desire for getting quarter micron capability online. The short, uh there's a shorter cycle in the acceptance of Deep-UV in the factory, and that's being promoted by, uh there's a desire to shrink that period of time to [unintelligible. Could be 'smaller and smaller'] scales to be able to bring the new equipment online even quicker to realize those supershrink versions.
Mayer: Uh huh. Is there any sort of a recent update as to the stepper manufacturer's abilities to ramp-up? There were some issues regarding some technical difficulties of Nikon?
Akins: Yeah. Uh I think that, again, we are not in the business of adding to any speculation there on any particular customer, and my response to that Robert has to be our basic response, that in general these are very difficult tools to manufacture and we expect that all manufacturers of steppers and scanners are facing significant manufacturing challenge in ramping-up so quickly. To the best of our knowledge, those problems fit under the category of those kinds of fits and starts to be expected when you are producing a brand-new type of tool like this. We don't see any fundamental technical problems.
Mayer: Ok, and one last question. There were some issues or concerns [baby in background] regarding inventory at some of your customers. Are you doing anything to uh, or have you instituted any new changes to work on that issue, or have you been in contact with customers recently to see what the exposure is there?
Akins: Well a couple of quarters ago we mentioned the fact that we had actually - and we had on our road show at one time - a bullet on one of our slides that read customer inventory levels, because we were doing back-of-the-envelope calculations to convince ourselves that inventory levels weren't reaching the excess or excessive levels. We are, as we speak, getting more sophisticated understanding of our customers' manufacturing processes, and trying to do more sophisticated modeling of that. And using those models as tools for management of the whole business with our customers. So yes, we're trying to get more sophisticated and try to understand the basic drivers, both in the factories of our customers and by their, by the demand from their customers the chipmakers.
Mayer: Ok, thanks. Congratulations again.
Akins: Thank you very much, Robert.
Jay Deanha[?] Morgan Stanley and Company
Deanha: Thanks. Good afternoon guys.
Akins: Hi Jay.
Deanha: My question relates to lead times for lasers going into stepper manufacturing, and I'm going to ask you two questions, and what I'm trying to calculate is the ratio of Deep-UV steppers to lasers for 1998. So the first part of the question is can you talk about what kind of lead times you're seeing now in general going into steppers and scanners and what kind of rate of decline you see in that over the next three or four quarters.
Akins: As we had discussed in the past, Jay, I think it characterized the integration times pretty exactly when we said it took about six months to get a laser into a stepper at one of our direct customers prior to delivery. I don't have exact figures in my head right now, but I would say that the integration times, depending upon the customers, are probably in the four to five month range right now for the standard stepper design. For the more sophisticated and newer stepper models, that's probably closer to five or six months. And scanners of course are a brand-new tool, and a number of additional more-demanding challenges optically for the laser on the stepper itself, are probably in the six month area themselves. Going forward we see that the last generation stepper will probably become faster, coming down to may only a quarter or so of integration time. But at the same time we'll have a mix with that with the more sophisticated tools that will take longer integration. It's difficult for us to come to an exact forecast right now of what the integration time will be. But certainly it's safe to assume that the integration of lasers into steppers will be significantly faster that those into scanners.
Deanha: Ok, and if you were going to characterize a ratio of lasers to steppers for 1998, would you be using something in the one point five area or would you be using something smaller than that?
Angus: Jay, this is Bill. We haven't done that analysis. I'd have to take that under advisement and get back to you. I just can't respond off the top of my head.
Deanha: Ok, that's fine, I was just kind of trying to calculate the size of the laser market next year given the assumption that the Deep-UV market would be kind of in line with Dataquest numbers which I guess are what, around the 500 level or something like that?
Angus: Uh, yeah. A [sounds like 'buck'] or so, and obviously it's going to be greater than the number of steppers and scanners delivered. But, our model and its sophistication and taking into account the blend of integration times of anywhere from three or four months to five or six months is not sophisticated enough for us to really have a good coefficient of that. It's probably, uh obviously it's between one and one point five, exactly where we can't estimate right now.
Deanha: Right. And Bill, what tax rate should we be using next year?
Angus: Uhhh. For next year, 37% right now.
Deanha: Great. Thanks.
Leonard Sanders, Needham and Company
Sanders: Hi. I was curious if there are any kind of developments at Seiko, as far as the quality of their laser versus the quality that you've had in San Diego, and also anything on the litigation front there.
Akins: No. Basically, as we have previously discussed, we took our time with Seiko, diverting all of their production for quite some period of time to San Diego. Those lasers were then completely retested and inspected in San Diego and used for internal applications in manufacturing, for tooling, or in R&D and so-on and so-forth. So, by the time we felt comfortable in actually shipping systems directly from Seiko's factory to our customers in Japan, we had sorted through those issues and we feel highly confident that the lasers that are assembled and tested by Seiko are the same quality, consistency, and performance as those from San Diego. We routinely have people from our factory in San Diego in the factory in Seiko to assure that and to make sure that the processes used or any improvements in processes that have been contemplated in San Diego are introduced at exactly the same time in exactly the same way in our facility in Seiko. And what was your second question again?
Sanders: Anything on the litigation front there?
Angus: You mean with Komatsu?
Sanders: Yeah, with Komatsu.
Angus: Uh yeah. Just a reminder - it's not litigation yet. [brief chuckle]
Sanders: Yeah, the threat of litigation.
Angus: The threat of litigation, right. We continue to negotiate and talk with Komatsu in hopes of in fact avoiding any legal entanglement of that nature, and there's nothing new to report.
Sanders: Ok. And then, just to sort of follow on to the comments about process changes, can you talk about how that's when you find a process improvement how you implement that and then how you go about in the field of implementing that?
Akins: Well we certainly move cautiously. There were the days of Cymer when we were selling ten lasers a year when we were very cavalier about cutting a new process and changes into lasers. But obviously new processes have to be extremely well-characterized before one can cut them into production. So we typically, once a new improvement is has been proposed and feasibility has been demonstrated, then we usually cut that in in a controlled way for lasers that will be used for internal consumption. And try to get enough statistics that one can make a statistically valid decision as to the efficacy and consistency of that change. Of course at the same time that may mean that there are similar issues to be dealt with at critical suppliers, so we launch both an internal and external program at the same time. Many of the changes that we make will be put into new models only. In the case of the big five, we deemed it prudent that we take some of those changes and actually make them available into all the lasers in the field of the 5000 series that we've shipped, as I mentioned before to help proactively ensure the correct operation of those lasers and the correct cost of ownership of those lasers in the full production mode, which is obviously to our benefit.
Sanders: Ok. Thanks.
Akins: Thank you, Lenny.
Gunnar Miller, Paine Webber.
Miller: I was over in Asia about two weeks ago, and there were some increasing indications from Nikon, Canon, as well as ASM Lithography and SVG that they are taking a pretty close look at Lambda Physic and Komatsu. There was a recent announcement in Japan that Ushio is also entering this market. I wonder if you could frame out for us what you'd view as your sort of worst-case market share next year and in 1999.
Akins: Well, let me describe our view of the competitive situation in perhaps a little bit more detail than I did already. I don't know that it would be our job to speculate on the types of market share that those companies may or may not be in a position to obtain over time. I think we have time and time again described our competitors Lambda Physic and Komatsu as uh, Lambda as our closest technical competitor, our closest technology competitor. Why? Because Lambda does make its laser performance and basic technology more or less available through various public channels, as does Cymer, as does - as do most equipment suppliers. And of course Lambda had the most significant market share prior to Cymer's inception. They have lost market share principally as Cymer has gained market share. Komatsu is still a black hole of information as far as we are concerned from a competitive front. Again, we get no access to real [stresses 'real'] information about their product's performance or its reliability, and that information continues to be very very closely guarded by those companies who talk or use uh, talk to Komatsu or use that company's product for evaluation in their factories. So it's impossible for us to make a real statement there. Of course, Lambda Physic and Komatsu claim on paper - and have for years now - that their performance is equal to or in most cases in excess of the specifications of Cymer's product, and I will have to observe that certainly that is a position that a company attempting to gain market share most assuredly must take. At some point in time, we completely agree, that those companies will develop lasers that are equivalent in performance to a product that we may have already introduced. And of course ongoing competitiveness depends on our ability to manage our customers, manage our technology roadmap, manage our new product introduction, so-on and so-forth. This is not a new situation for us, as we have been fighting both of these companies as competitors for approximately ten years now. Ushio is a different kind of competitor. You know they all, that they manufacture the mercury bulb used in steppers. Ushio has had a press release recently discussing their work on a solid-state alternative light source technology - alternative to excimer laser light source technology. This is not a new alternative. It was proposed oh, maybe seven years ago or so by, principally by Sony. Sony developed this technology. It's a GAG laser, frequency-multiplied to the Deep-UV area. There are significant limitations to average power in these machines. There are significant problems with spatial coherence which provides [here a word that sounds like 'speclum'] noise in the wafer. You have to gang these lasers together to produce the types of average power required. I ask anyone interested to check into the literature - which has been in place for years now - about the real competitiveness issue and cost of ownership associated with these lasers. We actually hired a graduate student from UC Berkeley about five years ago who did his thesis on comparing solid-state laser alternatives to excimer lasers for Deep-UV lithography. He worked in [here a word that sounds like Dote] Oldham's group, and we stay very abreast of these technologies, and at this point in time as in the [unintelligible] past five to seven years we feel that excimer lasers are the best solution to this problem.
[Long silence]
Akins: I'm finished.
Miller [waking-up apparently]: Thank you.
Ali Irani, Oppenheimer and Company
Irani: Thanks, my questions have already been answered.
Michael O'Brien, Soundview Financial Group
O'Brien: Hi. Question: if Komatsu and Lambda Physic do come out with competitive lasers, how would you model that into your gross margin improvement plan. Would you expect some pricing pressure from that?
Angus No.
Akins: No, and I'll just add to that, that in general, over our ten year history of competing with these companies, our laser uh, I guess I can't say always has been, but for the most part has usually been the most expensive laser offered.
O'Brien: Thank you.
Angus: I think that will wrap us up for today, and we thank everyone for listening in. |
