Axcelis Technologies Inc (ACLS) 2003 Q3 法說會逐字稿

Good afternoon and welcome, ladies and gentlemen, to the Axcelis Technologies third-quarter earnings call. At this, I would like to inform you that this conference is being recorded and that all participants are on a listen-only mode. At the request of the company, we will open the conference up for questions and answers after the presentation. I will now turn the conference over to Mark Namaroff.

Good afternoon. This is Mark Namaroff, Director of Investor Relations and business development for Axcelis Technologies and welcome to our conference call for third-quarter of 2003. I'm sure all of you have received a copy of our press release issued this afternoon announcing our third-quarter results. If not you can download the release via our website at www.Axcelis.com. Discussing our results today are Mike Luttati, Executive Vice President and Chief Operating Officer, and Stephen Bassett, our acting Chief Financial Officer. Unfortunately Mary Puma will not be able to join the call today as she has had a death in her family over the weekend. Our thoughts and prayers are with her and her family. Also joining us to answer your questions are Lynnette Fallon, Senior Vice President of Human Resources and Legal, and Don Palette, Vice President of Finance and controller. The prepared remarks will last for approximately 15 minutes after which there will be time for questions.

Playback service will be available via our website or by telephone at 1-800-428-6051, as described in our press release. Under the SEC Safe Harbor provisions, please note that comments made today about our expectations for future revenues, profits and other achievements are forward-looking statements made based on management's current expectations. We urge you to review our most recent forms 10-K and 10-Q on file with the SEC, particularly the exhibit entitled factors affecting future operating results. As you know, due to the risks inherent in our business which are described in detail in the exhibit, our actual results may differ materially from our current expectations. We do not assume any obligation to update these forward-looking statements. Before I turn it over to Mike to begin today's discussion, I would like to briefly comment on remarks that both Mike and Steve will make regarding our future performance and projections.

When we speak of worldwide revenue, we are referring to the aggregate revenues of Axcelis Technologies and those of SEN, our 50 percent owned, unconsolidated subsidiary in Japan. Please understand that we do not currently consolidate SEN's revenue under generally accepted accounting principles. We use the term Net revenues to mean Axcelis Technologies only revenues determined in accordance with GAAP. We provide data on worldwide revenues with SEN because we believe it is useful to investors. SEN's ion implant products are covered by license from us and therefore the combined sales of the two companies indicate the full market penetration of our technology. And now I would like to turn the discussion over to Mike Luttati.

Good afternoon and thank you for joining us today. There are three key take aways from our third quarter that I would like to focus you on during this call. First, we experienced a very strong bookings activity during the third quarter; a 29 percent increase in net orders combined with what we believe will be continued strong fourth-quarter order activity. It bodes well for Axcelis Technologies growth prospects into 2004. Second, the Company executed its business objectives during the quarter, scoring several major design wins and significantly reducing annual operating costs. These actions will contribute to Axcelis commitment to reach breakeven in the fourth quarter. Third, the Company revised its revenue recognition policy to comply with EITF 0021, which became effective July 1, 2003. I will review the first two points and Steve will cover the third.

So let's start by talking about our view of where we see the market headed. We continue to track about 30 fab projects worldwide, most of them 300 millimeter. Several of these customers have started placing orders for new equipment. And orders expected from several others could lead to continued strong bookings into the fourth quarter. For the first time during a cycle we're seeing real capacity buying continue to pick up at many different types of customers around the world. This is a positive sign. Japan continues to be very hot, as witnessed by SEN's near record orders and shipments. Heavy spending for 300 millimeter capacity and technology upgrades is what is driving the strength of the Japanese market. Our net order rate this quarter was up 29 percent, one of the highest order quarters we have had since going public almost 3.5 years ago. This order growth will translate into strong shipments for the next few quarters.

Even though the industry is not firing on all cylinders at this point, these signs lead us to believe that we are in the midst of an industry recovery. Given where we think this momentum will take us, we are very positive about the future. Next let's review our progress against a number of critical business goals. Especially in the areas of cost out and market share growth. During the third quarter, we finalized necessary decisions about how our business needed to be structured in order for us to be profitable through the cycle. While we executed these actions in the third quarter, the planning began early in the year, as we recognized that even with the pending recovery we needed to reevaluate our business model for the future. We felt that we had to do things differently and we took actions across a number of areas of the business to improve our overall operating leverage.

Since the peak of 2000, we have reduced our operating expenses by $45 million. Third quarter actions alone will save us 18.5 million annually. Over the last three years we have also reduced our headcount over 30 percent. This (indiscernible) structure will give us tremendous cash generating opportunities as we move into the upturn. On the market share front, we believe we continue to make progress in all of our product lines. During the quarter on a worldwide basis we totaled 12 new customer wins across all geographic regions. These are not just repeat buys but are wins for either new fab projects or for new product design wins. For example we are pleased with the progress we have made in integrating and selling the matrix dry strip product line. We received an order during the quarter from a large European ship manufacturer for 200 millimeter dual chamber strip tool, which we are now calling the rapid strip 220. This is evidence that the matrix technology coupled with Axcelis Technologies global infrastructure is a winning combination.

In RTP we received two orders from a new 300 millimeter customer in Japan. We now have a total of eight customers worldwide using our non lamp based summit platform for advanced 130, 90, and 65 nanometer silicide and implant annealing applications. As we have said before, customer penetrations are key to the adoption rate of this new technology. And the progress we have made will pay off when high-volume production occurs. In ion implantation, we continue to win high-volume production orders with our high current and high energy systems. In fact we believe that we continue to take share in the high current segment at customers around the world, and are gaining share in the medium current segment through SEN's activities in Japan.

Now I would like to turn the call over to Steve, who is going to talk to you about our third point, the accounting changes that we just adopted, and provide some further detail around our financial results.

Thanks, Mike. As we discussed in our press release distributed earlier this afternoon, Axcelis Technologies changed its revenue recognition policy to comply with the provisions of recently issued emerging issues task force number 0021, which became effective for revenue transactions occurring after June 30, 2003. This new accounting standard requires deferral of a portion of revenues derived from system sales until certain future deliverables are met. Prior to the change, Axcelis Technologies recognized 100 percent of revenue from system sales upon shipment.

Under the revised accounting method, net revenues for the quarter were 59 million on shipments of 63 million, compared to 87 million in our second quarter. Worldwide revenues again under the new accounting method were 79 million on shipments of 125 million up slightly from 123 million in the prior quarter. Our implant business accounted for 79 percent of revenues with our complementary products dry strip, rapid thermal processing, and photostabilization contributing approximately 21 percent. Year-to-date, the implant business represents approximately 76 percent of the total.

During the quarter, system sales of 200 millimeter products constituted 77 percent of total shipments and 300 millimeter products, 23 percent. For the year, 2003, we have forecast shipments of 300 millimeter products at 35 percent of the total. Service revenues for the quarter were 31 million, flat sequentially compared to Q2. On a geographic basis, systems shipments to Asia excluding SEN, accounted for 56 percent, with 26 percent to North America and 16 percent to Europe. Including SEN, approximately 83 percent of systems shipments went to fabs in Asia.

From a customer perspective, memory IDMs and foundries represented 56 percent of systems sales and larger IDMs and foundries made up 44 percent. System and service bookings of 94 million were at a record level, an increase of 29 percent over the second quarter with system bookings up 50 percent. Backlog at September 30, 2003 including deferred revenue was 81 million, up 77 percent from the beginning of the quarter. Based on the geographic location of the fab, 86 percent of system orders were from Asia, 12 percent from Europe, and 2 percent from North America.

On a product basis, approximately 48 percent of systems orders in the quarter were for 300 millimeter tools. We expect this trend to continue, and as Mike indicated, most of the projects we are following are 300 millimeter. SEN's business remains strong reflecting the strength of the Japanese consumer electronics market as worldwide orders increased to near peak levels at 162 million, up 15 percent from the second quarter. Gross margin at 23.6 percent was lower than expected. Our revised revenue recognition policy and adjustments to warranty expense for increased costs associated with 300 millimeter tools reduced margins by a combined 9.1 percent. We are forecasting margins to increase significantly in the fourth quarter as a result of expanded volume and cost savings that will be realized from the initiative taken in Q3.

Operating expenses, excluding restructuring costs for the quarter were 39 million, up 7 percent sequentially from Q2. Most of the increase relates to operating costs and integration expenses associated with the July acquisition of Matrix Integrated Systems. As Mike discussed, we restructured certain functions in our organization during the quarter and eliminated 200 permanent positions. The annual savings of $18.5 million from these actions will begin to be realized in the fourth quarter of 2003. In connection with the reorganization, we incurred restructuring expenses of 4.7 million for severance pay and other benefits associated with reductions in force.

The contribution from SEN for the quarter, royalties and Axcelis' 50 percent share of their net income was approximately $200,000. SEN shipments during the quarter were 62 million, of which 43 million was deferred as a result of the change in accounting for revenue recognition. This had the effect of reducing the SEN contribution by approximately 7.8 million. We expect the impact of the change in accounting to have a minimal effect on SEN contributions in the fourth quarter as revenue deferred from fourth quarter shipments will be offset by revenue recognized from amounts deferred at the end of the third quarter.

We reported a net loss for the quarter of 31.9 million or 32 cents per share. Approximately 10.3 million, 10 cents per share is attributable to the revised revenue recognition policy, and 4.7 million, 5 cents per share, relates to restructuring expenses. I also want to point out that we did not recognize any tax benefit associated with the loss in the quarter. We ended the quarter with cash and equivalents of 115 million. A cash outflow of 38 million was slightly higher than our forecasted 30 to 35 million, due principally to a $10 million increase in inventory levels in anticipation of higher fourth-quarter shipments.

Looking forward to the fourth quarter, we expect worldwide revenues to be in the range of 150 to 160 million on shipments of 155 to 165 million. Net revenues, excluding SEN, are projected at 88 to 93 million on shipments of 95 to 100 million. Gross margins in the fourth quarter are estimated in the mid 30 percent range. Research and development spending is forecasted to remain essentially flat, and SG&A expenses are expected to decline by as much as 10 percent as we start to benefit from the cost savings initiatives we instituted in Q3. We expect SEN's income and royalty contribution in the range of 6.5 to 7 million as the market in Japan remains strong.

At these projected revenue and shipment levels, we are forecasting results of operations at breakeven with positive cash flow. I will now turn it back to Mike for closing remarks, and then we will open it up for questions.

Thanks, Steve. I would like to finish with a couple of comments before we take your questions. During the quarter we have filed an appeal in the patent litigation case against Applied Materials. The cost of pursuing the appeal is negligible, and we expect to hear something from the court within 18 months. More importantly, I want to make sure that you know just how confident we are that things are improving for Axcelis. The fourth quarter holds significant improvement and promise. The 29 percent increase in third-quarter bookings, which is driving a 77 percent increase in backlog. A 50 percent increase in revenues. A breakeven bottom-line, positive cash flow and sequential growth into Q1 2004. Thank you and we would be happy to take your questions now.

(OPERATOR INSTRUCTIONS) John Pitzer of Credit Suisse First Boston.

Congratulations on a good quarter. A couple questions. One can you quantify a little bit what kind of bookings growth you might see in the December timeframe, and maybe give us a sense of what looks like memory versus the logic mix of the order rates that could come in on Q4? And Steve, a couple questions. Could you quantify what Matrix might have represented in Q3, what it might represent in Q4? And help me understand how the SEN numbers are impacted by the dollar/yen relationship?

John, we typically don't give bookings guidance going forward or quantify the order rates, but we are tracking, as I mentioned, a number of projects. We're seeing spending a big increase in spending in Q3, and we expect the additional orders to be placed into Q4 and Q1, so we are very optimistic about that based on the projects we are seeing and the CAPEX spending that is coming through. One of the things that you maybe thinking is why is the implant space picking up so quickly perhaps compared to some others? And I think we have shown you in the past that it's the cyclical buying of the implant business, the higher highs and the lower lows; this is a very positive sign for us because we do think it is beyond the technology buying that we have been living through here over the last year and it seems to be moving more to capacity orders.

Has that move been beneficial to pricing?

It is getting better. We're seeing improvements overall in pricing, so we are encouraged by that as well.

Improvements in the rate of decline or the absolute AFB's (ph) are starting to go up?

Leveling and improving and we're seeing differentiation and delivery slots are becoming more of a concern to customers and pricing.

And then Steve, just the Matrix contribution Q3, Q4, and the dollar/yen relationship and how that impacts the SEN revenue stream and contribution?

I will start with the Matrix. I think as we said before we expect that we are going to have net losses including integration expenses for Matrix of approximately $2 million a quarter through the end of the year. By the end of the year, their operations will be fully integrated with our operations both in Rockville and here in Beverly. So those costs will essentially go away. About 2 cents a share in each of the quarters, Q3 and Q4. With respect to the yen, both on an orders and shipments basis, the strength of the yen against the dollar in the third quarter had the effect of about 5 percent.

John, one last thing, the only other color I could add for you is we commented on how strong the memory, IDM and foundries and we sort of grouped them that way, because as you know many of the Taiwanese foundries are memory focused. There was a big part of the order rate in Q3. We believe that in the fourth quarter we would start to see a strengthening in the logic IDMs and foundries. I think that is again a positive indicator of a more balanced recovery.

Great, thanks.

Glen Yeung of Smith Barney.

Just a follow-up, can you give us a sense as to what you think Europe will do for you in the fourth quarter in terms of orders?

I think you know the three major players there. We are starting to see through the memory manufacturer in Europe activity, but primarily through Taiwan. And we are anticipating that the logic supplier there will start spending. We have seen some spending in Q3 but we anticipate that will pick up again in Q4 and the beginning of next year.

Are you talking spending or orders in Q4?

Orders.

And then I wanted to also question you on the RTP sale that you had in Japan. Can you just remind us is that through a SEN or is that through Axcelis Technologies own?

That is through Axcelis.

And also on SEN, if you could give us a sense as to -- you mentioned gaining some share in medium current. I wonder if you could just quantify that somehow in terms of some of the strength you're seeing there?

Basically it is, there are two (technical difficulty) in Japan, Mission (ph) and SEN when it comes to medium current. And what we have seen happen over the last 12 months, 18 months is Mission has moved off of Japan and has moved into Asia and their focus has shifted and so SEN has taken advantage of some of that. In addition, some of the advanced consumer devices have a very high degree of sensitivity towards energy contamination and the MC3 tool has period performance in that particular space and so many of the selections are coming down to technology requirements. I don't want to speculate on share numbers, but I can tell you that we believe that it is certainly north of the 25 percent range.

One other just a clarification. I think you said guidance would be in the mid 30 percent range. Just to be absolutely certain, that's not 30. something, but 35 something? Is that right?

35 is correct, approximately 35 percent.

Stephen Paleyo of Morgan Stanley.

Mike, could you make some comments about the high current market? You talked a lot about taking share and one of your competitors here in the US had a press release, I think about 7 customers in Asia in high current as well. Is it just the third guy losing everything or what is going on in the high current world?

We are trying to count those 7 too, but the volume buying that is occurring is still predominantly -- if you look at the mix of where people buy, its productivity gains. And most of the volume buying is still in the multiwafer area between we think, between Axcelis and Slide, (ph) frankly. It is a duopoly if you will at a high level. There are some orders that are going based on momentum that our other competitor here has built on, but we think it is limited and it is for very niche applications and in addition several customers I have spoken to have indicated that the tool is really not a high current tool. It is actually bridging some medium current applications were there are higher doses and which is traditionally the medium current space. So I think they are stretching frankly the space, and we are okay with that. In the end when the marketshare numbers come out as we indicated last year, we would gain share and the evidence from Dataquest proved that to be the case. We believe the same will be true in '03.

And Steve, could you help me understand the deferred impact here? I think you talked about SEN as a result was about 7.8 million or so impact in this quarter but next quarter's guidance talks about roughly 6.5 to 7 million contribution from SEN. So is there -- how come it is not a higher number for this quarter I guess since a lot was pushed into next quarter?

The business practices at SEN are a little bit different, where their payment terms are contingent upon formal customer acceptance. So a very, very significant level of their significant amount of their shipments were deferred, and 100 percent of the revenues associated for those shipments were deferred. I think I said about 43 million out of 62 million in total shipments ended up being deferred. We expect that most of those will be accepted during the fourth quarter. Similarly at the end of the fourth quarter, we will have a number of shipments that will be deferred until acceptance, and we are forecasting that those will offset one another. So we expect the effect on SEN in the quarter to be minimal.

All right, great. Thank you.

Ali Irani of SIBC.

This is Peter Wright for Ali. Congratulations again on a great quarter. My first question for you, Steve, if you could give us a little more color on gross margins, more specifically you alluded to pricing easing going forward with lead times being more important. If you could comment on what percentage of shipments in the third quarter came from shipped from sale and what direction you see that going forward? Specifically, if there is any pricing pressure still existing in what product line that exists? Secondly on operating expenses, you guided operating expenses down 10 percent quarter-over-quarter, but it looks like once again there is potential for operating expenses to decline in the first quarter of 2004, taking out the 2 million of Matrix. If you could clarify it by understand that correctly, and I have one follow-up.

Okay. We get back to shipped from sell. We are currently shipping about 20 percent of our systems are shipped from sell. We expect that that percentage will increase going forward into 2004 as our 300 millimeter tools are qualified. What was the next question, Pete?

The 300 millimeter crossover, do you have any visibility into when gross margins on your 300 millimeter platform versus 200 millimeter platform will be comparable?

We're looking late 2004, 2005 as to when they stand in margins for the 300 millimeter tools will approach those of our 200 millimeter tools. A lot of that is dependent upon the volume. We have not had volume orders yet, so as the volume increases the manufacturing efficiencies will be built into that and some of the warranty costs that we have experienced will also decline and by the end of 2004, we expect that those will be much closer aligned.

Robert Stern of Needham & Company.

You talked about the business practices at SEN being a little bit different, but typically my understanding American companies defer maybe 10 percent of the cost of the equipment for later, but SEN apparently has deferred almost two-thirds, so I am wondering if you could discuss what does business practices are that would have the lion's share of the equipment revenues deferred?

You are right. We are deferring here at Axcelis in the U.S. about 10 percent of our systems sales attributable to future deliverables. From the SEN perspective, their payment terms are structured as such that payment is not due until formal customer acceptance of the tool, which under the guidance prescribed by EITF 0021, they are not permitted to recognize revenue on shipments until acceptance. So whereas we are recognizing 90 percent of revenue on systems shipped at the time of shipment, they are recognizing none until formal customer acceptance. It is based on the standard payment terms in Japan.

Okay, thanks. I have a second question, if I might. You talked about having strong order, strong revenues for two or three more quarters and you said that in that case strong meant that revenues would be increasing in at least the fourth quarter, and the first quarter and possibly even the second quarter. Does strong bookings also mean that bookings must increase, or could they be flat or slightly down and still be strong in your view?

I don't think we gave guidance beyond fourth quarter in terms of revenues, although I think that what we did say was first quarter based on orders we received in Q3 and projected orders in Q4, we anticipate strong revenue growth in Q1.

Does that mean up, does strong mean up?

Sure, strong means up, usually. How much? We have pretty good forecast growth, but into Q2 we are still looking at a three to six-month window with pretty good visibility and beyond that it is not really too clear. I'm sorry, I missed the second part of your question.

It was the same question on bookings. Does strong bookings in the fourth quarter mean that bookings will be up sequentially?

Yes, they should be.

Thanks a lot.

Raj Seth of SG Cowen.

Just a quick follow-up on gross margins. If we do not assume that pricing improves much more, let's just assume its stable, is there any way you can help me think about gross margins and how they will trend on increased volumes, or does the mix effect 200, 300 make that difficult? In other words, how should I think about gross margins progressing as the revenue line starts to increase across '04?

Mix will have an effect and margins on our 300 millimeter tools are less than our margins on our 200 millimeter tools. But at these shipment levels that we are talking about now in the 95 to $100 million range, we expect the drop through on the margins it will depend somewhat on mix, but we are forecasting that those margins will drop through in the high 40 percent range.

That is helpful. Thanks.

Jim Covello of Goldman Sachs.

A couple questions. Stephen, you had gone through the earnings impact of the Matrix acquisition this quarter. Can you help me understand the impact from a revenue and order perspective on this quarter? Then I have a couple follow-ups. Thanks.

The revenues from Matrix all sources service and products were approximately 1.2 million in the quarter. And the bookings were essentially that. Orders were essentially about 1.2 million also.

Thanks for that. Next, I have a question on the revenue recognition. You stated that the policy change was as a result of a policy that was enacted or went into effect on July 1. How come we haven't heard about it before today? You reported your last quarter after July 1. Did you know about that then? And can you help us understand that a little bit better?

We knew that there was in all likelihood that our revenue recognition policy would have to change. We were not in a position to discuss that or the effect of the change at that because of complications of measuring the effect at SEN. We had previewed the effect on Axcelis, but the effect on SEN was substantially more and that took a much longer term. So we were just silent about it until we had it quantified.

Are you going to be able to help us with any kind of pro forma financials restated for the impact of historically so we can maybe make some apples-to-apples comparisons versus your past performance?

We have not gone back and restated the past performance because there are certain deliverables that are difficult to measure, specifically those associated with extended warranties and we have not gone back and evaluated those, but we will provide information going forward on gross shipment dollars. Gross shipment amounts under both the new GAAP revenue recognition policy and our old methodologies so that you can align those and we do have a measure of the effect on gross margin, so you will be able to align that.

That will be helpful, thanks. Final question. There was a $10 million charge associated with the change in the revenue recognition policy. Can you help us understand exactly where that shows up in the P&L?

7.8 million of that is in the SEN contribution. SEN's contribution without the change in revenue recognition would have approximated $8 million for the quarter and the balance is in gross margin.

Thanks very much.

Patrick Ho of Moors Cabot.

A lot of my questions have been answered. Can you just give me a little more color on the high current market share gains again from the perspective of the technology? What are your customers saying about batch versus single wafer again?

In general what we are hearing is that customers see the ability of extending multiwafer platforms below at least to 65 nanometer geometries and potentially below that. And in fact even below 65 nanometers they are saying that if their applications were vertical structures which would drive high-tilt, high does applications that even with that, there would still be a mix of multiwafer and single wafer platforms. Rather than calling it single wafer I would rather call it multiwafer and high- tilt platforms. It is not clear whether or not a multiwafer platform would in fact deploy some high-tilt applications going forward. So we are very confident based on what our customers are telling us, the momentum we are seeing, the drive on productivity.

It is a no-contest from a productivity point of view between the existing single-wafer products that are out there and the multiwafer products that are offered. The throughput and cost of ownership is a no-brainer, and as long as people can see that kind of an advantage, they are not going to switch.

Fair enough. And I know this is difficult to forecast going forward, but since you mentioned that the gross margins' levels of 300 millimeter tools will not reach the same level of 200 millimeter tools until later in 2004, can (technical difficulty) that margins would flatten out sometime in early 2004 as more 300 millimeter tools are recognized as revenue?

We are not forecasting that. We think there will be -- as volume increases during the upturn that we are going to get increases in margin.

Let me just comment because it might help to frame the things we have been doing on margin improvement because we talked about volume and leveraging our fixed costs, but we also have productivity gains that we hope to get as we start shipping 300 millimeter ship from sell and as volume picks up on those, we will get a big productivity gain, labor productivity. We've already mentioned some of the sourcing we have been doing both in terms of external sourcing, but also long-term agreements with our supply base and sourcing internationally, where we are seeing some significant benefits. And then finally we mentioned the warranty and (indiscernible). The warranty costs in the 300 millimeter tools have been intentionally expanded because we are trying to make sure we get these tools validated and production worthy so that when the ramp occurs we will be able to take advantage of those and in all of those areas we're making progress. Clearly the big lever in '04 because of the volume change will be volume driven, but these other things could help us through the year as well.

Thank you.

Mark FitzGerald of Banc of America Securities.

This is Murani Ramuri for Mark FitzGerald. Can you tell us what percentage of your revenues came from your top five or top ten customers and how is that going to change going into next quarter?

Good question. Hold on a minute. This is a question of what percentage of the revenues came from the top 10 customers?

Actually if you look at system revenue, our top 10 customers accounted for 90 percent of system revenue, 2 or 3 percent of total revenue.

Total revenue it’s probably just under 90 percent of the top ten customers.

And do you expect that to broaden going forward or is that more or less going to remain around that number?

I think it will broaden some, but frankly the big players tend to -- asset heavy guys tend to spend the most money.

Okay, because I think a couple of quarters back you mentioned the top five was as high as 88 percent of the revenues, so I guess do you have the top five number by any chance?

Was Q1.

Exactly. Has it broadened out now or is the top five still around 85 percent?

It's much broader. The top five were probably closer to 60 percent.

Excellent. Thanks a lot.

Peter Wright of CIBC.

Just a follow up on the high current single wafer platform that one of your competitors is introducing. To better understand that it seems there is 2 of 30 high current implants that require high-tilt of the 65 nanometer node. Are these tools being used to address -- they can be thought of that these tools are addressing a niche market within the high current segment specifically or is there kind of a replacement cycle that will occur probably not this cycle but the next cycle? And what are your plans of development to address the single wafer pipeline?

Our view is in fact that up from 65 nanometer northbound, we are not seeing the kind of structure that we are going to anticipate at 45 and below, so let's separate those out. You are right. There's limited applications for high-tilt high-dose implant steps to 65 nanometer. We think that actually -- I don't know whether it is two whether it is four, whether it is 3, but as I mentioned earlier if you look carefully at the space that they are covering, they are gaps in median current performance. In fact, several of our customers have asked us whether or not we could stretch our medium current implants to cover these because it would be much more productive to do these on one tool, capture the other medium current, then to have a niche tool that is going to be very underutilized. I do not expect based on my other point that I made on productivity, that there will be a massive shift to a nonproductive or low productivity tool to cover 2 of 30 implant steps. I think that would be insane.

Great. One follow-up actually on the tax benefit going forward, you haven't been planning a tax benefit in the losses. What has that accrued to on a go forward basis?

As you recall from last quarter we wrote off deferred tax assets and took a charge of approximately $70 million. We will not recognize a tax benefit in any quarter in which we have a loss until we have a history of profitability. Similarly, while we are developing this history of profitability, we will not be recognizing any income tax expense because any taxes that we would have payable will be offset by net operating loss and other research and development tax credit carry forwards. So I do not anticipate that we're going to have any taxes recognized at least to 2004, and then it will be a determination as to whether we would restore the deferred tax assets at that time. But there will be no tax benefit or no tax expense or minimal tax expense through at least 2004.

Great, thank you.

I couldn't answer one of your last questions which was on the roadmap for single wafer, and I just want to state that we have on the road map that 45 nanometer an introduction of a high-dose, high-tilt tool, so we will be ready for those applications and have reviewed that with our current customers.

John Pitzer of Credit Suisse First Boston.

Yes, a quick follow-up question to the revenue recognition. Could you give us the deferred revenue balance at the end of Q3?

Deferred revenue balance the end of Q3 was approximately 4.2 million.

Thank you.

Glen Yeung of Smith Barney.

I got cut off, so sorry if somebody asked, but what was service as a percent of revenues in the quarter?

Service revenues were about 31 million, so they were just over 50 percent.

Okay, great. Thanks.

Ted Berg of Lehman Brothers.

Could you talk about the adjustments or warranty costs you talked about, the 300 millimeter, what are some of the issues prompting higher warranty expenses with those tools? And then on the productivity issue of high current, I know in the past you talked about a lot of customers talk about the productivity benefits of the high-energy tools for chaining implants, but with the high current tool, I thought the beam was the limiting factor there for throughput between batch and single wafer. So I was wondering if you could again recap what some of the productivity advantages you're talking about for the batch tools specifically at high current are?

On the warranty piece, there are two particular drivers there. One is simply 300 millimeter tool performance. New tools -- warranty liability as we're getting those tools stabilized, we are seeing dramatic improvements as we're getting more of the tools out into the field, so we're very pleased with the improvements we're seeing there. The other one is more centered around the extended acceptance criteria and the qualifications that customers are putting us through. This is one we decided rather than fight we should ensure the success of these tools, which is one of the reasons we added additional cost there. I do not expect this to be ongoing. We have aggressive w&i, warranty and install, cost reductions in place and we're getting the results that I think we need and we will see those improvements come through during the course of '04. On -- could you just restate your question on the high current again?

I know what the high-energy tools that there are productivity benefits from doing the chained implants, but I wonder why they do chained implants with the high current tools and I thought the very low energy of those tools are one thing that eliminates some of the advantage of single versus batch for throughput issues, so I was wondering if you could remind me of what the productivity benefits are for the batch tools at high current specifically?

You hit on one of them which is it turns out that in the low energy region, sub 1 ked implants, we are able to generate much higher beam currents, which translate into throughput, than a single wafer tool. The single wafer tool has a very long beamline and you just do not get -- you think about duty cycles and in one respect translated into implants, that is what you get inefficient duty cycle for the implant step. So there's a significant issue there. The second major area is the ability to do quad implants, multiple implants inside without moving wafers in and out of the process chain. The huge setup and processing time advantage that we have over not only single wafer platforms, but also happens to be a big advantage we have over our other multiwafer competitor. It depends. This whole argument of single wafer/multiwafer/chaining/doing implants on a high-energy or high current is really starting to blur a lot of these traditional segments and obviously if I am variant, I'm going to pitch it one way, and if I am Axcelis I'm going to describe it another way. We are all fighting for the same economic space, but in the end it really does come down to productivity. It comes down a lot to particular customers, recipe mix, and the process flows, and how they design those. And in general in high current for the reasons I mentioned multiwafer wins 95 percent of the time.

Jim Covello of Goldman Sachs.

Quick follow-up on the service revenues. It seems like a high percentage. Can you talk about that a little bit? Thanks.

I think it’s more a reflection of the systems revenues from systems sales being down in the quarter. Actually the revenues for service have approximated they tend to be flatter and have approximated 30 to 31 million in each quarter of this year. So I think the percentage is more a reflection of the timing of the system shipments and the low volume in the third quarter of systems sales.

Thanks.

If there are no further questions, I'll turn the conference back to Mr. Namaroff to conclude.

Thank you very much. I just want to pass on one comment from Mary. Mary just sent her regards to everyone. I want to let you know she will be back in the office on Wednesday, for those of you who would also like to ask questions of her personally. Please contact me if you would like to set up an appointment with her and I would be happy to do so. Thank you very much and have a great evening.

Ladies and gentlemen, this concludes our conference for today. Thank you for participating and have a nice day. All parties may now disconnect.