Axcelis Technologies Inc (ACLS) 2005 Q3 法說會逐字稿

Thanks very much for holding, everyone and welcome to the Axcelis third quarter 2005 earnings release conference call. Just a reminder that today's conference is in being recorded. Later we will conduct a question and answer session and instructions will follow at that time. For opening remarks and introductions, I would now like to turn the call over to the director of investor relations for Axcelis, Jim Kawski.

Jim?

Good afternoon. This is Jim Kawski, director of investor relations for Axcelis Technologies. Welcome to our conference call to discuss our results for the third quarter of 2005. I'm sure all of you have received a copy of our press release issued earlier today announcing our third quarter results. If not you can download the release via our Web site at www.axcelis.com.

Discussing a results today are Mary Puma, chairman and chief executive officer, and Stephen Bassett, our executive vice president and chief financial officer. Also joining us is Mark Namaroff, our senior vice president of marketing.

The prepared remarks will last for approximately 15 minutes after which there will be time for questions. Playback service will be available via our Web site 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 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. In addition, I'd like to briefly comment on remarks that Steve will make regarding our revenue performance and projections.

When we speak of worldwide revenue we are referring to the aggregate revenues of Axcelis and those of SEN, our 50% owned unconsolidated subsidiary in Japan. Please understand that we do not consolidate SEN's revenues under generally accepted accounting principles. We use the term net revenues to mean Axcelis-only revenues, determined in accordance with GAAP.

We provide data on worldwide revenues with SEN because we believe that it is useful to investors. SEN's ion implant products are covered by a license from us and therefore the combined sales of the two companies indicate the full market penetration of our technology.

Now I would like to turn the call over to Steve.

Thank you, Jim.

Overall, our results for the quarter were in line with expectations. We've recorded a net loss of $5.2 million, or five cents per share, which was better than expected in spite of incurring higher than anticipated restructuring costs due to timing associated with the move of our Rockville, Maryland operations.

Total cost attributable to the restructuring effort approximately $3million and increased net loss by three cents per share. Net revenues were $87million and worldwide revenues, which include SEN, or $126 million, exceeding our guidance. The sequential decrease in worldwide revenues was primarily related to the timing of shipments in Japan. We expect the level of business in Japan to increase significantly in the fourth quarter of '05.

Our service business continued to show strong performance with revenues of $39 million. Our service operations continue to be accretive to our overall gross margins. Revenue from system sales came in at approximately $47 million, which was in line with our expectations. Shipments of 300 mm products constituted 47% of the total for the quarter and 61% of total shipments year to date.

From a product perspective, our implant business accounted for 76% of total shipments. Systems and service bookings for the quarter were $69 Million compared to $90 million in Q2, and on a worldwide basis total bookings were $118 million down from $149 million in the previous quarter.

We attribute the sequential decline in bookings to the timing of placement of customer orders and while we do not give specific quarter guidance, we expect the Q4 order great to approximate the quarterly levels we realized in the first half of the year. We ended the quarter but a systems backlog of $38 million and deferred revenue of approximately $42 million.

Based on the geographic location of the fab, Asia accounted for 68% of systems orders and with 15% coming on the U.S. and 17% from Europe. Including SEN, approximately 83% of new systems orders were from Asia.

Memory manufacturers accounted for 48% of new systems bookings with logic at 28% and foundries 24%. While the percent of foundry orders has increased from what we've experienced over the past several quarters, it appears to be a matter of timing as foundries remain cautious about adding capacity.

On a product basis approximately 61% of systems bookings were for 300 millimeter. Including SEN 73% of new systems orders were for 300 millimeter products. Gross margins at approximately 41% were in line with expectations.

Operating expenses for the quarters including restructuring and related costs were approximately $43 million, in line with our forecast. As mentioned earlier, operating costs include approximately $3 million of restructuring and other costs related to the move from our Rockville, Maryland facility.

The contribution from SEN for the quarter, which includes royalties and Axcelis' 50% share of their net income was approximately $2 million. This compares with $12 million in Q2. As we remarked last quarter, the expected decrease in SEN's Q3 contribution is related to the timing of business activity around the close of SEN's fiscal year which occurs in our second quarter.

We expect SEN to continue to show strong operating performance and continue to dominate the Japanese implant market. SEN's average quarter contribution on an ongoing basis is expected to be in the range of $4 million or $6 million.

Income tax expense benefited by approximately $500,000 from adjustments made to prior year accruals. Cash flow, although negative at $4.7 million, was better than expected due mainly to accelerated collections could be viewed as a positive indicator of customer satisfaction.

Looking forward to the fourth quarter of 2005, net revenues excluding SEN are projected to be $85 million or $95 million. Worldwide revenues are forecast in the range of 145 to 160 million higher.

Gross margins are expected to decline to a range of 37% to 40% due mainly to product mix. Looking beyond the fourth quarter and into 2006 we do expect continuing margin pressure as we start to recognize revenue from new product acceptances.

Rolling out new products requires incremental investments that typically result in lower product margins in the short term. We expect the Optima platform margins to return to normal levels as we achieve volume production.

Operating expense levels in Q4 are expected to be approximately the same as Q3. SEN's income and royalty contribution for the fourth quarter is expected to be $6 million. As I mentioned earlier, we expect SEN's quarterly contribution to average $4 million or $6 million but it can fluctuate quarter to quarter based on the timing of product delivery and acceptance.

Income taxes are forecast at 300,000 and we expect cash flow to break even. We are currently forecasting a loss for the fourth quarter ranging from 3 cents or 7 cents per share, reflecting the lower gross margins and the effect of costs associated with the move from our Rockville facility. Approximately 2 cents per share.

I will now turn the call over to Mary.

Thanks, Steve.

I would like to use this time to discuss some of the specific challenges and opportunities we have in the implant market. Let us start with some background on the market environment. 2005 and 2006 are transition years in implant products and technologies.

While customers continue to buy multi wafer tool, leading edge customers are shifting to a single wafer tools. These customers are beginning to sort through the different beamline architectures that single wafer systems offer.

Ultimately, they will select tools that offer the highest productivity, the most flexibility, the best process performance, and the greatest extendibility. In the midtovestates (ph) traditional medium current tools appear to behaving difficulty meeting customers' emerging needs at 65 nm.

What customers want are tools that offer broader applications coverage. They need a tool that can do both traditional medium current implants in addition to implants requiring lower energies at a higher doses.

The Optima MD is the only tool market with these capabilities. The Optima MD has the widest application range and therefore the greatest flexibility. It has proven productivity advantages, in particular for the advanced tailor (ph) application. The Optima MD's patented energy filter design provides unmatched contamination control resulting in higher device yields, and finally the Optima MD is designed for the future. Its use in Imec's (ph) 32 nm development program validates its extendibility.

Four customers have already installed our new mid dose tool and one of these customers has receives multiple units. The mid dose market is a growth opportunity for Axcelis and we expect our investment in the Optima MD to begin to pay off in 2006.

In high dose, or what was traditionally called high current, competitive offerings are temporarily filling the applications gap. A common misconception is that the single wafer high current market is closed, that customers have made their final decisions and Axcelis is too late to the game.

We continue to hear from numerous customers that they are not satisfied with the single wafer high current offerings that are currently available. There are two main concerns that customers consistently raised that we believe the Optima HD soundly addresses.

First, customers do not like the productivity and process compromises they must make using a ribbon beam system. Our customers know that simple spot beam technology is superior because of its simplicity. A spot beam requires less manipulation, less focusing, and fewer magnets and electrostatic systems. This allows for a shorter beamline and better beam transport, more consistent process performance and greater productivity across the board.

These fundamental advantages coupled with Axcelis' patented electron confinement technology results in optimized high dose performance. Second, customers are concerned about the reliability of competitive 2D mechanical stand end stations due to vibration.

The Optima HD pendulum scan technology, which we call our Radius Scan system, demonstrates smooth, vibration-free operations at higher scan rates. Customers familiar with both the Radius Scan and the competitive system confirm this. With a highly reliable single wafer end station, the Optima HD now brings competitive particle performance, superior process performance, and required reliability to market.

We believe that the Optima HD is the only system on the market there will address customer concerns and meet process performance needs. As initial validation, I am happy to announce that we have just received our first customer commitment for the Optima HD. We expect to finalize this multi system order shortly. There is serious customer interest in the Optima HD and we will share more with you throughout the coming months as we will be acutely focused on gaining traction for the Optima HD in 2006.

Finally, I would like to comment on the competitive environment in Japan. One of our competitors is making much out of gaining a new customer in Japan. To understand the sale in context, we should look at the larger picture.

The Japanese implant market is dominated by three Japanese suppliers who together have 88% market share in 2004. This competitor had a four percent share of the Japanese market in 2004 and eight percent in 2003.

The company with a largest share at 62 percent is our joint venture, SEN. Is important not to underestimate the determination of SEN to retain its market share position. But, like most Japanese companies, SEN will take along-term deal and will not compromise its business to compete at unsustainable price points.

SEN will retain important customers, relying not on price but on an in-depth understanding of customers, on cost-effective, high-quality products, and on an unrelenting focus on customer satisfaction. This strategy has driven SEN's success to date and we see no reason to assume that price driven penetrations represent a change to these market dynamics.

In 2006 we will be focused on capitalizing on opportunities and addressing our challenges. Our priorities in implant are first, to penetrate the 65 nm single wafer high dose implant market with the Optima HD based on superior low energy and contamination free performance and its highly reliable single wafer end station.

Second, to begin reaping the benefits of the Optima MD based on its flexibility, productivity advantages, unmatched contamination control and extendibility.

Third, to ensure that SEN continues to dominate the market in Japan because of their cost- effective high-quality products and their unrelenting focus on customer satisfaction. And finally, to continue to aggressively sell our market share leading multi wafer implant products.

At Axcelis, we are energized, we are excited, and we are committed to turning our challenges into a winning future. We would be happy to take your questions now.

[Operator instructions]

First up in the roster is Robert Maire at Needham. Go ahead, please, sir.

Yes, in terms of orders going forward, could you give this a little more color between ion implant and strip and strengthened strip versus strengthening plant in a little bit more detail, please?

Robert, this is Steve. I think that going forward you will see is what we have seen -- We do not see any change right now in the mix, really. We have been averaging 75% to 80% implant and we think that that will continue, at least for the near term.

OK. So there has been no changes in that?

No significant change in mix, no.

Would you attribute part of the current comparative weakness that you have experienced due to customer concentration, i.e., other firms having more concentration in the memory market, which is perhaps doing a little bit better now as compared to other markets which may not be doing quite as well right now?

I do not think so, Robert. I think that we have our fair share of the memory market, and I do not believe that that is contributing at all.

And one last question. Your new ion implant product, do you have any of those placed with current customers of your competitors' products?

Are you talking -- the only product we have a place to date are the HD -- I am sorry the Optima MD, and SEN has placed the SHX and yes, those are in locations that have displaced competitive products.

OK. Thank you.

Next up question is a question from Jim Covello at Goldman Sachs.

Can you hear me OK?

Yes, sir.

OK. Great. Thanks. Good afternoon. Thanks for taking the question.

A couple quick ones. First of all, your revenue guidance for December is obviously well above your order intake for September, and you were saying you expect orders to return back to the level of the first half of the year. What level of orders to you need in the September quarter to keep revenues flat with the December quarter guidance? In other words, are you close to eating too much into your backlog and then meeting declining revenues into the first half of 2006? So what level of orders do need to keep revenues flattish going forward?

I think what I said is we are currently looking at our systems orders will approximate the levels network in the first and second quarters of '05, and one of the things and we are experiencing is that we are receiving orders and shipping products in the same quarter, much more frequently than we have in the past.

We manage our business so that we build products based on what we think anticipated ship dates are. But we frequently do not receive the orders and until the same quarter when we are shipping. So we're comfortable that if we get the order levels that we saw in the first half of the year, that we will be OK going into 2006.

OK. So the order levels that we're talking about for December would lead to flattish kind of revenues in the first half of '06?

We do not see much of a change in the overall environment that is going to change our revenue base significantly.

OK. Second question then. What is it or to take to get a return to profitability in first half of '06? So if we can keep our revenue -- assuming the discussion we just had, we can keep revenues kind of flattish, what does it take to get back profitability in the first part of '06?

The biggest variables for us on profitability are revenue levels, which we need to see in the neighborhood of 90 to 95 million, and then margins, as I said, we have had margin pressures with the new tools, and if we can - the pace of gaining traction with the new tools and allow us to get into volume production and reduce the cost, get the tools in to ship from sell will determine how fast we can accelerate the margins back up into the 41% or 42% range. So from assessing our profitability, I would be looking at those major variables, revenues and our margin performance.

And I am sorry if I missed this on the call, going back earlier this year we had talked about being able to maintain profitability through the downturn, and it seems like the big variable there, the big difference now is that some of the new tools are going to create margin pressure. Didn't we know that the new tools would create margin pressure, or is there incremental pressure to the margins?

Well, the -- we have a model that we have modeled that we had a break-even level at $90 million in quarterly revenues, and it assumes a gross margin percentage of 41.5%. That is the way we modeled our business.

We do expect anytime introduce new products we will have margin pressures downward. To the extent that our new product introduction provide incremental revenues then our revenues would go up and our model would still work. To the extent that our revenue levels on other products are running below $90 million, you could expect in that scenario to see slight losses.

We do not anticipate no matter what levels that we're at that our losses are going to be significant. We're going to be bumping along at or close to the break even level.

Fair enough. Thanks so much.

Moving on now to Bill Lu at Piper Jaffray.

Yeah, hi. Good afternoon.

I'm looking for just a quick clarification here on your fourth quarter guidance. Are you actually going to recognize revenues on the Optima in the fourth quarter?

Probably not in the fourth quarter. It looks like the first revenue recognitions on the Optima will come in in '06.

So when you said margin is down due to mix, it is not related to Optima, right?

No. It's a product mix. It's a shipment configurations that we have got planned for the fourth quarter, principally related to non implant products that are pressuring margins. It is a fairly big range because we've got some eval tools that could convert, they could push the margins back up to the closer to 40% range or if they don't convert, we may be in the 37% range.

So you're expecting non-implant to be up next quarter for shipments or revenues?

No it's - I don't expect non-implant to be up. It's just certain product configurations. We do have certain products that are lower margin and we have a disproportionate amount of those in Q4.

I see. And then also on your fourth quarter guidance, you said that bookings are likely to be up. Can you talk about that? Where is that geographically?

It is spread around the world. There's not a specific geography or type of customer where we're going to see those bookings.

OK. Great, thanks very much.

We have a question now from John Pitzer at CSFB.

Hi, can you hear me?

Yes.

Yes, John.

Hi, this actually Sept (ph) here for John.

Mary, can you give us a quick update on where we are with the customer progression, customer rents (ph) with Optima - I know that earlier you had 10 customer targets. How are you progressing towards that?

Well, if you include this Optima HD that we are talking about today, that would be seven customers that have now designed to end tools based on Optima technology in 2005. That is seven out of 10. And we're very pleased with this progress and we are still targeting 10 wins for the year.

The XT (ph) win that you got was a customer who is not currently an MD customer?

That is correct.

And I was just doing a little math on the backlog, and is their any adjustments to backlog or any order cancellations during the quarter?

No order cancellations during the quarter.

OK. And can you give us a little bit clarity on this product contribution (ph) affecting margins? Is that specifically the jumpforRTP (ph) or what ...

All we really can say is non implant products.

OK. Thanks.

From Citigroup this Timothy Arcuri.

Hi, guys. This is actually Dan Berenbaum for Tim.

Two quick questions. On the HD, if you look at the 65 nm node, how many decisions have been made there? If, say, you look at your top 10 customers, how many decisions have been made for 65 nm? How many are left to be made? And do you feel like you have missed an opportunity with maybe a few of them but there are a few of them that still remain? How do think about that

Dan, this is Mark Namaroff. How are you?

I just wanted to kind of maybe clarify maybe some of Mary's comments - Mary said in her comments about 65 nm. I think generally speaking we need to talk about the 65 nm, that is a very generic statement. All in all, when you look at customers across the board, first of all as you try to look at segments, memory, for instance.

Memory is nowhere close to being at 65 nm now. They're migrating to 90 nm, may be at 70 nm by the end of the 2006 or beginning of 2007.

So I think right there, there is a huge opportunity for us to be - design then at 65 nm. Most logic fabs are also moving to 65 nm, but their volume is very low. So when there is really a high-volume 65 nm production, we think that is going to be also later in 2006, maybe even later. So that also creates an opportunity for us. I do not think we have missed out on all those decision points.

And just to add another point of fact, that 45 nm evaluations are also beginning as early as the middle of next year and that is going to be a major part of what we're going to do with the Optima platform, is that we are going to be positioned for high-volume production at 65 and also the early design wins at 45.

Those are regarding memory. Maybe let me ask the question another way.

Dan, can you speak up? We cannot hear you.

I'm sorry, can you hear me better now.

Yes, thanks:

I apologize for that. If I rephrase the question on memory, if they are not at 65 nm then I might ask where the single wafer high current get designed in? How many of those decisions at the top three or four memory manufacturers have been made, and in terms of designing at the node that single wafer comes into play?

Like I said, I think Mary even read it in her comments, leading edge customers are looking to move to a single wafer now. They are looking at it so I would put that in the context of all large memory manufacturers at looking to move in that direction.

There are manufacturers that are still using multi wafer tools, at least for some of the leading edge fabs, particularly at probably 70 nm and below, they're looking for single wafer tools. And we are working with those manufacturers to get designd in. So I think that that fits well with our strategy.

Thanks for the color. And just one other question, Steve, on the guidance for the restructuring, that $2million restructuring guidance for Q4. Does that include everything, and maybe you can give a little color also on this quarter there, I see there is 1.5 million that is included on the statement, and then there is another half that comes out of SG&A, I assume.

How does that split up for the guidance for next quarter?

I think the 2 million for next quarter will be split between what we will report as structuring costs and costs associated with the move and under generally accepted accounting principles reported as SG&A expenses. And in this quarter, you're right, we had about $1.5 million below the line and we will have $1.5 million up in SG&A expense.

OK, great. Thanks.

Next up is Martin Teng with SG Cowen.

Uh, yeah, thank you. Just a couple of questions.

The first question was with regards to the commitment for the high density single wafer tool, can you give us a little bit more details in terms of geography and in terms of what kinds of device manufacturers they are?

At this point I cannot give you additional information regarding this customer commitment.

OK. And also in your press release, you mentioned that there are several new customers for your Optima MD. Did I read it correctly in that you have more than 10 now or is it still 10?

So far to date as I just mentioned we have seven total Optima -- customers that have designed in technology based on the Optima platform. So we're at seven right now, not on 10, the targeting 10, still, for the year.

And your comments about market share in Japan, do you see any market share losses or do think it is still stable?

Do I -- If you look at market share overall, I think SEN is facing some of the same challenges that Axcelis is facing. And we've talked about how the implant market is currently in transition as customers sort through some of the different beamline architectures that some of the single wafer systems offered. And if there a share loss, but we believe that any share loss in the high dose segment is temporary, and our Optima HD will be production ready for 65 nm production high volume production and we expect to penetrate the high dose market in 2006 with this tool.

Just two more questions. The first one is with regards to, you mentioned that customers with current single wafer high current tools are dissatisfied existing tools. Can you talk a little bit more about what exactly they are dissatisfied with?

Sure, Martin. This is Mark. I'll add to that. The current toolset out there, I think what customers are experiencing, those tools, particularly the current ribbon beam type tools, were designed probably more than eight years ago. And when you look at those designs back then, they were leading edge tools, but when you look some of the leading edge processes that some of our customers are looking to run they are pushing the productivity performance of those tools as the energies are going lower and lower.

And I think we mentioned this before. When you go to lower energies, the productivity drops off. And that really challenges a ribbon beam system.

And a lot of our customers now are starting to realize that. At 65 nm and when they know they need to have higher productivity even at 45 nm, they're starting to realize that some of those inherent designs in those systems are sort of flawed.

So they're actually looking at a spot beam technology as potentially a better solution for low energy process performance. Not only during higher productivity but better contamination control and better uniformity and repeatability just to name a few advantages of spot beams.

So I think that is sort of the biggest factor that we have seen. In fact, we have actually invited customers to come in and look at the design of the Optima HD and they validated our approach which is very, very good to see from our perspective. That they have actually said that we are on the right track and the tool design and the beamline architecture can deliver performance that they need at 45 nm.

OK. Thank you. And just one last question for Steve which is the gross margin impact on the new tool in '06, how should we think about that? Is it a one or two percent decrease for '06?

We're going to give guidance as we progress on into '06 quarter to quarter because the effects of that is so largely dependent upon the timing of the recognition of the tools and which tool it will be. It's just -- There will be some margin pressure. The total effect on margins will be driven somewhat by the volume levels. But again, it will vary significantly by the timing of the recognition of the tools.

We will give you good guidance. It is really hard to project out into the first quarter right now or even farther than that, Martin.

OK. Thank you very much.

Next up is Hari Chandra at Deutsche Bank.

Hello?

Hello.

Thank you. In terms of the cost structure, agreed (ph) you have new products rolling out, when can we expect the gross margin to roll back to the 40% range that you have targeted?

This is Steve. We said that when we get to volume production and the first several tools will have lower margins we do not have the efficiencies built into the manufacturing process, and we do not have the cost out of the product that we have - the sourcing costs out of the procut. The material cost out of the product.

As we get into more volume and as the tools gained traction, we will start to see these tools margins progress up to the normal product levels that we historically have achieved.

Could to give a timeline to that?

It is hard to give a timeline because it really depends on the traction of the tools at this point in time. When it will be. As we get further into 2006 and we see the order base and the continue ordered pace for the HD or the continued order pace for the MD, we will get a much better feeling there.

We are expecting the MD to begin to ramp in the first half of 2006 and as Mary said, we should see some progression with the MD and the HD is probably going to ramp more in the second half of 2006.

And the top line (ph) be motrosembe (ph) convertible in January '07?

I've actually said this before but we are quite confident, I will back up. We don't expect that we will convert but we are quite confident that we will be able to liquidate the debt without causing any kind of liquidity issue for the company. I would like to add that the company has no intention of raising equity capital or issuing another convertible to bench with a share price at the levels - at these current levels.

Thank you.

From DA Davidson, this is Matt Petkun.

Hi. Good afternoon. Mary, when you're looking towards '06, it does sound now like what you are seeing is more of a wholesale shift towards a single wafer especially from high-end customers. But what would your expectations be, assuming a normal year for growth in semiconductor demand next year both at the high-end and low-end.

What would your expectations be for batch? And what sort of opportunities do you see for your batch tools as we see more capacity related purchases, hopefully, in 2006?

Well, I think we will continue to selling number of batch tools. Our single wafer, I am sorry, our multi wafer tool high energy tools still have market leading share. We expect that -- to be able to maintain that very high market share throughout 2006.

And we are actually still selling a fair amount of the high current tools. To the less advanced customers. For capacity buys and just filling out their fabs.

So basically we expect that to continue on through 2006 and even beyond.

Now, the more advanced customers, perhaps the more advanced foundries, would you expect that even if their initial inclinations for something like a 0.18 or 0.13 capacity, would that still be looking towards the single wafer tool just due to extendibility?

Matt, this is Mark. There is existing fab and they are looking to add capacity at 0.18 or even 0.13 microns, so it will be a repeat buy from what they currently have in the fab.

You think they will mirror the old batch until then?

They would. Why would they change it the currently have a half a dozen batch tools in their existing production line. So it is easy to assume that it would just be a repeat buy.

And again, to sort of ask the same question in a different way, I am looking at your inventory levels which really haven't come down too much below -- your order rates have declined pretty significantly here. What do you see in terms of exposure for that inventory? How much of it is batch related? And are we seeing both -- I know you are guiding down margins in general. Are we also seeing pricing pressure on some of your more legacy tools?

I will take the inventory question first. There are a lot of things that affect the inventory and one of the things that you see is that evaluation tools are in the inventory. We also have to build in anticipation of building our new Optima MD tools and Optima HD tools which we have not shipped or recognized revenue on. So we're actually pretty happy but the inventory levels and how they have performed to throughout the year. We do not have any particular risk that we're concerned about with respect to our inventory.

And on the pricing side of it, no, we do not see any significant pricing pressure any more so than usual on any of our tools and I think he specifically asked about multi wafer, so the answer is no.

Although it did sounded like you mentioned SEN was seeing some pretty strong competitive pricing pressure in Japan.

Yes, I alluded to die and in one particular instance it worked.

OK. Finally, for me, expectations for R&D levels as we still rolling out the HD and MD tools and we have seen that pick up over the last two quarters from more of the $15, $16 million level. Should we expect that to come down at all next year? Some of that is timing related with materials cost for R&D.

A lot of it is timing related. I think if you get into the fourth quarter I think you should look at R&D expenses as being relatively the same as we've seen them and I think that as we get out into 2006, we would expect that overall R&D spending would decline, at least overall for the year. As we get to - we're putting our profit plan together now and as we get to the next quarter, we will be able to give you much better guidance on that.

OK. Thank you.

We have a question from Patrick Ho at Legg Mason.

Thanks a lot.

Just I wanted a clarification in terms of -- I think you said, Steve, in terms of the bookings for the fourth quarter, was this referring to system bookings or total bookings looking ahead for the fourth quarter, as being combined for the first half of the year?

Actually the service bookings have been pretty flat throughout the year. We have had a pretty high level of service bookings. So the variability has all been in the systems. You could look at it either way, Patrick, but we were referring specifically to systems.

OK. Great. That does help.

Secondly, in terms of your discussion with the foundries, I think the common theme out there seems to be the timing with the foundries, Mary what have you been getting feedback on what is going to trigger, I guess some of the spending trends that I think all of us are anticipating?

Well, I think it is still based on utilization. I guess UMC just reported today that their utilization is at 85% and that is not going to do it. In the past might have triggered buying, but what we're seeing is that the foundries are being much more conservative and it is going to take much higher utilization level before they turn back on in any sort of significant volume.

OK. Great. And a final question in terms of the Optima and I guess more on the supply chain management side of things. Given that you have two new products coming out, both the MD and the HD, how would you characterize gathering supplies and material costs, what type of improvements can you make on that front over the next few quarters?

Well, I do not know that you're going to see anything over the next few quarters. Over the long term you should see the gross margins on the Optimas actually outperform the gross margins on the older tools because we have done quite a bit in terms of outsourcing, global sourcing, designs for cost, ship from sell (ph) we're trying to accelerate right now. We're working on accelerating when we actually cut and into production, we are focused on warranties and installs, so we're really out there on the number of fronts, looking very aggressively at the cost portion of the new tools.

Thank you.

[Operator instructions] And we will move on to C.G. Mews (ph) at Lehman Brothers.

I'm attempting to put some hard numbers around your order outlook and what the mix will look like between service and systems?

You know, as I said, C.G., the service has been pretty flat. We have been running between $38 and $39 million for the first three quarters and I think that we anticipate it is going to stay about there. So the variable for us has been in the - in systems bookings. So we expect the systems orders levels to increase to approximate the levels that we saw in the first two quarters of the year.

Can you talk a little bit about mix in your implant business between high current and high energy and what the trend has been like over the last couple of quarters?

We do not sort out any of our results based on different implants segments.

Last question. In terms of your service business, do you typically see budget flush in the fourth quarter, see stronger seasonality?

It varies year to year so we do not know how much to anticipate in our forecast. We have not forecast any big upside in service. We have forecasted on a normal pattern, quite frankly.

OK. Thank you.

OK. I would like to thank you for joining us today, and we appreciate you calling in. We'll talk to you soon. Bye.

Thank you, again, everyone. That will conclude today's conference call. Have a good day.