Axcelis Technologies Inc (ACLS) 2014 Q3 法說會逐字稿

Good day ladies and gentlemen. Welcome to the Axcelis Technologies third quarter 2014 conference call. My name is Philip, and I will be your coordinator for today. At this time all participants are in listen-only mode. We will be facilitating a question and answer session towards the end of this conference. (Operator Instructions). As a reminder this conference is being recorded for replay purposes. I would now like to turn the presentation over to your host for today's call, Mary Puma, Chairman and CEO of Axcelis Technologies. Please proceed, ma'am.

Thank you, Philip. This is Mary Puma Chairman and CEO of Axcelis Technologies. With me today is Kevin Brewer, Executive Vice President and CFO, and Doug Lawson, Executive Vice President of Corporate Marketing and Strategy. If you have not seen a copy of our press release issued earlier today, it is available on our website. Playback service will also be available on our website as described in our press release.

Please note that comments made today about our expectations for future revenues, profits, and other results are forward-looking statements under the SEC's Safe Harbor Provision. These forward-looking statements are based on management's current expectations, and are subject to the risks inherent in our business. These risks are described in detail in our Form 10-K Annual Report and other SEC filings, which we urge you to review. Our actual results may differ materially from our current expectations. We do not assume any obligation to update these forward-looking statements.

Axcelis reported third quarter results essentially in line with our guidance. As expected a slowdown in memory spending continued to impact our system sales, but it appears that the pause is beginning to thaw. Recent announcements of the sale of both our new Purion H high current implanter, and our high energy Purion XE to a large DRAM customer, provide a strong signal that the anticipated DRAM build is imminent. While the pause has created several difficult quarters for us financially, we believe that history will show that this slow down has been very positive for Axcelis.

On the product front it gave us time to offer customers the full family of Purion products for selection and upcoming memory build. With the Purion H as the most critical addition, as it reopens the important high current market segment to Axcelis. On the financial front, having completed key development projects during this time period we reviewed and reduced our operating expenses, further lowering our quarterly operation cash and breakeven level.

Coming out of the pause we have the full product family in place, we have a reduced cost structure, and we have the opportunity to strengthen our balance sheet as a result of the sale leaseback. Although regaining a strong position in high current and medium current will take time and effort, Axcelis is now ready and well-positioned to begin taking back market share in the ION implant market. Moving forward into 2015, we will focus on gaining share with Purion H and Purion M in both DRAM and NAND. The Purion platform already has a significant presence in these segments, due to our strong position in high energy. We will concentrate on placing Purion H evaluation units with foundry and logic customers, where the scanned spot beam architecture provides significant technical benefits ,and we will telephone to serve the second tier market with both Purion and legacy products, and will maintain our highly accretive service business.

Turning to Q4, our guidance reflects the beginning of the thaw. In Q4 we are expecting revenues to increase to between $50 million and $55 million. Gross margins will be around 30%, impacted by the lower margin associated with the shipment of our first Purion H revenue systems, and a significant increase in deferred revenues associated with increasing shipments. As a consequence our operating results will be lower than our breakeven model would suggest, ranging from an operating loss of $1 million to $3 million, with a per share loss ranging from $0.01 to $0.03. Our cash balance will be in the low-$30 million range before expected proceeds from a sale leaseback transaction.

In the quarter we continued to make progress with our Purion strategy. We shipped our second Purion H evaluation to another leading memory customer in Asia. Two customers now have the full Purion platform, and are able to take advantage of the power that it brings to their fabs. We also penetrated a new advanced power device manufacturer with the Purion XE, leading the way for the adoption of the Purion M and the Purion H at this customer in the future. The interest in our Purion products continues to grow in the leading-edge device market, where the technical advantages of Purion are critical. Mature logic and foundry customers who are often four wall constrained place a significant value on the productivity advantages Purion offers.

Our first two Purion H evaluations are proceeding well. It's always exciting to put systems into the field, as it provides an opportunity to truly exercise the tool in the customers' environment. As a result of the commonality of the Purion platform, the first evaluation has moved rapidly through the initial proven phase, and quickly into the recipe qualification phase. This customer is testing and qualifying a myriad of different implants on the system, ranging from material modification implants, to more traditional implants that create the electrical characteristics of the transistor. This positive result early in the evaluation period has led to a multi-system revenue order to support this customer's upcoming DRAM ramp.

These systems will ship in Q4. Receiving a revenue order only a few months after the start of a one year evaluation is a highly unusual event, and one that significant that the Purion H is a very special implanter. The second evaluation is expected to follow a similar path. Field performance confirms the advantages of the system's scanned spot beam and source technology, among numerous other benefits that include improved defectively, 500 wafer per hour throughput, and superior angle dose and energy control and uniformity. We are excited about the opportunities that the Purion H provides us, as this customer adds both DRAM and NAND capacity in 2015.

The Purion XE high energy implanter continues to add to its already leading market share position. It's linear accelerator or LINAC beam line, provides the highest level of productivity and lowest level of metal contamination in the industry. It is the system of choice, not only for memory chip makers, but also for customers making CMOS image centers. Customers new to the Purion platform who are using the Purion X for the first time are pleased with the system, and are excited about evaluating both the Purion H and the Purion M. The Purion XE is a strong weapon in Axcelis' arsenal to gain footprint and market share.

Our medium current offering the Purion M is qualified and running production at three customers in both the memory and foundry segments. The Purion M offers customers improved productivity, lower electricity usage, and a broader energy range than the competition. When combined with the Purion H and Purion XE, it provides the customer with both manufacturing and technical advantages. Although the medium current market is very competitive, with do inspect follow-on production orders for the Purion M during the 2015 memory build. With that I will turn it over to Kevin to provide more details on third quarter results, including sale leaseback and restructuring actions.

Thank you Mary. In the third quarter we spent considerable time evaluating financing and sale leaseback opportunities for our Beverly facility. On October 6th we announced the signing of a $50 million sale leaseback agreement with Middleton Partners which is expected to close in Q4. We also took swift action to lower our cash and operating breakeven levels by approximately $10 million per quarter, through headcount reductions and other discretionary spending cuts.

Due to the high degree of Purion platform commonality, we were able to proactively reduce R&D expenses, while keeping key strategic initiatives and funding in place. In addition, we streamlined our customer facing operations to improve customer satisfaction and reduce SG&A costs. In Q3 our results were consistent with guidance, but it was a challenging quarter, due to the continuing industry pause. During the quarter all of our tools shipped to foundry logic customers the same as in Q2, reflecting the magnitude of the pause in memory spending.

Turning to the details of our third quarter results, revenue finished at $38.5 million in the low end of guidance. System sales were $8.4 million, down from $10.9 million in Q2. GSS revenue finished mostly flat at $30.1 million, compared to $30.3 million in Q2. Q3 sales to our Top 10 customers accounted for about 65% of our total sales, compared to 67% in Q2, with two of these customers at 10% or above. Q3 system bookings were $15.9 million compared to $9.2 million in Q2. And our book-to-bill ratio almost doubled to 1.83 compared to 0.95 in Q2. Combined Q3 SG&A and R&D spending was $18.2 million, slightly above our guidance of $17 million to $18 million. SG&A for the quarter was $10.3 million with R&D at $8 million.

Moving forward we expect our SG&A and R&D run-rate to be below $18 million per quarter, as we realize the full impact of Q3 restructuring actions. Q3 gross margin finished at 39.3%, above our guidance of mid-30's which was driven primarily by a higher mix of GSS. Q3 operating loss of $5.4 million was within guidance including higher than forecasts restructuring costs related to headcount reductions. Our operating breakeven level was $46.4 million, compared to $59.9 million in Q2, driven by significantly lower expenses and higher gross margins. Net loss for the quarter was $4.7 million, with a loss of $0.04 per share and better than mid-point of guidance. Loss per share was increased by $0.02 of restructuring costs, partially offset by $0.-1 of favorable FX.

Q3 inventory ended at $107.8 million compared to $106.7 million in Q2, up slightly due to Purion material purchases. Q3 Accounts Payables were $14.7 million, compared to $17.4 million in Q2. Q3 Receivables were $29.9 million compared to $32.1 million in Q2. Q3 cash and cash equivalents finished at $32.5 million and within our guidance. In summary, our actions to lower operating expenses will provide additional financial strength and leverage to our business model moving forward. Closure of the sale leaseback in Q4 will provide significant liquidity to the business, with proceeds from the sale used to payoff the $15 million mortgage from Northern Bank and Trust, and the remainder of cash used to strengthen our balance sheet. Gross margins in Q4 will be negatively impacted by approximately 450 basis points, due to higher revenue deferrals associated with the timing of shipments, and delivery of our first Purion H revenue tools.

Lower gross margins are typical in our first units shipped due to the unique customer requirements and upgrades associated with the product development cycle. Our detailed gross margin improvement roadmaps drive higher system margins over time, as volume increases and Lean initiatives are implemented. Overall, we feel we are well-positioned from both a product and financial point of view to grow share during 2015. Doug will now provide some thoughts on the 2015 cycle and Purion's technical advantages. Thank you. Doug.

Thank you Kevin. We expect 2015 to be a solid year for capital spending relative to the implant market. There are two Greenfield DRAM factories under construction that Axcelis will participate in. Both are expected to ramp an initial phase of between 30,000 and 50,000 wafer starts, the first project will ramp in Q1 into Q2, and the second is expected to ramp more slowly beginning in Q2 and accelerate in the second half of 2015. Axcelis has already received initial orders for the first project for both the Purion XE and the Purion H.

NAND is expected to ramp beginning in the middle of the year, and continue through the second half. Axcelis is well-positioned to support this ramp, especially due to the more intensive high energy requirements of NAND devices. Leading Edge foundry is forecasted to ramp 14 and 16 nanometer FinFET processors during 2015. We expect this to be second half-weighted due to yield issues, and the challenges associated with this difficult technology. Axcelis is targeting placement of Purion H evaluation units in leading-edge foundry customers in 2015. The Purion H magnetically scanned spot beam architecture offers significant technical advantages for this node.

Mature foundry and logic customers producing 28-nanometer and older logic MEMS, image sensor and power devices, will also be adding capacity throughout 2015. The mobile, wearables, automotive, and Internet of Things markets will drive demand in this segment. This market is less predictable to forecast, but these customers face challenges of continuous product and process mix changes, often within a four wall constrained factory. The productivity advantages of the Purion product family are highly valued in this environment. This view supports a more gradual but sustained up-cycle compared to the short but steep 2011 upturn. We believe the shape of this recovery offers more opportunity for Axcelis, as we continue to roll out the Purion product family across all market segments.

Now I would like to discuss a little history regarding the technical advantages of the magnetically scanned spot beam. Throughout the nearly 40 year history of semiconductor ION implants, a spot beam has been the preferred beam line architecture for all types of implanters, high energy, medium current, and high current. The physics of the spot beam provide for superior across the wafer uniformity of the three critical implants parameters, energy, dose, and angle. Until 2004, the spot beam dominated the high current market holding approximately 80% share, compared with the competing ribbon beam. At that time customers discovered a particle problem related to the batch end station used on the spot beam based high current implanters.

To address this problem customers migrated to the only high current implanter using a single wafer end station. This implanter utilized a ribbon beam architecture. The ribbon beam suffers from uniformity problems inherent in the physics of its design. These uniformity issues are becoming more problematic for customers on advanced nodes, as control of dose, energy, and angle become more critical. Axcelis has developed a magnetically scanned spot beam allowing the high current implanter to once again benefit from the spot beam capabilities. In the Purion H, this new magnetically scanned spot beam joins with the common Purion end station, to give customers the uniformity and control they sacrificed ten years ago, but require more than ever on today's advanced devices. Now I will turn the call back to Mary.

Thank you, Doug. The bottom line is that we believe that we are executing against the right strategy, and with the pause thawing, we are positioned to begin to grow our market share. Our innovative Purion products provide a very competitive portfolio that will grow our top line. We expect that we will increase our share to greater than 30% of the implant market over the next three to five years, based on the strength of the Purion H, the Purion M, and the Purion XE.

Our SG&A and R&D run-rate is now below $18 million per quarter, which allows us to continue to invest in our future, while protecting profitability and cash. A sale leaseback will strengthen our balance sheet and provide stability for our customers, employees, and investors. We have a talented organization driven by measurements, and a culture focused on execution and success, the management team is committed and incentivized to meet both short and long-term objectives.

We have had a Board refreshment process underway over the last several years. We have added three very qualified Board Members to replace the three senior Directors who will not stand for re-election in May 2015. We are currently recruiting for one to two new Board Members, and have reached out to our shareholders for their input into the recruiting process. We continue to work with Blackstone Advisory Partners, to identify strategic opportunities to strengthen our long-term position in the semiconductor industry. And perhaps most importantly, our customers recognize our advantages, and are supportive of our success.

Customers are realizing that new emerging implant applications are best served by using our innovative Purion products. Our customer satisfaction as measured by VLSI Research is rated highest in the industry above all of our competitors. We are well-positioned to meet our goals for 2015, which are very simply to gain share with all three of our Purion products. The recent multi-system revenue sale is an exciting indicator of this opportunity in 2015. By increasing top line sales, holding the SG&A and R&D run-rate below $18 million, and strengthening our gross margins through commonality of the Purion platform, we expect to have a business that delivers improving profitability and cash for our shareholders. With that, I would like to open it up for questions.

(Operator Instructions). Our first question comes from the line of Patrick Ho from Stifel.

Thank you very much. Mary, can you give a little I guess color in terms of your exposure overall to NAND, and how some of the recent planer conversions and potentially the push-outs on 3-D NAND, impacts both I guess your XE as well as some of the evaluations that you have been doing for both M and H? How does some of the shifting on the NAND side impact you?

Okay. Patrick, I'm going to let Doug take that.

So Patrick, the NAND we've been expecting to be later in Q2 and into the second half kind of all along, so it's really not changing anything relative to our plans. The DRAM build in the first half has kind of been the way we have seen it. Leading me, when we discuss the shape or slope of this upturn, we see the DRAM stuff happening initially, NAND coming in second, and then the foundry filling in towards the end of the year, obviously with some overlap.

Right. That's helpful. And maybe just going to comments regarding 28-nanometers and some of the more mature technology nodes. Now that some of the second tier players are starting to ramp-up, or get their plans for 28-nanometer capacity for 2015, how does their projected spending plans and their build-outs impact you? Did their delays or their timing give you guys more opportunities potentially to break in at those second tier accounts?

Patrick, we have said all along that having more time to get our Purion products qualified has been a good thing. So we continue to work with all of those customers to get really the full portfolio of our products qualified. So I would say yes, that having a delay at those customers at the 28-nanometer node in particular, will give us an opportunity to potentially have a larger opportunity than we would have had.

So, Patrick, let me and to that. That market is a little less predictable, and so it's more difficult to plan. It's not like one of the large memory companies building a greenfield, where you have an idea of exactly when they're being to build and how big. Having said that, the second tier market really looks at Axcelis from three perspectives. First if they're looking for productivity improvements the Purion really can bring that to them. If they're looking to get a tool quickly that they don't have to re-qualify, they may buy an Axcelis legacy tool that's the same as what they have in their fab to ramp quickly. Or if they're more concerned with cost, then they would look on the used tool market, and so in all three areas, Axcelis offers opportunity, or there's opportunity for Axcelis to participate with all of that second tier.

Right. And final question. Maybe for Kevin in terms of the variables on the gross margin line. I understand how it impacts Q4. Given that you're still kind of in the evaluation phase, particularly with the Purion H, what kind of time frame are you looking at for the H to eventually equal I guess the XE margins, and meet the corporate average that you have targeted? How much of a time roadmap are you looking at?

Patrick, I guess the two variables that are driving the gross margin overall will be the systems mix to GSS, because as you know we have talked about GSS being very accretive, and then obviously the mix within the systems, which is I think what you're poking at right now. Our gross margin roadmaps are fairly aggressive, in terms of driving the costs out of these tools and quickly improving things, but if you are looking for a sense of timing, we have been talking about a model in the mid-30's, and obviously, I think you have just heard, we have got a little bit of pressure going in this quarter. I would think probably the first half of next year, we will continue to see pressure as we roll out more M's and H's, and then I think looking at your overall model 35, you just want to slope it, use an average of 35 maybe a little bit tighter on our front end, a little bit more in the back end of the year. So the volume will drive a big component of that.

Great. That's helpful. Thank you very much.

(Operator Instructions). Our next question comes from the line of Christian Schwab from Craig-Hallum.

Hey guys. Congratulations on the seeing some good business improvement. Mary, what do you believe your market share is today in aggregate roughly in 2014?

I think our market share is probably around 10%. We said that we have been around that now on average, and I think in 2014, given the unfortunate pause that we have experienced the momentum that we came into the year with was obviously slowed down. So I would say we're probably around 10%, and as I mentioned in the script, we expect over the next three to five years to be able to get to 30%-plus market share.

Great. And then looking out three to five years what do you think that the TAM is going to be relatively stable at what we're at now? What had you expect would be an appropriate TAM expectation for ION implant three to five years out?

I think probably three to five years out it's probably slightly less than $1 billion, if you look at DataQuest and what they're forecasting right now. I think as you know we go into our strategic planning period, what we're probably look at is something maybe around $800 million. Now again, we have said this many times, that we believe that DataQuest is really only capturing traditional implant, and perhaps not taking into account the increase in implant steps that we're seeing in material modification. So I'm just sharing with you basically I guess the common wisdom at this point in time, based on data from an outside source, but we feel that potentially could be larger than the $800 million.

Great. And, Kevin, when you guys start ramping revenue a little bit more materially than what you are seeing currently, let's just assume at some point in the future, we have a call it a $300 million business, plus or minus, right? And gross margins at 35% plus or minus, how should we be thinking about OpEx at that type of run-rate? In other words, another way of asking it is how much corollary revenue can $18 million plus or minus support?

Yes. So I think the only difference you will see, Christian, at those higher levels would be in maybe some of the variable components. Commissions for sales people, and potentially bonus programs that maybe get funded, but in terms of core headcount that we have right now, I don't think we need to increase that. So the commonality on the parent platform is really starting to show the leverage. I think the fact that these two H's just shipped as quickly as they did, also shows the benefit of that commonality across that whole family of products. So at $300 million it's going to be a variable component of that, so I wouldn't say significant changes, the days of the heavy spending are gone. So we have modeled up a little bit, but again, core headcount will stay essentially flat with where we are.

And can you remind us, Kevin, what headcount is, and if you have it in front of you? Otherwise I can pull a K where was that say a year or two ago?

Headcount a year or two ago was probably close to 900, 950 employees, and our recent headcount after the restructuring that we just did is around 750 people.

Okay.

That's worldwide.

Yes. Absolutely. Great. Thank you. No other questions.

Alright. Thanks, Christian.

Thanks, Christian.

Our next question comes from the line of David Duley from Steelhead Securities.

Yes. Thanks for taking my questions. Just a clarification. What is your breakeven now in dollars, both I guess on a cash basis and on a GAAP basis, or however you would like to present it?

Yes. So what we have talked about, David, is we brought our breakeven level at mid-30% gross margins down to $50 million on an operating level, and on a cash breakeven level it's around $43 million. Again, that's using mid-30 gross margins and operating expenses, R&D and SG&A, slightly below $18 million a quarter. Actually you could model it $18 million a quarter and 35% gross margins are going to be right at the $50 million breakeven and we were $60 million prior to doing this most recent restructuring in Q3.

Okay. And I think you mentioned in your prepared remarks that your impact to gross margins in the December quarter from the revenue deferral and shipping first time systems to customers was 450 basis points?

Yes. It's approximately 450 basis points. There's a fairly large revenue deferral just because of the timing of the shipments. And then the two Purion H's are bringing a little bit lower gross margin through with them right now.

So I guess I was wondering if you could help us all understand because one is kind of a one-time thing. What is the impact of the revenue deferral? Because you're going to capture that some time in the future. So if we just kind of understand was it a couple, was it most of the 450 basis points or--?

It was a good portion of it, yes.

Okay. And I think you mentioned that the foundry ramp for you guys would be in the second half of 2015. Could you just talk about exactly, I guess I would have thought that maybe would have been earlier in the year for you, and maybe just talk about some of the dynamics that are going on around why you think that's the timing for Axcelis?

Okay. Well, I wasn't specifically talking about Axcelis when we said that. We feel that the memory market is ramping pretty aggressively in the first half, led by DRAM, followed by NAND. If you look at what all the foundry companies have talked about during their earnings call, they have talked about ramping FinFET in 2015. Our view based on what we hear, in terms of where yields are, and where they are in terms of qualifying process and so forth, just suggests to us that it's probably a little weighted towards the second half of the year.

For Axcelis our goal in 2015 is to grow Purion market share in the memory segment with all three of the Purion products, and to penetrate the foundries with Purion H evaluation. So we would expect to continue to sell Purion XEs, in foundries for image sensor applications, power device applications, and so forth, but our real goal is to penetrate evaluation units into the leading-edge, and so that could happen any time. It's not really dependent on that ramp, but the comments that we made in terms of the timing, is just based on where we feel the industry is at this point.

Okay. And maybe just another clarification there. You do have an evaluation with a Purion M or H at a major foundry going on now, correct?

No. All of the evaluations regarding Purion and foundries are closed. So Purion M is in a foundry in production. There are multiple Purion XEs in foundries that are running in production, and so our next goal evaluation-wise is to penetrate with a Purion H high current tool into a leading-edge foundry.

Okay. And would you expect that when they do start to spend more significantly, would you expect your Purion M that's in production to ramp-up more than it is now?

Yes. We would expect Purion M will ramp across the different market segments. It currently is in three customers running in production for both memory and foundry processes, and so as they ramp and need additional medium current capacity, we would expect to participate in that.

Okay. Thank you.

Thanks.

Okay. Thanks, Dave.

Our next question comes from the line of Christian Schwab from Craig-Hallum.

Hey, guys. Just a quick follow-up. At $300 million in product revenues plus or minus similar to 2011, is there any changes on a go forward basis that would materially change services revenue versus last time you were there? In other words, is that going to be kind of like a $32 million , $33 million business plus or minus at that point at that type of revenue, or are there some positive or negative puts and takes to that number?

I think that is fair. Right now if you take a look back on the last 12 to 18 months we have been running around $30 million, and obviously as we move forward, we have a number of initiatives in place to grow revenue. We think we have made significant progress in some areas like spares win backs, spares and consumables win backs, upgrades, we have sold many more used tools than we ever have in the past. So we're going to continue to focus on a number of those initiatives and would expect to grow it, but I think if you want to model it in like a low-$30 million range, that's probably fair.

I think the other thing, Christian, is longer-term as we continue to sell more Purion, refreshing the installed base, also contributes to the service business over time. So if you look out depending on what you're target $300 million number is, and you look out a little further than that, then we start to see benefits from that installed base growth.

Great. Great. And are we doing, are there any further adjustments, I know you have guaranteed up time in all of your tools, and are in short order for very obvious reasons. As we look at that inventory number, just remind me just as we think about working capital, how much of that inventory is positioned around the world for guaranteed up time?

About a third of it. About a third of that number.

Remember, evals still sit in that number as well.

Right. Right. Right. Right. Perfect. Thank you.

Okay.

Thank you, Christian.

Thank you, Christian.

Alright. Our next question comes from the line of Edwin Mok from Needham & Company.

Hey. Thanks for taking my question, sorry I jumped on a little late. I apologize if you guys already touched on this, but on the sequential increase in guidance, how much of that is increased demand from your customer, and how much of that is coming from this new Purion H win?

Well, we don't split our revenue out by product, but we referenced in our prepared remarks that it's being driven by this buying, in memory spend, in particular the first phase of a DRAM ramp at a major customer in Asia-Pacific. So I think it is safe to say that there's a fair amount of our systems revenue coming from that customer, and it's not just that customer. There are at least two other customers that we're also getting some significant business from.

So Edwin, I think a little bit of follow-on to that, is so as we have been saying as the memory business recovers we, given the pause, we expect all three products will participate, and so I think the other way to answer your question is, the memory market is coming back to life, at a time when all three of our products are available, and so the combination of the two is what will impact our business projections going forward.

I see. And then on the booking that you guys had the press release I guess two days ago, right? Are all those products shipping in the fourth quarter, or is that a booking for multiple quarters that we expect to see more shipments in first half of 2015, or something like that?

No. Those are Q4 shipments.

Yes.

I see. Okay. That's helpful. And then actually a question on your Purion XE, right? So for this year we have seen more, actually incremental order you guys got from a couple of foundry customers, which traditionally only buy it for image sensor application accounts, a little bit of a niche application. Are you guys seeing increased demand or an increased number of applications on that XE platform, or is it just those kinds of applications that you guys typically see, just because of customers investing in this area? I'm justifying to understand, is it a broadening of application, or is it just more customer and building capacity?

It's a combination. There is a broadening of application. The more advanced image sensors are requiring lower metals contamination and more high energy implants, and the power device, is our power device market is also beginning to require more high energy, especially in the automotive side of power devices. So there is a bit of growth in terms of where high energy is being used, and most of that is coming from custom logic, or specialty logic, and their foundry partners.

I see. Okay. That's fair. And then just a quick follow-up question on the margin side. Actually last quarter you guys have actually high product gross margin, right, or probably high system gross margin as well. And I understand this quarter because of Purion H initial ramp, you always have some drag on gross margin. How can we kind of think about gross margin longer-term, assuming the X ramp, and you have multiple customers buying that product, together with a healthier year spending environment? Do you expect your margin to actually march back up to this, last quarter you did 39% overall comparable gross margin, right? Do you expect that too, and how fast can you actually get it back up to that range, I guess?

Yes. So Edwin, the Q3 39.3% was really driven by a fairly significant mix of the service side, which is really where all of the spare parts and consumables are. And GSS is very accretive, so that was a big piece of it. But in terms of how we think about it, as we roll out more systems, because the GSS is more accretive, that will obviously pressure the margins a little bit, but we have always kind of said that we would ramp up this Purion product line, and hold our margins in that mid-30's and then the roadmaps take us back to a 40-plus% gross margin business, but if I was looking out at, obviously we're not guiding next year, but if I was looking at next year, I would say the mid-30's is still the right place, because any growth we have is going to come from the systems, and just the fact that we would hold in the mid-30's without serious erosion with systems coming, says that the systems are continuing to improve as the year goes on.

Okay. Great. Actually that is good color. Thanks. That's all I have.

Alright. Thanks, Edwin.

Thank you.

And we have a follow-up question coming from the line of Christian Schwab from Craig-Hallum. Please proceed.

Sorry. I keep playing with my models. Just a follow-up on Edwin's comments. Maybe I will just ask it a little bit more directly. If we get to $300 million in revenue at some point down the road, and we have $30 million, $35 million in services revenue, is the gross margin objective to have 35% gross margins plus or minus, or is it somewhere between 35% and 40%?

When we get to 38%, the answer is somewhere between 35% and 40%.

Okay. Somewhere in between there, 40% would be heroic, and 37.5% would be kind of what you would love to do, and 35% at that point would be a disappointment, is that fair?

At $300 million, yes.

Great. No other questions. I promise. Thanks.

Alright.

This concludes the Q&A portion of the call. I will now turn our call back over to Mary Puma, who will make a few closing remarks.

Thanks Philip. We spent a great deal of time with investors on the phone and in meetings both in Beverly and on the road, and we have scheduled a number of non-deal road shows, as well as conferences. We will be attending the Midtown Cap Conference in New York City in December, the Needham Conference in New York City in January, and the Stifel Conference in San Francisco in February. We are looking forward to catching up with all of you. Thank you.

This concludes the presentation. Thank you for your participation in today's conference. You may now all disconnect. Good day.