Axcelis Technologies - Earnings Call - Q4 2024

Executive Summary

• Q4 2024 revenue $252.4M and diluted EPS $1.54; gross margin 46.0% and operating margin 21.6%—both ahead of internal outlook due to stronger CS&I upgrades and mix.
• Company beat prior Q4 guidance of ~$245M revenue and ~$1.25 EPS; margin outperformance was the key driver—EPS came in $0.29 above guidance and revenue ~$7.4M above.
• 2025 setup: management guides Q1 2025 revenue ~$185M, EPS ~$0.38, GM ~40% (low point for the year), OpEx ~$63M, tax rate ~15%; expects Q2 similar to Q1 and slight H2 improvement as China digestion abates and mix improves.
• Subsequent capital return catalyst: Board approved an additional $100M to share repurchase authorization (now $215M total), intending to increase quarterly buybacks while maintaining a strong balance sheet.

What Went Well and What Went Wrong

• What Went Well

  • CS&I strength and upgrades: Q4 systems revenue $187.4M and CS&I $65.0M, with better-than-expected CS&I driving margin and EPS beat; GM 46% vs 42.5% outlook and EPS $1.54 vs $1.25 outlook.
  • Mix and geographic resilience: China was 49% of shipped system sales and Korea improved to 11% on DRAM shipments; bookings stabilized at $84.5M and backlog ended at $646M.
  • Management tone on secular drivers: “Exited the year on a strong note… anticipate near term cyclical digestion… focused on long-term growth” (CEO); CFO highlighted higher full-year GM (44.7%) despite lower revenue and strong balance sheet with $571M cash/ST investments and no debt.

• What Went Wrong

  • YoY contraction: Q4 revenue fell to $252.4M from $310.3M in Q4 2023; diluted EPS declined to $1.54 from $2.15 YoY, reflecting mature node softness and China digestion.
  • Mature node and IGBT softness: Power mix fell sequentially to 51% (from 57% in Q3) and general mature expected to decline sequentially in Q1; silicon IGBT is soft into 2025.
  • Q1 reset and margin trough: Q1 2025 guide implies volume/mix headwinds (GM ~40%, CS&I seasonality, lower China systems); CFO called Q1 GM the year’s low point, with gradual improvement thereafter.

Transcript

Operator (participant)

Good day, ladies and gentlemen, and welcome to the Axcelis Technologies call to discuss the company's results for the fourth quarter and full year 2024. My name is Dee Dee, and I will be your coordinator for today. I would now like to turn the presentation over to your host for today's call, David Ryzhik, Senior Vice President of Investor Relations and Corporate Strategy.

David Ryzhik (SVP of Investor Relations and Corporate Strategy)

Thank you, Operator. This is David Ryzhik, Senior Vice President of Investor Relations and Corporate Strategy. And with me today is Russell Low, President and CEO, and Jamie Coogan, Executive Vice President and CFO. If you have not seen a copy of our press release issued yesterday, it is available on our website. In addition, we have prepared slides accompanying today's call, and you can find those on our website as well. 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. Now, I'll turn the call over to President and CEO Russell Low. Russell.

Russell Low (President and CEO)

Good morning, and thank you for joining us for our fourth quarter and full year 2024 results earnings call. Beginning on slide number three, we exit the year on a high note with revenue for the fourth quarter at $252 million and earnings per diluted share of $1.54. Revenue was slightly better than our expectations, as strong demand for our aftermarket CS&I sales partially offset the anticipated sequential decline in system sales. This strength in our CS&I business was the primary driver behind our better-than-expected margins and EPS in the quarter. Within our system sales, a sequential decline in power and image sensors was offset by an improvement in general mature and memory sales. Our backlog declined during the quarter, but remained at healthy levels. We also saw bookings stabilize, which came in flat on a sequential basis.

Turning to slide four, in the quarter as well as for the full year, sales to mature node applications remained the lion's share of our business, in particular power and general mature. Now, let me review shipped system revenue by end market, and I will begin with mature nodes on slide five. Revenue from our power market was 51% of our mix, down sequentially from 57% in Q3 2024. Shipments to silicon carbide applications moderated slightly in the fourth quarter. However, on a full year 2024 basis, our system sales for silicon carbide grew approximately 6% year over year. Over the past several years, Axcelis has established itself as a market and technology leader in ion implantation for silicon carbide, one of the defining process steps in device manufacturing. We were first to identify this emerging opportunity several years ago and quickly leveraged our Purion platform to drive necessary innovation.

This included the development of a differentiated medium current implanter, followed by an extension of capabilities to our high-energy tools, close collaboration with customers to understand their production needs, and finally, we launched our high current implanter optimized for silicon carbide. This all translated into our system shipments to silicon carbide growing from approximately $8 million in 2020 to over $300 million in 2024, and while we expect revenue from silicon carbide to decline sequentially in the first quarter of 2025, as customers undergo a digestion period, the fundamental long-term drivers remain intact, as we expect adoption of silicon carbide to continue to increase, particularly as costs come down and new applications become economically viable. A case in point, we expect the EV industry's transition from 400 V to 800 V architecture to greatly improve charging times, and this will require silicon carbide.

We're also closely monitoring power applications in the data center, where the demand for energy is rising rapidly, and silicon carbide can be used to deliver more power more efficiently. But our engagement with customers is not just confined to addressing their capacity needs. We are deeply embedded with customers on their technology roadmaps, which include the transition from 150 mm to 200 mm, our wafer size, the transition from planar to trench MOSFETs, the transition from trench to superjunction, and some customers are even exploring wafer splitting applications to improve yield and lower cost. In all of these cases, Axcelis is a key enabler, and we believe the need for higher performance devices with higher yield and lower cost will only unlock new opportunities for silicon carbide in power applications. We believe we're in the early stages of the silicon carbide market growth.

Turning to silicon IGBT, system sales declined in the fourth quarter, and we anticipate this market to continue to soften in 2025 as our customers continue to work on managing capacity amidst a slower-than-expected industrial auto recovery. In general mature, revenue increased sequentially in the fourth quarter, led by investments in China, while other regions remained muted. As a reminder, general mature represents a broad array of semiconductor applications requiring a 28 nm process node or above. This includes RF, analog, microcontrollers, and other semiconductor applications. We continue to monitor key end markets, namely auto, industrial, and consumer, which generally are drivers of our general mature segment. Given recent industry commentary of a slower-than-expected recovery in the auto and industrial markets, along with an anticipated digestion of mature node capacity in China, we expect our general mature revenue to decline sequentially in the first quarter.

Over the long term, as inventory levels normalize and demand recovers in key end markets, we anticipate our general mature business to benefit accordingly as ion implant intensity is particularly high for process nodes at 28 nm and above. Turning to image sensors, as we anticipated, revenue moderated in the fourth quarter, following a large customer order in China in the third quarter. Image sensor production will continue to rely on a large part on smartphone volumes, but also to a lesser extent on auto, as we are seeing increased camera content in autos. As we think about the first quarter, we expect image sensor revenue to be flattish on a sequential basis.

Turning to slide six, in advanced logic, we shipped a system to a new advanced logic customer in the fourth quarter, following a previously announced order received in the second quarter, and we had discussions for a follow-on order. We continue to work actively with customers, as well as with a leading European advanced logic research center, in understanding next-generation advanced logic applications for ion implantation. Growing footprint within the advanced logic market is a strategic goal of ours, which is a multi-year initiative, and we are still in the relatively early stages. Moving to memory, as we anticipated, we saw a sequential improvement in sales to memory market, specifically for DRAM. Looking ahead to the first quarter, we expect sales to memory to be relatively consistent on a sequential basis, entirely in DRAM.

In NAND, we believe customers have ample capacity given current demand trends and expect this to remain the case in 2025. As we think about our memory business over the long term, we are quite excited about the opportunity, both in DRAM and NAND, given the following drivers: one, growth in AI and its structural impact on high bandwidth memory, which is absorbing DRAM capacity. In fact, not only are AI server unit volumes expected to grow significantly, but HBM content per server is also expected to grow, enabling a multiplier effect on HBM capacity. Two, AI's impact on new data creation, particularly with inference, whereby new data sets need to be manipulated and stored, which we believe will be a tailwind for DRAM and NAND.

Three, rising memory and storage content in smartphones, servers, and PCs as devices need to process and store more data, and four, device volume growth resulting from improved macro and potential refresh cycles. We believe that the confluence of these catalysts translates into an attractive long-term market for Axcelis, and while a market recovery is instrumental to any growth in our memory business, we are not standing still. We are focused on penetrating new customer opportunities within memory, where we historically had a low share, and I'm pleased to say we've had some initial progress in this regard. Turning to slide seven, as we look back on 2024, I am proud of how our team executed amidst a dynamic demand environment. For the full year we saw continued growth in sales to silicon carbide, while silicon IGBT softened considerably. In memory, demand remained soft as customers navigated through lower utilizations.

Despite this, we focused on what we can control, and this included working closely with our customers to enable their technology and production roadmaps, placing evaluation units into the field, seeding new opportunities, continuing to invest in our R&D to maintain our robust pace of innovation, and maintaining strong margins due to favorable mix and cost control. In fact, despite a year-over-year decline in revenue, we grew our gross margins by more than 100 basis points. On slide eight, let me now discuss some of our initial perspectives on 2025. We anticipate overall revenue in 2025 to decline on a year-over-year basis. As we think about the trends by segment, we expect a digestion of capacity in the power and general mature markets, primarily in China. In memory, we expect year-over-year growth in 2025, specifically tied to DRAM investments, while NAND remains muted.

And we expect modest revenue from our initiatives in advanced logic, consistent with our expectations of being in the early stages of a multi-year growth effort. In summary, while the near-term demand backdrop is muted, the fundamental long-term drivers of our business remain intact, namely long-term secular growth in power, particularly silicon carbide, which we believe will continue to proliferate within existing and new applications, given the world's insatiable demand for more power and greater efficiency. Market recovery in memory and general mature, share again in advanced logic, and geographic expansion into Japan, which is a sizable market for ion implantation, where we have relatively low penetration. As a result, we are taking actions today to increase our technology engagement with customers to help accelerate their roadmaps.

On that note, before I hand over to Jamie, as you can see in slide nine, I'm particularly proud of the Axcelis team and the recognition we've received from customers in 2024. We received 22 customer awards covering overall supplier excellence to support safety, health, and others, and this represents a significant increase compared to 2023. The core of our culture at Axcelis is customer first, then company, and then self, and this is a shining endorsement of how we operate. With that, let me turn the call over to Jamie for a closer look at our results and outlook. Jamie.

Jamie Coogan (EVP and CFO)

Thank you, Russell, and good morning, everyone. I'll first start with some additional detail on our fourth quarter and full year results before turning to our outlook for Q1.

Starting on slide 10, fourth quarter revenue was $252.4 million, with systems revenue at $187.4 million and CS&I at $65 million. This was above our outlook, largely driven by better-than-expected CS&I sales. As a reminder, CS&I is driven by our installed base and represents consumables, spares, services, and upgrades. In the quarter, we saw stronger upgrade activity as customers are looking for ways to enhance their technology within the same factory footprint. We also executed well on service contracts. From a geographic perspective, China remained our strongest region at 49% of total shipped system sales, with the sequential decline quarter over quarter, primarily due to an anticipated decline in image sensor following a large order in the third quarter, as well as a moderation in sales to the power market.

On the other hand, we saw system sales to Korea improved to 11% in the fourth quarter, compared to only 1% in the third quarter, mainly due to improved shipments in memory. Bookings in the fourth quarter were $84.5 million, or flat on a sequential basis, while backlog exiting the year was $646 million. On a full year basis, 2024 revenue totaled $1.02 billion, consisting of $783 million in systems revenue and $235 million in CS&I revenue. Turning to slide 11 for additional detail on the fourth quarter and full year results, gross margins in the fourth quarter was 46%, which exceeded our outlook of 42.5%, driven primarily by stronger-than-expected CS&I revenue, which carries higher-than-corporate average margins. Operating expenses totaled $61.7 million, slightly above our outlook of $60 million, partly due to higher variable compensation associated with our stronger performance.

As a result, operating profit was $54.5 million, reflecting a 21.6% operating margin. We generated approximately $4.1 million in other income, a sequential decline due to an FX gain we saw in the third quarter. Our tax rate in Q4 was 15%, in line with our outlook. Our weighted average diluted share count in the quarter was 32.5 million shares. This all translates into diluted earnings per share of $1.54, which exceeded our outlook of $1.25. The higher-than-expected EPS was primarily due to better-than-expected revenue and gross margins. For the full year, we delivered gross margins of 44.7%, a 120 basis points increase year-over-year, despite lower revenue volume. This was due to favorable mix and the continued focus on cost control. Operating expenses for the full year of 2024 were $244 million, and operating income was $211 million, translating into an operating margin of 20.7%.

Our full year tax rate was 13%, and our full year diluted earnings per share were $6.15. Moving to our cash flow and balance sheet, we generated $8 million of free cash flow in the quarter. The lower cash flow in the quarter was primarily due to the timing of cash receipts associated with deliveries in the fourth quarter. In the quarter, we repurchased $15 million of shares and exited the fourth quarter with $130 million remaining in share repurchase authorization. For the full year, we repurchased $60 million in shares, which amounted to 47% of our free cash flow. We exited the year with a strong balance sheet consisting of $571 million of cash, cash equivalents, and short-term investments on hand, with no debt. This provides a solid foundation for our capital allocation strategy, which falls into three main categories. First, continued organic investment.

Our strong cash position allows us to continue to invest in product innovation, despite the near-term digestion in some of our end markets. While the semi industry has had many cycles in its history, the overarching trend is one of strong secular growth, and the current environment is a great opportunity to increase our engagement with our customers on their technology roadmaps as we work to best position the company as and when markets return to growth. Second, our strong cash position allows us to continue to execute on our buyback program, which more than offsets the dilution from equity compensation. For historical context, over the past five years, we've repurchased more than $200 million in shares. Third, we continue to evaluate opportunities for inorganic growth. We remain disciplined in our approach and consider opportunities only if they deliver sustainable long-term shareholder value creation.

Before I move to our outlook, I'd like to discuss a few reporting changes, beginning with our first quarter 2025 report. Following a thorough review of our peers, we've decided to add non-GAAP measures as part of our quarterly reporting process. We believe this enhanced layer of disclosure will improve transparency on the underlying performance of the business. This will also help us align with the practices of our peer group, which can result in an easier benchmarking process by our analysts and investors. Second, starting in the first quarter, we will begin including image sensor revenue as part of our general mature category. Given the relatively small size of image sensor business, we believe it's a logical fit within our general mature category, which already consists of a broad array of applications. We believe this further simplifies our disclosures.

With that, let me discuss our first quarter outlook on slide 14. We expect revenue in the first quarter of approximately $185 million. The sequential decline is primarily a result of lower systems revenue from China customers for the power and general mature applications, as well as a seasonal decline in our CS&I revenue. As we think about the balance of the year, we expect revenue in the second quarter to be relatively consistent with the first quarter, and based on our discussions with customers and our view into our current backlog, we anticipate that revenue will improve slightly in the second half compared to the first. Turning to gross margins, we expect first quarter gross margins to be approximately 40%. The primary driver of lower gross margin is lower overall volumes, as well as anticipated mix.

While gross margin in one quarter can be dictated by a variety of factors, we expect gross margin in the first quarter to be the low point of the year, and we anticipate a gradual sequential improvement resulting from mix and our continued cost controls flowing through over the balance of the year. We expect first quarter operating expenses of approximately $63 million, with a slight sequential increase resulting from the seasonal increase in payroll taxes. For the full year, we anticipate operating expenses to be relatively flat on a year-over-year basis as we continue to manage our cost structure with discipline while ensuring we are making the necessary investments to capture the long-term growth opportunities that lie ahead. We expect our tax rate for the first quarter and the full year to be approximately 15%.

This all translates into an estimated diluted earnings per share in the first quarter of approximately $0.38. Finally, on January 10th, we filed an 8-K discussing our preliminary review of the new restrictions put in place by the U.S. government on December 2nd, 2024, and at that time, we estimated an approximately $20 million-$50 million impact to our revenue to China in 2025. We now estimate the full year impact to be closer towards the low end of that range, and this is factored into our outlook. In summary, we are pleased with our performance in 2024. Despite a decline in revenue, we were able to deliver higher gross margins, generate solid free cash flow, return capital to shareholders via our existing $200 million stock buyback program, and exit the year with a stronger balance sheet than we did coming into it.

With that, let me hand the call back to Russell for closing remarks. Russell?

Russell Low (President and CEO)

Thank you, Jamie. We exited 2024 on a strong note and are focused on capturing the growth opportunities that lie ahead. We are investing in innovation, managing our costs, and working closely with customers on their technology roadmaps, which we believe will help us in an even stronger position for the next upturn. In closing, I want to thank our customers, employees, shareholders, and partners for their continued support and trust in Axcelis. With that, operator, we are ready to take your questions.

Operator (participant)

Thank you. As a reminder, to ask a question, please press star one one on your telephone and wait for your name to be announced. To withdraw your question, please press star one one again. In the interest of time, we ask that you please limit yourself to one question and one follow-up.

Please stand by while we compile the Q&A roster. Our first question comes from Charles Shi of Needham & Company. Your line is open.

Charles Shi (Senior Analyst)

Hi, good morning. Maybe the first question, I want to get a sense of your assessment, slightly better second half for the year. I heard you guiding, I mean, roughly speaking, flat-ish into Q2, but what are some of the factors you are seeing that could get your second half slightly higher than the first half? That's the first question. Thank you.

Russell Low (President and CEO)

Yeah, hi, Charles. It's Russell. Thank you for your question. So I think, why do we believe the second half is going to be stronger than the first half? So this view is based on our backlog, our conversations with customers, and our own internal work. So obviously, we're continuing to monitor bookings for the course of the year.

They seem to have actually stabilized, and I'd say at this stage of the quarter, we're actually ahead of bookings compared to, say, Q4. So that's giving us some encouragement there. Each and every one of the tools in the forecast now actually belongs to a project, a customer project. So with the turn of the year, customers have put their budgets in place. They're beginning to formalize and stabilize their plans for the year, and that's allowing us to work with the customers to make sure we understand their shipping plan. So I would say that we have fairly good confidence that the second half is going to be better than the first half.

Jamie Coogan (EVP and CFO)

Yeah, and just to add into that, although we're not giving specific revenue figures for 2025, Charles, and as I said in the prepared remarks, we do expect a slight uptick in the second half of 2025. But again, all the factors Russell just mentioned, our backlog, the customer conversations, it doesn't take a whole lot of incremental systems for us to sort of be in and around that $800 million number for the full year period.

Russell Low (President and CEO)

And then, Charles, as you know, the secular drivers are still very much intact. This is a cycle. We've been through many, many cycles. So we do expect the cycle to disappear, start to see growth again. And we'd like to think that 2026 will be a growth year for us as well.

So we'd like to see the second half of 2025 be better than the first half, and we'd like to see some of that strength continue into 2026.

Charles Shi (Senior Analyst)

Thanks for walking us through a few of the market assumptions behind the outlook. Maybe another question, maybe this is not a big part of the business. I do want to ask about memory. It sounds like you guys still think the vast majority of the memory revenue in 2025 is going to come from DRAM. But over the course of the last couple of weeks, we did hear a little bit more positive commentary from some of your peers on NAND side of the spending, but it sounds like you're not expecting a pickup in NAND.

I wonder if this is still a difference between greenfield versus no upgrade, or maybe there's some hope of maybe at some point you're going to see some upside in NAND. I just want to get your thoughts on that.

Russell Low (President and CEO)

Right, we do believe that NAND is still going to be very new in 2025. Yes, you're correct. What we're seeing in 2025, which, yeah, we are seeing an improved memory situation in 2025 relative to 2024, albeit 2024 is a low base. Thinking specifically about NAND, we only sell more implanters to memory in general, both DRAM and NAND, when they expand the number of wafers out. If they change the technology node, and in terms of NAND, they put more and more layers on, that might be great for dep/etch. It's not really helping the number of wafers out.

So right now, I'd say that customers in the quiet times, they use that time to do node changes, and those node changes may drive some revenue to our peers. But until we start to see more capacity, and bear in mind, a lot of these customers do have actually capacity planned, but until they start filling those new factories, we won't start to see the benefits of that.

Charles Shi (Senior Analyst)

Got it. Maybe a last question about export control. I recognize some of the expected China weakness probably has very little to do with export control, but you are actually seeing potentially the impact that could be at the low end of your previously guided range. I wonder what's the reason for that, and what do you see?

Why do you feel like you are able to actually ship some of the, I don't know if it's a product or service, but that you are able to sell a little bit more than you thought?

Jamie Coogan (EVP and CFO)

Yeah, no, Charles, thanks for the question on that. In the Form 8-K that we prepared, we ultimately that was a preliminary estimate based on our review of the rules. Out of an abundance of caution, we included some system shipments in that number to build us up to the high end. So you may recall the low end represented sort of the CS&I impact, and the high end included some incremental systems that could be at risk based on the interpretation of the rules.

Since then, we've received some incremental information and data that provides confidence in our ability to be able to deliver on those system shipments, which is why we're now predicting the impact to be closer to the low end of that range. And that low end of the range has already been baked into our guidance for the full year.

Charles Shi (Senior Analyst)

Thank you.

Operator (participant)

Thank you. Our next question comes from Craig Ellis of B. Riley Securities. Your line is open.

Craig Ellis (Director of Research and Senior Semiconductor and Capital Equipment Analyst)

Yeah, thanks for taking the question, and congratulations on 2024's gross margin performance, guys. I wanted to start just focused on some of the near-term dynamics. So what are the things that you're seeing in the business that might indicate that the first quarter would be a bottom for the digestion that's occurring? And alternatively, what might be indicating that that may play out more in the second quarter?

Related to digestion, since power and general mature digestion is mostly in China, does that mean there are some more positive things going on in other geos or just an easier base coming off of the second half of last year?

Russell Low (President and CEO)

Hey, Craig, thanks for the question. So I think we've said that we believe the first half will be lower than the second half. We haven't kind of broken it down into quarters at this point. We do think that, obviously, there's a lot of digestion in mature technologies in China. That digestion isn't necessarily just oversupply. In some cases, it's getting to grips with the technology, ramping, and making sure the yield holds together. So as you know, China certainly has ambitions to be very self-sufficient. They're not close to that at this stage.

So I think right now, really, it's all about kind of getting the technology under control before ramping. And on the side of China, I'd say that we've, it's been resilient, let's say. I'd say that we do see a business outside of China still actually holding together for 2025. And obviously, you've heard about automotive, and in general, IGBTs have been very soft. But I'd say, in general, we're actually seeing positive spots from the U.S., from Europe, from Korea, and other locations.

Jamie Coogan (EVP and CFO)

Yeah, just to add into that, right, I guess each of the, based on our customer conversations and what we're hearing, right, each of them are at a different point in their investment cycle, Craig. And so we've got some who are continuing to push ahead.

And we think about this in the vein of silicon carbide, right, that we still have customers outside of China who are pushing ahead with their investments and building out capacity. We have others who are being a little bit more cautious in the spend of those CapEx dollars in 2025. But that mix, as you noted, relative to where we were in 2024, 2025 outside of China is still going to be relatively resilient, as Russell noted. We are seeing a bit of weakness in that silicon IGBT market. And so I think that the level of digestion there is a little bit greater in the silicon IGBT space. A smaller portion of the business for us coming into 2025 relative to where we had been historically. And then, generally speaking, memory is, again, looks like there's some, again, continued uptick relative to where we were in 2024.

We do expect memory, specifically in DRAM, to be good for us.

Russell Low (President and CEO)

Yeah, and just kind of going on that. So we mentioned IGBT power is down. So when we look at our power business, in 2024, we were up in silicon carbide relative to 2023. And even into 2025, we believe that it's still going to be a resilient market for us. So it's the silicon IGBT bit that is soft. The silicon carbide part, it seems to be remaining resilient for us.

Craig Ellis (Director of Research and Senior Semiconductor and Capital Equipment Analyst)

That's really helpful color, guys. Jamie, I wanted to follow up with some of the full year color that was provided on the shape of potential revenues to see if you could provide some color, not precise guidance, but some color on what the contours of gross margin and OpEx would look like, especially on the former. Does it look like 2025 can be a year where the business realizes another material gain in gross margin year on year? Thank you.

Jamie Coogan (EVP and CFO)

Yeah, so coming into the quarter, we expect gross margins to be at the low point in Q1 here. There's some cost absorption flowing through as well as mix within the period and a little bit lower CS&I as we saw some nice volumes in CS&I in the fourth quarter, Craig, specifically in the upgrade space as customers were going through and getting those throughput efficiencies and better utilizing the factory footprint they have. That is not uncommon for us to see sort of a little bit of a tick up in the fourth quarter as folks are going through their budgets for the full year and trying to sort of clear some of those budgets out and stock them on the shelf. So we're going to see a little bit of offset here in Q1 on lower CS&I mix. We are expecting memory to be a little bit higher in Q1 as well, which is going to add to some of that gross margin pressure.

But ultimately, the team does a really nice job of trying to control cost relative to volume. And as those plans go into place and continue, we do expect the margins to uptick throughout the course of the year given the mix and the volume that we see in the back half of the year. On the OpEx front, right, I think it's important that we've talked about our capital allocation strategy a lot, and we've said time and again that our number one priority is organic growth and making sure that the business is positioned to be able to hit the long-term technology trends that our customers expect from us. So we're going to use 2025 really as an opportunity to continue to make those investments.

And so OpEx is going to be slightly higher as we see RD&E, our research, development, and engineering expenses as a percentage of sales probably uptick relative to where they were in 2024. We're going to continue to make investments in that space. We want to make sure that we're positioned to meet those customer ramp cycles that we know are coming just given the sort of transitory nature of what we believe is the current situation in the market. And really, a nice little proof point on this is if you look back at 2018, 2019, 2019 was a down year for us. We maintained OpEx at similar and consistent levels to what we had in that 2018 timeframe.

Not predicting that we're going to see the same level of growth, but in the periods that followed, we grew revenue by almost 300% coming out of 2019 given the fact that we positioned the business to be able to execute and hit the markets in the way that we needed to. So that's our plan on OpEx. As always, though, we'll manage costs. If this seems to be longer than normal, we'll continue to look at our cost structure as appropriate and manage cost accordingly as we move forward if this seems to be more prolonged than it currently is.

Craig Ellis (Director of Research and Senior Semiconductor and Capital Equipment Analyst)

Thanks for the granularity on those, guys.

Operator (participant)

Thank you. Our next question comes from Jed Dorsheimer of William Blair. Your line is open.

Jed Dorsheimer (Group Head–Energy and Power Technologies)

Hi, thanks for taking my question. And I guess first one, if we look at the silicon carbide business and we look at last year, roughly about 60 tools that you guys did, as you're looking forward, is China still about half of or was China roughly half of that? I'm just trying to gauge as you look into 2025, the total exposure. I know you talked about the 8-K and around trade, but I'm looking in absolute terms what that exposure may look like. And then I have a follow-up.

Russell Low (President and CEO)

Yeah, hey, Jed, it's Russell. So just to kind of recap on silicon carbide. So we were up in 2024 relative to 2023, and we're basically saying we might be down slightly in 2025, but it's like we say, it's still a robust business for us. I would say that we have exposure to every silicon carbide project globally. China is a piece of it.

Obviously, it's a very important piece to it, but we don't necessarily break out our silicon carbide out of power by region. What I would say is that when you talk about the export compliance issues, the most recent export compliance issues that were released on the 2nd of December, we have had the opportunity to have kind of a Q&A with the BIS guys to help clarify, and coming away from that, it's very apparent that it's very similar to the previous administration and the continuation of the trend, which is 300 mm equipment going into advanced memory and advanced logic, so the silicon carbide business at this stage, we do not believe will be impacted by that, so the silicon carbide, like I've said, it obviously varies from customer to customer.

We've talked about the low penetration of silicon carbide into EVs and the low penetration of EVs into the automotive industry. The Chinese definitely want to be self-sufficient. They do have some capacity. It's not a huge amount of capacity. And I think a lot of them are at the moment trying to optimize their architecture, the devices. They're looking to improve their yields. And I think you're going to see the digestion in silicon carbide that would occur in China is not necessarily because of oversupply. It might be the fact they haven't got the applications they're looking for qualified, like automotive, or they may not have got the yield and the processes fully wrung out.

Jed Dorsheimer (Group Head–Energy and Power Technologies)

That's helpful, Russell. Then just a two-part follow-up, if you will. I guess, and maybe for Jamie, what percentage of your backlog is secured by customer deposits?

Then, Russell, just as you mentioned on some of the DRAM and advanced logic, is there a technology trend that you've developed that we should be aware of in medium current and in high current that positions Axcelis vis-à-vis the competition of what we saw with high energy for silicon carbide? Thanks.

Jamie Coogan (EVP and CFO)

Yeah, Jed, thanks for the question on that. Again, I would say the majority of our backlog to new entrants into the market and new geographies into the space, that's where you're going to find our customer deposits as we sort of manage the collection risk associated with that. And the more traditional semiconductor customers that may be domestic or European-based, we're less likely to secure customer deposits on those just given the long tenure of our relationships and the strength of their balance sheets that they have in place.

We haven't broken out specifically that number to kind of give a frame of reference for that. But the backlog itself, I'd say it is a smaller portion of the backlog than you would imagine all things being equal.

Russell Low (President and CEO)

Hi, Jed. So just to clarify, so you were talking about, do we have a differentiation in our high current and medium current products that we believe will allow us to take market share in memory and logic? Was that the crux of your question?

Jed Dorsheimer (Group Head–Energy and Power Technologies)

Yeah. Your market share in high energy is materially different than medium current and high current. And so I'm just wondering, as you start focusing on these applications for medium and high current, has there been a recent development in the use of a linear accelerator, for example, or something that differentiates you vis-à-vis the competition for those other applications? Right. Absolutely.

Russell Low (President and CEO)

So advanced logic, for example, you're absolutely right. It doesn't use high energy. It really is high current and medium current. The real opportunities are in high current. That's where a lot of the sales are. And I would say that the front end has an established competitor there. So really, when we look at advanced logic, we're looking at new application in the middle of line and the end of line. So think about the kind of applications that would have been in the old back end, the metallization of, for example, backside power. So that's where we're looking to go. And a lot of that is material modification. So we are always looking to get very close to our customers, make sure we innovate with and solve their valuable problems. And that's exactly what we're doing.

In addition, I should point out that not only are we working with our customers on advanced logic problems, but we're also working with a very advanced institute in Europe to work out what are going to be the next set of issues as the architectures change. So we're always looking to see an inflection point where we can take advantage of that. And we've said in the past, we do have architectural differentiation in our products. And in some cases, we've been able to take really good advantage of that, like in our Purion M, where that really is architectural entitlement. Regarding memory, I think it's fair to say that the tool has been battle-hardened for memory. And what you see for memory is a bunch of specific applications. And in many cases, what looks like dedication of species.

And our tool has been optimized to really perform well under those situations. So basically, since it's been battle-tested and fielded, high current and actually medium current, but specifically high current, we're now looking to fan that out to other customers who are in that same memory segment.

Jed Dorsheimer (Group Head–Energy and Power Technologies)

Thank you.

Operator (participant)

Thank you. Our next question comes from Jack Egan of Charter Equity Research. Your line is open.

Jack Egan (Equity Research Analyst)

Great. Thanks for taking the question. I was hoping you could go over the big increase in CS&I. I mean, 17% sequential growth is pretty big for that business. So you mentioned that it was stronger upgrade activity and execution on some of your service contracts. Was that strength particularly pronounced in any end market or region?

Jamie Coogan (EVP and CFO)

Yeah. No. I mean, honestly, it was fairly broad-based in the fourth quarter. It was multiple customers across multiple regions where we saw the strength.

I think we did see people taking advantage. We've talked a lot about how as our customers start to slow down, they try to find ways. They take advantage of these points to find ways to improve the efficiency and yield and throughput of their devices and to optimize the current fab space that they're utilizing. And this is, I think, the yin and the yang to our business at the end of the day. We've increased the number of Purion products out into the field fairly materially over the last few years. The power of that installed base is what provides us the opportunity to take advantage of these cycles and get our team to go in specifically targeting upgrades as we have them available.

We focus on making sure that our research, development, engineering team is working on differentiated upgrades that provide efficiencies to our customers and to their fab space. So I think that's kind of what we were seeing here in the fourth quarter, as well as a little bit of the budgeting, as we talked about before. Some of our fab partners and customers were looking through and working through their budget opportunities for the period. Going into 2025, we've got some relative expectations for upgrades throughout the course of the year, although we are not currently forecasting the same level of upgrade activity in the first quarter of this year, which is why we are seeing a little bit of moderation in margin.

Upgrades typically provide our highest margin relative to the consolidated average, just given how valuable they can be to the customer at the end of the day.

Russell Low (President and CEO)

Yes, I can just kind of follow on from that. This is a very focused strategy. We have been investing heavily in upgrades because it's a great opportunity for us. And the good thing about an upgrade is you get to sell it to the entire installed base. And then, obviously, we've been making progress on contracts as well as we try to look to kind of create an annuity stream of aftermarket.

Jack Egan (Equity Research Analyst)

Yep. Great. Yeah, that's super helpful. And then on silicon carbide, we're seeing more weakness crop up there, specifically in the financial results of the device manufacturers.

But you mentioned that silicon carbide is still generally pretty resilient and that I think you said it'd only be down a bit in 2025. But with that lower EV adoption and industrial demand kind of starting to flow through the financials for that industry, could that be kind of a risk to your guidance for the second half to be slightly better than the first half? Or is that largely contemplated in your guidance as it is today?

Jamie Coogan (EVP and CFO)

Yeah, I think it's Jack, as we sit here and say, it's expectations relative today based on customer conversations, discussions, reviewing the current backlog, as well as some of the bookings activity we've seen through the first quarter so far year to date.

As Russell sort of said earlier, where we sit right now, we're a little bit of an encouraging sign, probably too early to call it a victory, but encouraging signs relative to bookings where at the same point in the fourth quarter, we're ahead of our bookings rate in 2025 where we were at the same time in the Q4 timeframe. As we think about silicon carbide broadly, right, I think what we are seeing is truly diversity in the way our customers are engaging with the product, right? So some of them align very well with their public commentary in terms of how they're thinking about making investments going over the course of 2025, whereas others, again, not that they're inconsistent, but they have a path in place to spend the CapEx dollars and build out the capacity, and we see those in our expectations for 2025.

What I think helps us is we have a broad range of customers in the silicon carbide space, and we're not overly tied to one in particular customer in order for us to meet our financial results and objectives in the course of the year. I think that is really what maybe differentiates the expectations of our performance relative to a single customer's expectations or public commentary.

Russell Low (President and CEO)

I think that's exactly right. Each customer is slightly different. You've got some that want to build ahead of demand, some that are optimized in their yield because they want to get into opportunities, some that are looking to gain market share, but it's difficult given the time it takes to get qualified. Some are taking the opportunity to transition to 200 mm, go to 400 V to 800 V architectures, which would be trench and superjunction.

So a lot of customers are doing different things. So I would say that it is very customer-specific.

Jack Egan (Equity Research Analyst)

Great. That's super helpful. Thanks, guys.

Operator (participant)

Thank you. Our next question comes from David Duley of Steelhead Securities. Your line is open.

David Duley (Managing Principal)

Yeah. Thanks for taking my questions. I guess first, just a clarification. Can you help us with what silicon carbide revenue was in Q4 and for calendar 2024, dollar or percentage? And then as a follow-up.

David Ryzhik (SVP of Investor Relations and Corporate Strategy)

Dave, this is David Ryzhik. Yeah, it should be in the slide presentation. I think we can pull that up.

Jamie Coogan (EVP and CFO)

36% for the quarter and 41% for the full year for silicon carbide. Yeah. Yep. 6% up year over year.

David Ryzhik (SVP of Investor Relations and Corporate Strategy)

Okay. And then. Is that helpful, Dave?

David Duley (Managing Principal)

Yes, very much so. Do you expect the silicon carbide business to be part of the second half improvements? Or, I was a little. This is a follow-on to Charles's question. Is what exact geographic regions or is silicon carbide expected to be up in the second half of the calendar year?

Jamie Coogan (EVP and CFO)

Yeah, so geography-wise, we talked about, right, again, we're very broad-based across the board in silicon carbide, Dave. And so, again, we are seeing digestion. We expect China to be a lower portion of our revenue base right across the entirety of the portfolio in 2025. And so there are strengths. There are other geographies that are having strengths in the product offer to offset some of that. But we haven't given specific details on geographical expectations or anything like that.

Russell Low (President and CEO)

But what we can tell you, Dave, is now that customers have worked out their plans for the year, we are working very, very closely with them. We are looking at our bookings. We are looking at the forecast. And so we do know pretty much project by project where we believe these tools are going to go to. And we've double-checked that the fab will be ready for the tools, etc. So this is, and I'd say that we've talked about the first half being a China digestion. We've talked about us being very broad-based on silicon carbide. And we've talked about outside of China actually remaining relatively resilient, I'd say.

Jamie Coogan (EVP and CFO)

Yeah. And just, again, broadly speaking, it's going to be down year on year, down year on year, but not as much as other parts of the mature book of business for us, Dave.

David Duley (Managing Principal)

Yeah. So general mature is going to be down more than, and other power stuff is going to be down more than SiC is your message. Silicon carbide.

Jamie Coogan (EVP and CFO)

Yes, that's correct.

David Duley (Managing Principal)

Did you? I'm sorry, you talked about not wanting to break things out on geographic regions. Do you expect silicon carbide to be up in the second half of the year, even though there's digestion in the first half?

Jamie Coogan (EVP and CFO)

Yeah. We're not going to get into that level of granularity just yet, Dave.

David Duley (Managing Principal)

Okay. Thanks.

Operator (participant)

Thank you. Our next question comes from Tom Diffely of D.A. Davidson & Company. Your line is open.

Tom Diffely (Managing Director and Director of Research)

Yes. Good morning. Thanks for the question. Russell, I was hoping you could give us a little bit of an update on your plans to expand more into Japan in that geographic region.

Russell Low (President and CEO)

Hey, Tom. Thanks for the question and good morning. So yes. So Japan represents about $450 million in any given year. Our percentage of that is, as you know, sub 5%. So it's a really good opportunity for us.

We've actually managed to make a bit of a beachhead there in power and actually in some other areas as well, but power has been the great opportunity. So obviously, the local vendors of our implantation have often had the lion's share of the business. They haven't innovated and moved forward as quickly. So as, sorry, as Japan got into power, silicon and silicon carbide, they were looking for the right equipment that you need to get into high volume, which is the whole portfolio. You can't just have a medium current machine. You need the high current and the high energy. And that's where we've been able to come in. And so now we're working with those customers, working on their roadmaps, and making sure that we continue to expand in those applications of power, but obviously, it's not just power. I think it's interesting.

We just shipped. I think in Q2, we mentioned we'd got a system shipping into Japan for advanced logic, so that's all shipped in Q4, and we're now in discussions for actually follow-on businesses there as well, so naturally we want to build our footprint through our strongest products, but as you know, once we get in with a product, we want to fan out the applications and bring the rest of our product portfolio through with it, so I think year over year, we're going to see a modest improvement in revenue in Japan, and this is what I consider to be the seeds. You want to get into these projects early in where it's a mini line, and then when that mini line starts to ramp up, you want to make sure you get the volume there.

So hopefully, that gives you a little bit more clarity there, Tom.

Tom Diffely (Managing Director and Director of Research)

Yeah. So is it safe to assume that both Japan and advanced logic are more of a 2026, 2027 story?

Russell Low (President and CEO)

I think so. When we built our $1.6 billion model back in the summer, we want to see secular growth in silicon carbide. We've talked about that. We want to see recovery and memory in general mature. When it came to advanced logic in Japan, we had modest numbers in the model for that in 2027. So you're going to see, again, a modest improvement year over year building towards a modest component in our $1.6 billion business. And remember, of that $1.6 billion, we're expecting, say, $400 million of that to be aftermarket.

So really, we're looking to have a very modest amount of equipment in advanced logic where we're sowing seeds in Japan, where we're sowing seeds as a component of that.

Tom Diffely (Managing Director and Director of Research)

Great. And then just a quick follow-up. Jamie, on the backlog side, have you seen an extension or expansion of the duration of your backlog? And has anything fallen off the backlog?

Jamie Coogan (EVP and CFO)

So again, we see customer push-outs occur, right? And so these will be for purchase orders in backlog, Tom. And that was some of the activity that we saw between our third-quarter call and fourth-quarter call here was it's a customer request to push out some delivery dates and expectations, not inconsistent with what our customers have said externally relative to their views on this. So yeah, to some extent, there's a bit of extension of the duration in the backlog as a result of that.

Tom Diffely (Managing Director and Director of Research)

Okay. But does the backlog represent the next 12 months, or is it total backlog going forward?

Jamie Coogan (EVP and CFO)

Total backlog.

Tom Diffely (Managing Director and Director of Research)

Yeah. Some of it goes to the first couple of quarters of 2026.

Jamie Coogan (EVP and CFO)

Yeah.

Tom Diffely (Managing Director and Director of Research)

Okay. All right. Thank you.

Jamie Coogan (EVP and CFO)

Thanks, Tom.

Russell Low (President and CEO)

Thanks.

Operator (participant)

Thank you. Our next question comes from Mark Miller of The Benchmark Company. Your line is open.

Mark Miller (Equity Research Analyst)

Thank you for your question. I just had a question. You're projecting the significantly lower sales in the first quarter. Yet OpEx is going to be somewhat higher. I'm just wondering if you can kind of break that down in terms of SG&A and R&D, what's going on there?

Jamie Coogan (EVP and CFO)

So, as it relates to, I think the general commentary is our view is to have to kind of in the first quarter inherent specifically. It's going to be relatively flat to what we saw with a slight uptick just given the seasonal component of some of the way the expenses roll through, and that impacts all the line items fairly equally across the board, and then throughout the rest of the year, we anticipate finding ways to continue to invest in our R&D and the business, and we'd expect as a percentage of sales, that number to be slightly higher than what we've had historically for that business. Sort of compensating that, the difference would go through the SG&A line item. Yeah.

Russell Low (President and CEO)

Mark, I think it's clear that we believe investing in our products and services, working close to our customers, is one of the best returns we can have. So we actually kind of see the downturn slightly as an opportunity in the sense that our customers have more bandwidth. We can work with them. We can develop new products, get those products qualified. So then when the upturn arrives, we have new products and services to offer. And that's where we take advantage. So yeah, we are definitely, as Jamie mentioned, focused on products and services, working with our customers.

Jamie Coogan (EVP and CFO)

Yeah. And through this transitory cyclical digestion period, right, we do expect kind of resumed growth in 2026. And so we want to make sure that we're positioned for that, Mark.

Mark Miller (Equity Research Analyst)

Yeah. Okay. So basically, for the first quarter, both SG&A and R&D is flat to slightly up. Is that correct?

Jamie Coogan (EVP and CFO)

Yep.

Mark Miller (Equity Research Analyst)

Okay. Thank you.

Operator (participant)

Thank you. Our next question comes from Christian Schwab of Craig-Hallum Capital. Your line is open.

Christian Schwab (Senior Research Analyst)

Great. Thanks, guys. Just in the mix of business, if we look at it by end market, call it auto and industrial, is it as simple as taking silicon carbide and IGBT versus general mature to get the mix of business between end market shipments to, say, auto and industrial?

Russell Low (President and CEO)

So again, I don't fully follow the question, but I would say, Christian, that basically our general mature nodes are very driven by consumer, industrial, and automotive. And the biggest use of silicon carbide is, number one, the biggest market is automotive. But the second market that's growing actually really quite quickly is the industrial part.

Jamie Coogan (EVP and CFO)

Yeah. Similar dynamic.

Russell Low (President and CEO)

Yeah. I mean.

Christian Schwab (Senior Research Analyst)

Yeah. So that was my question. I'm just trying to get end market exposure. So in 2024, what percentage of your revenue do you believe went to automotive and what percentage went to industrial/consumer applications?

Russell Low (President and CEO)

Yeah. As best we know how we track it. So we don't know. It's very hard for us to know what our customers ship their products into.

Jamie Coogan (EVP and CFO)

Correct. We don't get discrete data like that from the customers on where they're ultimately putting. So we're using the same sort of general data on the utilization of silicon IGBT and silicon carbide in making those general assumptions. We don't have specific end product applications necessarily.

Christian Schwab (Senior Research Analyst)

Great. And then follow-up to that, on your automotive exposure, what percentage of that do you think services the domestic Chinese market versus global players?

Jamie Coogan (EVP and CFO)

So okay. On silicon carbide products we sell, right, I would say our European customers are largely the ones putting those silicon carbide devices into the automotive space today. Yeah.

Russell Low (President and CEO)

I think it's fair to say that the domestic Chinese manufacturers, even the BYDs as well, we don't believe they're qualified yet to put their devices because of the reliability requirements into their own electric cars. I can think of maybe one company in China that may be ahead of the curve. But ultimately, pretty much if you're going to put silicon carbide into an electric vehicle, it's going to come from the North Americans, Europeans, and the Japanese.

Christian Schwab (Senior Research Analyst)

Perfect. Thank you for that clarity.

Russell Low (President and CEO)

And then bear in mind, that's the question. Sorry, guys. Bear in mind, that's for the all-important MOSFET in the drive system. Remember, there's an awful lot of diodes that go into these cars as well. And I think the Chinese do manufacture a large part of those components, the silicon carbide diodes.

Christian Schwab (Senior Research Analyst)

Correct. Yep. Got it. Then we talked about an investment year for customer ramps. I guess I'm kind of confused on what technology investments you're investing for. Is this for new upgrades to machines, and if you could provide greater clarity of what that exactly means, is there transitions in what you're providing as far as a box structure for different applications, high current, high energy, etc., for 400 V to 800 V changing transition? I'm trying to understand the investments that you're making to drive future growth. What's changing that you're sustaining the investment?

Jamie Coogan (EVP and CFO)

Yeah, so part of it, it's a little bit all the above, right, at the end of the day, so part of it is incremental upgrade opportunities that then flow into new product development and technology.

So as we think about our power customers very broadly, our team is working through solving some of their critical challenges to improve throughput, efficiency, reliability of the devices at the end of the day. And so we're making investments to ensure that, one, we can go back and upgrade the suite of tools that we have available for those and continue to transition our products to sort of the next generation of the technology. There's examples of investments we're making in memory tools and technologies to meet our customer roadmaps in that space.

Russell Low (President and CEO)

Yeah. I think it's fair to say, even though we might call these mature technologies, the equipment that goes into them are anything but. So if you think about what we've been going through with silicon carbide, we started off with a medium current machine.

We knew that as people moved to trenches and superjunctions, they needed a high energy. In fact, the energy's actually been creeping up and up and up. That's an example of where we have to develop the right equipment. That's where we're putting our resources. The other thing is, I mean, we haven't talked much about it, but proton implantation for IGBTs, those energies have gone up significantly. That takes a lot of innovation to be able to support our customers with a highly productive tool that can achieve the energies. Our stated strategy of driving with the secular growth in silicon carbide, making sure we hit the recovery in memory and general mature, and that still takes work while actually expanding into advanced logic and Japan, all of that strategy is very well aligned with our spending on our products and services.

Jamie Coogan (EVP and CFO)

Yep. And those are multi-year, right? Those are multi-year projects that we're committed to.

Christian Schwab (Senior Research Analyst)

Great. And then my last question, your implied guidance for revenue in 2025, call it $800 million plus or minus, is substantially below your 2027 goal of $1.6 billion. I'm having a tough time reconciling what would have to happen in end markets for that to still be an attainable objective.

Russell Low (President and CEO)

Right. So that was the model we provided back in July. And we showed a very kind of clear path from where we were to where we need to get to. Yeah. It's fair to say that things have changed. But that was a very thoughtful model. And what I would say about that model is, one, we're not looking to update it every quarter. But the strategy remains exactly the same. The strategy we've outlined doesn't change.

So we actually feel fairly comfortable with the $1.6 if the timing that's a little bit uncertain. So obviously, trying to predict the timing in the middle of a downturn when people are all kind of doom and gloom wouldn't be productive. So I think we are focused on the long-term secular growth in power, market recovery in memory and general mature, share gain in advanced logic and geographic expansion in Japan. And we believe, like I say, that'll get us to the $1.6. The timing is unclear right now.

Christian Schwab (Senior Research Analyst)

Great. No. No other questions. Thank you, guys.

Operator (participant)

Thank you. Thanks. This concludes our question-and-answer session. I would like to turn it back to David Ryzhik at this time for closing remarks.

David Ryzhik (SVP of Investor Relations and Corporate Strategy)

Thank you, Dee Dee. I want to thank everyone for joining our call and your interest in Axcelis. Dee Dee, you can now close the call.

Operator (participant)

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