An organic investment opportunity set up, you know, and we've been doing some shareholder return over the period of time to capitalize on where we are with the market. I'll go into more detail on what an implanter is in a second, but we were founded in 1978 by really the, I'll call it the fathers of implantation. You know, we've been around for a long time, over about 1,500 or so employees, 2024 revenue of about $1 billion in total. That makes up both sales of systems, so we sell systems, as well as aftermarket, which is spares, consumables, and upgrades for the systems that we have in our install base. A robust set of patents today that help protect a lot of our technologies and our inventions, and really, really strong financial performance. For those that are new to the story, what is Ion implantation?
Again, we're a fundamental process step. You think about deposition, you think about lithography, that's sort of the ASML products, you think about EP, you think about measuring and inspecting to make sure your device is working. Implant helps to change the physical properties of the silicon or silicon carbide or GaN wafers to allow your device to operate in the means that's necessary for it to perform its task. We are one of the very critical steps in that process space. There are two players in the space today that make up 80% some odd % of the overall market. It's us and Applied Materials. Applied Materials is number one in the market overall. We're number two in the market overall. We have some relative strengths given our product technology and suite that I'll talk about in a second here.
What is an ion implanter? You can see the picture here. We tried to show that for scale purposes, but it is roughly the size of a hotel room. It is a very large system and tool, takes up a little bit of the fab footprint here. Ultimately, again, it is designed to modify the electrical and physical characteristics of the material in order to meet the device performance requirements. That can come in multiple different flavors. It can be what we call a medium current tool, which is a little bit of the seven iron. That represents around 25% or so of the market. It can do a lot of things. It does not do everything well, but it can do a lot of things for you in the fab.
Ultimately, as you continue to specify your device requirements, you're going to need either a high current tool or a high energy tool. A high current tool is for more dopant, so you want more coverage of ions on your substrate, so you will use a high current tool to achieve that. Whereas a high energy tool will put the dopants deeper into the substrate. From a market size, high current is around 50% of the overall market, medium current is around 25% of the total market, and high energy is around 25% of the market. Ultimately, what we do is we shoot ions or your source materials down an electrical beam line into the wafer. The tool itself is designed inherently of multiple complex sub-assemblies and subsystems, all integrated using software to ensure that you both get precision and productivity.
Implanters are designed to be not necessarily the most precise tool in the fab and not necessarily the most productive tool, but it is the most precise and productive tool in the fab today. Carved out a niche opportunity set inside of fabs for the manufacturing of devices. Every chip requires this, and it is one of the most complex steps, you know, in the fab today. Behind lithography is probably one of the most complex steps. As you can see here, you know, it still makes me, you know, more than 50 quadrillion ions per square centimeter of wafer can be productively inserted. I mean, that is just a phenomenal amount of ions that are going in there. We can basically put any element you would like.
Not that you'd put all the elements into the wafer, but we can certainly put any element you would like into a wafer to get the relative properties that you're looking for in your device design. We jokingly say that this isn't rocket science, but it is nuclear physics. And so, you know, at the end of the day, this is a particle accelerator at scale, is what we are doing here to ensure that the semiconductor devices work as they're supposed to. So our ions are everywhere. So a lot of these are the larger trends which have grown our market. There's a slide, a couple slides up, we're going to talk about the market size itself. It started at about $1 billion and has grown today to be somewhere in the $2.5 billion-$3 billion range.
The primary drivers for this is the proliferation of what we call general mature or foundational technology devices in essentially everything that we do. This is the introduction of image sensors more broadly into automotive, outside of cellular devices. This is communication RF technology devices. This is the introduction of power semis into the market very broadly. We think about compound semis like silicon carbide. Those are driving the market expansion and the addressable market for us over this period of time. One of the things I think that gets lost is that, yes, we have a relative strength today in silicon carbide power, but we are a broad-based semiconductor capital equipment company. We serve the memory markets. We have capability to serve the advanced logic markets. Obviously, power device has been a really strong suit for us.
Ultimately, we continue to have a strength in general mature. General mature power are the most implant intensive devices today because they're typically made on the 28 nanometer and above nodes, in which, so for those that are familiar with semiconductor manufacturing, you know, the implant intensity sort of peaks at those 28 nanometer nodes and above. We see a particular strength there. On the advanced logic side, you know, there's an opportunity set for us, and we'll talk about what that opportunity is for advanced logic to expand our presence into that market and some of the actions that we're taking to do that. I'll talk about that in a couple slides.
I think fittingly for the conference that we're at and some of the work that we've done with Jed, you know, we look at energy consumption as a real one of the more significant secular trends here. We look at what we're able to do and what we provide to the market as a significant opportunity to help, you know, alleviate some of the pressure, right, that energy utilization has. When we think about AI, one of the things we think about is the energy use of AI, AI servers, and AI devices. They're absolutely power hungry. As you can see here, over time, both the growth in traditional data center, AI data center is significant over this period of time. Associated with that is the cooling of those devices.
One of the benefits we'll talk about in a minute is what silicon carbide can do to help alleviate both the energy use and consumption as well as reduce the cooling needs and requirements over time as adoption of the technology grows. In addition to that, we see automotive electrification also driving, you know, a long-term trend inside of both power device and more broadly the technologies that we support today. This is really across the portfolio. This is not just full EV adoption that we require. It's also hybrid, plug-in hybrid, and all points in between. We'll talk about some of the trends that we're seeing there that are helping to push silicon carbide beyond just full electric vehicle into the hybrid and other parts of the chain. Where do we sit today?
Today, as we look at the market and break it down by segment, we are the number one implant technology provider in the power segment. We're number one in high energy, and we're overall number two globally. On a growth perspective, we have a number of vectors of growth and opportunities for growth for us. We're going to continue to grow in the markets in which we serve today through technology, implementation, and integration, working closely with our customers on their roadmaps. We have some geographic expansion opportunities. We're underrepresented both in Japan as well as in advanced logic, and that's primarily the Taiwanese market. We have opportunities there to continue to grow our footprint and our exposure into those markets over time. We have the Purion product family. That is our single wafer technology today. The Purion product is a growing portion of our install base.
We have a large portion of historical tools that are in our install base, but Purion is our latest technology, and that continues to become a larger portion of our overall TAM for us. That is going to open up aftermarket opportunities, which has been a really durable portion of our business for us over time. Ultimately, we're working to identify some new implant applications. These are identifying the introduction of implant into process steps where historically you would not have used implant. Typically, implant is used in the front-end production of semiconductors. We're now introducing middle end of line and back end of line applications for implant technology to do some material modification steps. We're working to identify, you know, both customer opportunities and incremental steps there that we can help to support the growth. Where were we?
In 2024, and this is our systems revenue. Total revenue for 2024 was about $1 billion. $250 million or so of that was our aftermarket business. The remaining portion of that was related to our system sales. This is sales of systems across the board. As you can see, silicon carbide represented around 41% of our total system sales in that period, whereas general mature, so this is non-power devices, was 41% as well. We have a really strong franchise today in the general mature process steps. That is general mature, silicon IGBT, and silicon carbide. Memory representing 2% here, that is at a really historic low. We have a very strong set of customers in memory, but memory spending has declined annually since 2022 as that portion of the market has been in a cyclical downturn.
For reference, going back to 2022, total revenue in memory was around $120 million of sales. It diminished to $90 million in 2023 and was roughly around $20 million or so here in the 2024 timeframe. That decline has more to do with customer spend cycles given the downturn in that market and not some shift in market or market share or technology trend that moved away from us. We do expect that to recover as the memory market recovers over time. Advanced logic is the opportunity set for us. We typically sell maybe one, two systems a year here. There is a market size there we will get to in a slide or two that goes through what that opportunity is.
The TAM itself has grown significantly over the last decade, and that has, like I said, has more to do with the rapid growth and the utilization of general mature technologies, RF communication devices, image sensors, and power devices in the everyday technologies that we use. The market for a long time hovered around $1 billion. It has more than tripled since then as you have seen the adoption of these foundational or mature technologies grow over time. When we break power down very specifically, what you can see here is the trend from 2023 to 2029, both in silicon IGBT, silicon carbide, and GaN. As we look, the fastest growing part of this market really is in silicon carbide.
That has more to do with the growth in the automotive and industrial parts of the business as those silicon carbide finds price parity points that allow it to creep into what were traditionally silicon IGBT use cases. As manufacturers become more efficient at the production of silicon carbide devices, as the input cost associated with those devices come down, we do expect the broader use of silicon carbide, not just in full EV, but ultimately into hybrid EV, industrial applications such as HVAC, into power management systems and the facilitation of data centers, but ultimately potentially even into the data center racks themselves over time. I still think there's a really long, long road here to run for silicon carbide. Here's how we got into the market. We started this journey, you know, 10 years ago at least. Ultimately, what did we do?
We recognized there was a need in the marketplace. We partnered with our customers very closely. We launched the first product set in here. We developed that product set to include our medium current implanter, again, the seven iron of implanters, because this allowed our customers to play, right, within the silicon carbide substrates, determine what the right characteristics were that they needed. We then worked with them to develop a high energy implanter, ultimately growing that to a high current implanter. What we saw is our sales from this from 2020 to 2024 grow from $10 million to over $300 million. This is going to be a durable portion of our portfolio going forward. We believe that there's incremental growth opportunities as we look ahead to continue to grow our position within the power segment, specifically within silicon carbide.
What this shows, though, is that we do partner with our customers and we take a long-term perspective in view in terms of how we attack our market opportunities. We are developing, you know, next generation implant technology today to support market opportunities in the future. I'll touch on those in a second. One of those is in the Japanese market. We are underpenetrated today. We're about less than 10% within the Japanese market, which is approximately $450 million. This market actually serves power, image, memory, and advanced logic. We're using our strength in the silicon carbide power market today to get a foothold in the Japanese market. We're able to find success today with Japanese power device manufacturers who need a silicon carbide solution. We're working with them today. We're using that to now fan out our technologies into other parts of this market.
You know, we're making some really good progress here. The next is in the advanced logic phase. Maybe I'll pause here for a second. We are underrepresented here. It's less than 5% current market penetration. I would say that's generous. We typically sell one to two tools a year into advanced logic. It's not because our tools are not capable enough, but we're not qualified as the process tool of record across the significant portion of the steps as they exist today. Our customers there are not as price sensitive to the sort of upfront cost of a tool or a system. We really need to solve a difficult challenge for them in order for them to uptake our tool or system into their fab. What are we doing now today?
What we've got are learner systems, we call them. These are systems that are at R&D facilities that are working on the next generation device technology. Think of these as node changes as we continue to move down nodes. We have our systems in with customer R&D facilities as they're working on high value problems. Ultimately, what we want is to, through these learning tools, identify new process steps where there isn't a tool of record today. We want to position ourselves to be qualified on the node changes so as they occur, we can then compete aggressively for position with a qualified tool on those process steps at the node change. Ultimately, what does that mean? It will lead to more high volume production.
The advanced logic market is typically served by the high current system. We are investing today in our high current systems and technologies to allow us to achieve these long-term goals. This is probably a three- to five-year play for us. Today we're maybe, like, you know, 12-18 months into this process as we look today. We would not expect any meaningful contribution here in 2025, not likely going to be a meaningful contributor in 2026. As we look out, we do believe the steps we're taking today will position us to be able to see incremental growth opportunities as we gain some market share in advanced logic.
The ancillary benefit to all of this, though, is that, you know, the technology we develop to support the advanced logic market ultimately will bleed their way back into the general mature memory and other markets that we serve today. We will be able to pull that tool through to those other markets and drive incremental growth opportunities. All in all, as we think about artificial intelligence and we think about electrification, right, these are two sort of big meaty items here that are going to drive near-term growth for us. On AI, we're not going to sort of directly benefit, right, from the growth in AI. The second and third order impact from artificial intelligence really impacts most of the segments where we have a particular strength.
We think about, you know, the amount of data that's being created and the fact that it needs to be stored somewhere. That's going to benefit DRAM and NAND production, right, in the long run. We think about im age sensors, RF analog, right? AI is hungry for data. These are the types of devices that feed that data into the large language models that allow those large language models to output what they do. Ultimately, we talked about power consumption. Both silicon carbide as well as silicon IGBT are going to be, you know, primary drivers for managing the power through there. The preference for us is for that to be a silicon carbide device.
We think that the thermal properties, power management, and efficiency that provides when you compound it across the racks, the facility, the cooling infrastructure, we'll see significant energy reductions relative to traditional means of providing power management, you know, through these devices. On the electrification front, the continued electrification of our world, whether that be through, you know, full hybrid vehicles or full electric vehicles, hybrid vehicles, and everything in between, ultimately are going to require more silicon carbide as well as silicon IGBT devices. This moves to solar, right? Infrastructure. This moves to wind infrastructure. This moves even towards enhancements to the overall power grid. If we think about better ways of transmitting power from, you know, source to use, you know, using semiconductors to work through those transitionary periods will provide significant benefits.
Overall, you know, we continue to be really well positioned to take advantage of these market opportunities over time. Moving to the obligatory financial slides, here's how we've performed. You know, since 2020 to 2024, you've seen a significant increase in our overall performance that's driven by growth both in our power segment as well as our general mature segment as we continue to see investments as we grow capacity throughout the course of that. I think what's more, you know, again, leverage is a very important thing. As we look at the performance of the business overall, we're able to, you know, grow margins. If you go back before this, we were sort of sub 40% margins. As we grew the business, we took advantage of the position that we were in, the facilities that we have on hand.
We were able to utilize those more efficiently and effectively, drive incremental leverage opportunity. In addition to that, we segmented the market and we priced our products from a value perspective. We continue to get incremental ASP on our products while driving incremental volume, you know, through the facility to get better leverage on gross margin. As we look to Q1 in 2025, just as a point of reference, you know, we ended the fourth quarter of 2024 with about a 46% gross margin. We had signaled as part of that fourth quarter earnings call that we anticipated volume to be lower in 2025 as there was some digestion of capacity, you know, as we came into this year. In the first quarter, that volume reduction occurred. We saw systems volume come down around 25% relative to where we were.
However, our gross margins within that period stayed relatively flat with what we saw in the fourth quarter. What we've been doing over time is managing costs relative to volume while at the same time positioning the business for incremental growth on a go-forward basis. We continue to have a disciplined cost structure in light of all of the opportunities we have in front of us. From an adjusted EBITDA, we did just recently introduce non-GAAP measures into our external reporting processes. You know, as you can see here, we continue to perform really well both on an adjusted EBITDA and a gross margin perspective. Our diluted earnings per share, you know, did peak there at around, you know, $8 a share, but it's going to expect to come down as volumes decrease and we continue to invest in the business.
Overall, you know, we are focused on maintaining, you know, really tight cost control. I kind of analogize, we're driving with a foot on the gas and the brake at the same time. We are putting more dollars into R&D despite the lower sales volume year-over-year. We're able to do that given the strength of our financial position. However, on the G&A front, and as we think about how we source and manufacture product, we continue to find opportunities to eke out incremental gross margin while maintaining G&A cost control to feed back into that R&D loop and cycle. This really will position us to grow as the markets that we serve today enter their recovery phase. CS&I, this is a really important part of our business. We talk a lot about system, system, systems.
You know, our install base today is approximately 3,300 tools. That is made up of our Purion tools. You can see the sort of dark blue line here is our Purion. That's our single wafer product. The light blue line is our multi-wafer. This is a legacy tool set. This is what implanters were before they moved to single wafer. That install base has really provided a really robust and meaningful and highly profitable floor for the business. As a point of reference, in Q1 of 2025, revenues from our CS&I business were relatively flat with the CS&I revenue we had in Q1 of 2024. It is a durable portion of the business, a highly profitable portion of the business.
Actually, David was sharing a little anecdote today that what we saw, we generated more profit in Q1 of 2025 than we did in the entire year of 2019. We have really shifted the nature of this business through the growth in our aftermarket business. There continues to be opportunity for us to expand our presence with our customers as we introduce service contracts into our portfolio, which will increase share of wallet with each of these customers and drive more meaningful customer engagement with the install base that we have today. That is in the early phases of adoption. We expect that to continue to grow over time, a total portfolio of our aftermarket work. R&D will continue to be an important part of us. We continue to put more and more dollars into R&D.
As you can see here, we expect, you know, over $100 million of R&D to likely go in in 2025 as well as we continue to invest in those parts of the technology that, one, allow us to expand the moat that we have. Increase our patent portfolio, increase the competitive advantage that we have in the space today, continue to pursue opportunities in advanced logic, which provide, you know, a potentially $400 million, better access to about a $400 million market today that we do not serve appropriately, and ultimately identify new implant technologies that could open up middle end of line, back end of line, new applications for implant technology for our customers. This slide, I think, I just like the way this bar looks. We did see a nine times increase in our cash.
If you look from 2015 to 2019, we generated $71 million of cash over that period of time. Between 2020 and 2024, we've generated $673 million of cash. We've really positioned this business for financial health, financial longevity, and durability. This allows us to invest in the down cycle very meaningfully. This gives us the comfort. We have no debt today on the balance sheet, just to be clear. We're debt free. We have cash today that we are using to invest in the business. We continue to generate cash through, you know, I'll call it sound working capital management. We had an 89% conversion ratio for EBITDA to free cash flow conversion ratio in Q1. You know, even during the downturn, we generated a meaningful amount of cash while continuing to invest in the business very, very strongly.
We expect these trends to continue, which is why we announced in March an incremental $100 million allocation to our share repurchase program. We had an existing $200 million authorization. We put another $100 million into that. In addition, we announced an upsize to our quarterly spend. We do anticipate spending on average, we spent about $15 million a quarter. We are going to, you know, upsize that. We did not give any specific number, but we do anticipate spending more than $15 million a quarter now. We anticipate to see, you know, continued strong cash flow generation for the business. With that, to wrap up, and I know we got about three minutes left here. If there is a question or two, we can certainly take them.
You know, at the end of the day, we think that Axcelis is well positioned for the long-term secular trends. We've got a very strong balance sheet that allows us to invest in the core business while taking advantage of those trends. We're well positioned within the market to take the inflections and secular growth that is coming our way. You know, with that, any questions, Jed?
I'll kick it off. We're going to be up in Jenny B for the breakout, but maybe just two parts. One, the cash flow is, I would agree, it's a phenomenal looking chart. First is, what would you attribute the majority of that 9X increase? The second question, first time I've heard you speak about gallium nitride with respect to implant. Usually that's deposition. That's a new market opportunity.
If you could just expand on what that is for that, because that seems, I certainly didn't have it in my TAM analysis there.
Yeah. As we think about, on the cash flow side, you know, I think we attribute that to obviously the improved profitability. You know, I think that as the prior sort of management team before myself and Russell came on board, did a phenomenal job of actually getting cost under control, driving gross margin and improved operating efficiency. I think you're seeing the benefit of those actions and the continued, I'll call it, focus on working capital management, right? We continue to manage working capital very effectively and efficiently in this business, which has allowed us to generate that cash. Those practices are continuing, you know, with Mary and Kevin, you know, retiring. Gallium nitride, though, is part of the TAM.
It is not as implant intensive as silicon carbide. You still need to implant gallium nitride. You just do not implant it at the same intensity as silicon carbide. As we think about what the opportunities are, clearly I think we have a preference towards silicon carbide, silicon IGBT being the primary, right? They are more implant intensive. As we move towards compound semis, gallium nitride does provide, you know, sort of a modest opportunity set to grow if, you know, use applications move to gallium nitride versus silicon carbide.
Another follow-up, I guess. You have a virtual monopoly in high energy for 90% plus market share for that particular segment.
Do you see any technologies that you can use your linear accelerator technology and either a medium current derivative or high current that, you know, positions you, you know, that might be on the horizon that's not here? Just curious, or do you see another opportunity for high energy like silicon carbide that, but a different market opportunity?
Yeah. I think part of what we see with high energy is the opportunity for high energy really comes around, and this really is in the silicon carbide space. As we move from planar to trench to super junction, you will need higher and higher energies. We have a tool called XE Max today, which is sort of like phenomenally high energies. It was a tool initially developed for a different application in the general mature process space.
However, we think it may have found its footing in the manufacturing of, you know, super junction devices in the silicon carbide space given what needs to occur. That again goes back to this ethos we have of develop interesting technology, novel solutions, and the applications sometimes find their way to those solutions. We think about XE Max as a potential opportunity set for us. It'll never probably be as big of a market as those other opportunities are in medium current or high current. I think on the high current side, one thing that we need to do there is invest, just continue to invest in our tool set because it ends up being the biggest part of the market overall.
As the advanced technologies move to the mature technologies over time, we need to have high current systems that allow us to continue to attack that market in the way that we have historically. Great. Yep.
Thank you. We'll let Jenny B on stage.