Good morning, everyone. Welcome to the New York Stock Exchange. For all the folks that came here today, as well as the folks that are on the webcast out there, we welcome everyone to the Onto Innovation Analyst Event for 2023. This morning I'd like to go over just a few introductions, housekeeping items, and then we'll get started on the program. First of all, safe harbor statements. Any statements that we make today are as of the information that we have as of today. For future market conditions and our own business conditions, we take no responsibility for updating as things change. Secondly, we will be presenting on a non-GAAP basis today. The GAAP to non-GAAP reconciliations are in the appendix in the back of the presentation. The presentation is online at investors.ontoinnovation.com.
For your reference, you'll be able to see those. In addition to the appendix with the GAAP reconciliations, you'll also see a glossary. We've put together a glossary of technical terms, if you see things on the presentation, like Forksheet and CFET, et cetera, and you don't know what they are, there is a glossary that can update you on those things. Today's agenda, we're gonna start with Mike Plisinski, our CEO. He's gonna talk about our growth strategy, lead things off, followed up by advanced nodes, how be the opportunities, the market opportunities, our business opportunities, why we win in those advances. Followed up with Mike Rosa, our Chief Marketing Officer, who then follow up with our specialty and packaging, advanced packaging devices.
Again, showing the opportunities that we have there, and why we win. Finally, Mark Slicer, our CFO, will talk about our operating model, the we're doing, some of the initiatives that the company has already started in order for us to get into that long-term operating model for now, as soon as we possibly can and into the future. Finally, we'll do the closing again with Mike Plisinski. We'll close with some remarks on the future, and then we'll do Q&A. I would ask you to please hold questions until the Q&A. We'll bring the entire panel back at the end of the presentation for Q&A, and we'll pass around a microphone so that you can be recorded and ask any questions that you have.
Following the presentations today, we'll be serving lunch outside in the vestibule. That'll be from 12:00 P.M. to 1:00 P.M. You're welcome to stay for lunch, or if you have to get to another meeting, we understand that, but certainly lunch is provided. In addition to lunch, we have our technology row out there in the vestibule, so feel free to stop by and see what's under the hood for our metrology, inspection, software, et cetera, and get a little advanced look at what we have that we think is certainly unique and if I do say so, better than the competition. With that, let's start the show.
On social media, to fighting diseases and combating climate change, people around the globe are collaborating on an unprecedented scale. We all know what's at the core of it all, silently enabling this connectivity, small but powerful semiconductor chips. High-performance computing, AI, and 5G, made possible by EUV tech breakthroughs, have unlocked incredible possibilities for connection, learning, and collaboration. With those new applications comes complexity, making it more difficult to manufacture new products with the risk of slowing the pace of innovation. Onto is at the heart of this groundbreaking work. As the only ones in this industry embracing the power of connected thinking, we are revolutionizing the manufacturing environment. Like the visionaries who break down conventional barriers to discover new worlds, we're pioneering a new, connected approach to chip manufacturing.
From close collaborative relationships with our customers and strategic partners, to the AI-driven technology and systems that increase yields and lower defect rates, we enable connections across fabs and tools, delivering value across the semiconductor supply chain. As the world continues to evolve and become more connected, Onto continues to innovate beyond industry expectations. The story of connected thinking is the story of winning. That is how we measure our value and continued impact in the semiconductor industry. Onto Innovation, achieving growth outperformance through the power of connected thinking.
I'd like to thank you... I'd like to thank all the employees who generously donated their time to be on camera and not get paid for it. now I'd like to welcome Mike Plisinski, our CEO, to the stage. Mike?
Good morning, everybody, and, again, a warm welcome to you all who joined us here in the New York Stock Exchange, as well as those of you who are joining through the webcast. As many of you know, actually, for today, we want to cover, provide a detailed, a little bit more depth into the markets we're serving, as well as the technology we've applied to those markets, and then what that means for future growth. As you know, we ended 2022 with record revenue and record earnings. We outgrew the semiconductor industry by about 500 basis points, 25% CAGR over the three-year period, and then we had earnings improvements of about factor of five over that same period.
During that time, I want to call your attention to the revenue of the Nanometrics and Rudolph Technologies businesses before the merger. In 2019, each company was at its peak, near its peak, at about $280 million-$300 million. Companies come together, we double revenue. A few years later, we double revenue again. That growth, the global pandemic, supply chain impacts, challenged this new team in very unprecedented ways, and they were able to deliver these results. Through the next several hours, I guess, we're going to talk about how this team is going to deliver additional results on the same platform or on the same line.
At the core of our success or at the core of our growth, is the differentiated value proposition we offer our customers, and the foundation of that starts with the sensor technology that we have. The breadth of the technology gives us the ability to measure, to inspect, and even print some of the most advanced semiconductor structures in manufacturing today. In addition, several of these technologies, when I talk about unique, several of these technologies have no peers. Others have only one peer or one competitor. That gives us the opportunity to grow, not in just niches that are served by those technologies, but also expand the opportunities of those technologies in the semiconductor manufacturing space.
You've seen that with our ability to grow our SAM, which we'll talk about a little later, but which has grown from about $1.5 billion to $2 billion over the last couple years, sorry, to $3 billion over the last couple of years. We take this further, and we differentiate even further by leveraging our Fab-Wide software, which you can see out in the technology row, by bringing in third-party information, third-party data, to augment the data coming out of our systems. Instead of just being a very fancy, nice ruler for our customers, we're actually providing the information or the decisions so that they can positively impact their process, either ramp faster, improve yields, or improve device quality, device performance.
We do this by integrating, again, third-party data from tests, from process equipment, from other third-party process control, yet leveraging some advanced analytics to then augment our systems and provide what we call connected solutions. That's what you saw in the video. This is not just about our technology, it's also about a culture change and a culture shift with the team focused on our customers' success, and focused on looking beyond just providing unique technology, but providing unique solutions that directly impact the manufacturing process. Is it working? It's a great story, but is it working? We believe so. In the last three years, you can see here that we were recognized by our customers 6x .
If I look back prior years for both Rudolph and Nanometrics, that might have been about one or 2x in five years, so 6x in three years. You'll note in blue here, that the majority of these awards were for technology and technical innovations. What's also exciting is that several of these awards were not just for one particular solution or collaboration initiative, but in the case of some of the bigger customers here, they acknowledged our contributions for both front-end and packaging technology evolvement. What's interesting about that is not just that we were acknowledged for the, for the contributions we made, but how important packaging is becoming to many of these customers in their future roadmaps.
Of course, that's led us to achieve the number three spot as a supplier of process control equipment to the global semiconductor industry. We were number four prior to this. We moved up a spot, we think that as you learn more, that this growth is not plateauing, and we're only just at the beginning. That's history. Now, what about the future? You're all very familiar with the secular trends that are driving our industry, I won't go into a lot of details there. Despite the slowdown that we're all experiencing today, there is a lot of growth left in these secular trends. We're still in the early stages of many of these trends. Connectivity, for instance, 5G rollout is still fairly new. The next generation, 6G, isn't going to be rolled out for another 10 years or so.
At least most estimates have it starting next decade. About 14 billion connected devices today, going to 30 billion in 2030. A lot of growth there. Those connected devices generate a tremendous amount of data. Today, in 2022, ISS estimated about 100 ZB of data globally stored up into the cloud. They project that by 2030, that's gonna go to 600 ZB or a factor of six improvement. You collect all this data to monetize it. Monetization of data has historically been search, and it's been business intelligence, primarily advertising.
That's now shifting with the advances in computational and high-performance compute, that's shifting to AI applications, leveraging all that increase in data to provide more capability through algorithms and computation, which we expect the revenue to be from about $100 billion today, going up to $2 trillion in revenue by 2030. Of course, the net zero, the globe. It's a little bit different driver now, but going to combat climate change. It's almost like the electrification of everything, where you see lawnmowers and tractors, lawn tractors, the combustion engine is being replaced, not just in EVs, which everyone's well aware of, but what's gonna drive and enable that or hold it back is the ability to charge quickly and the availability of these chargers. That's again, where the semiconductor industry is coming in.
What does that mean for us, in particular? Secular drivers, lots of growth, lots of opportunity for over the next decade. For chip manufacturers, it's gonna drive the transition to Gate-All-Around, so the next generation of transistor architecture, it's gonna drive high bandwidth memory, and it's gonna drive advanced packaging for chiplets. All that from storage and from AI. Of course, net zero is driven by specialty devices or driving the specialty device market. You're gonna see these CAGRs and say, "Whoa! That's a huge CAGR. How is that gonna, you know, impact our business?" These are all normalized to unit volumes, so chips, the actual chips. How that translate to Onto Innovation, is gonna depend on sampling size, and it's also gonna depend on what happens to the other nodes.
you know, does 5 nm node continue to stay at the current level and even maybe grow, even as they transition to Gate-All-Around? A lot of this is variables that are unknown yet. It's clear that these transitions drive a lot of growth. Onto Innovation is playing in this growth with a suite of products, not just one or two products tied to one or two of these inflections, but a whole portfolio of products, further enhanced by the software that provide that differentiated value proposition that we think is gonna lead to outsized growth and continued outside performance. Here you can see that several of these products and portfolios are relatively new, introduced only in the last two years. Again, part of that collaboration, part of that change in our strategy.
To talk in more detail about why we believe those positions are going to lead to the outsized growth and why we're going to win, we have Dr. Srini Vedula, who will speak to you about the advanced nodes and why we're winning in advanced nodes. Srini is our Vice President, responsible for our Global Customer Success Group. That group has a mission. It's combined, brings together our sales, our service, and our applications to focus not just on getting an order and getting sign-off, but really on making sure that we deliver the value proposition to our customer, so that we are driven by the customer's success. Dr. Mike Rosa, he'll be talking to you after Srini, about the advanced packaging and specialty device markets and why we think we're well-positioned to grow there. Mike joined us from AMAT.
He's joined us about a year and a half ago. He's our Chief Marketing Officer, and he's responsible for our corporate strategy. Looking at the trends in the markets, looking at where the inflections are, the high-value problems, and seeing how Onto Innovation can best benefit from those trends. Prior to joining Onto Innovation, Mike was at Applied Materials, where he was responsible for marketing of the ICAPS programs, which started with specialty and legacy, then moved to include packaging. Both of these gentlemen are experts. I think they have great material to share, and look forward to catching up or summarizing after. Srini?
Thank you, Mike, and good morning, everyone. Today, I will be covering Onto Innovation's growth and outperformance, opportunities in advanced logic and memory devices. Mike already talked about the broader drivers, but these technologies are really going to be driven by strong inflections, multi-year inflections, as the industry migrates into the artificial intelligence era. One of the things Mike touched on is high-performance computing. As the compute requirements continue to get a lot more stringent, more and more processors going into high-performance computing will be demanding a lot of GPU and CPU content. As AI servers get installed more into both commercial as well as consumer applications, we expect that there will be increasing need for DRAM content.
Industry sources estimate that the number of transistors within the package could explode by a factor of 10 over the next decade. Likewise, the bit density increase in DRAM could increase at a CAGR of over 45% over the next several years. The long-term growth drivers for the core logic and memory technologies are intact. What our customers, the device makers, are working on is to ensure that their roadmaps align with optimizing power, performance, cost, and area. That's what drives our industry's roadmap. I'll start off with taking a look at logic, architecture, and its complexity. Today, the current generation of devices is based on FinFETs, all the way from 14 nm technology node to 7 nm and sub 5 nm, and these devices will continue in high volume production for several applications.
Where it gets interesting is that as the compute requirements get really stringent and a lot more intensive, FinFETs are going to be hitting a threshold limitation in terms of their overall capability. With that, our customers, the device makers, have been working on new R&D technologies to introduce the next generation architecture called Gate-All-Around, or GAA. This is a transistor architecture that is well defined in terms of its ability to scale, its ability to drive better electrical performance, and reduce overall reliability impacting failure modes. As these new devices come into the play, as said, it's scalable, so GAA will be followed by forksheet, CFETs. I think there's a multiyear roadmap already lined up.
These devices will require significant enhancements in manufacturing processes with the introduction of new materials, with the introduction of complex geometries, as well as increasing usage of extreme ultraviolet or EUV lithography steps. All of these have direct implications on process control challenges and to Onto Innovation. When you look at the metrology area, you'll see that a lot of these materials have to be manufactured in very thin films. We are talking 10s of nanometers in some cases. These require extremely tight accuracy and precision across large wafers within a very small process window, right? These have to be measured, they have to be the yield parameters have to be well understood. Additionally, critical dimensions, which we so-solve with our OCD technology.
In addition, control of the reliability and the electrical performance will require precise shape control for a lot of the geometries that are going to get manufactured. That's on the metrology side. Incoming wafer quality going into EUV is a big driver, as we cannot afford to have any sources of defectivity, not just in the printed area of the die, but also along the edge of the wafer and the backside of the wafer. The non-printed real estate on wafers are going to have a significant impact on overall performance and yield coming out of the lithography steps. That gives us an opportunity in terms of demonstrating and delivering value through our macro inspection products. I'll go through some of these examples in detail.
One other facet that I would like to point to you is that OCD metrology, historically, traditionally, have been based on targets, program targets, that our customers would use to make both thinner films as well as critical dimensions. As the devices have become more complex, the correlation between what you measure outside the printed area to what you measure directly to predict device performance, those correlations are breaking. As a result, over the last several years, we have seen our customers demand more and more within-die active area metrology, and this has implications for us. It could be, for example, moving in from scribe line targets... Oops, excuse me. Scribe line targets to SRAM structures, standard logic cells. What's the implication for us is that, A, these measurements are becoming extremely complex.
We now have to measure structures of interest by removing background nuisance effects, that's very, very hard to do. At the same time, the sampling area and the sampling density that requires active metrology is increasing. Customers are now demanding a lot more uniformity measurements within the die, which means the number of measurements that they make within the active area is going to be increasing, right? The takeaways are things are getting harder to measure. At the same time, more things need to get measured to really control these processes. With that, as this migration happens from FinFET to Gate-All-Around, we see with our broad suite of metrology and inspection solutions to have a significant impact in terms of our entitlement. We believe the opportunity for ON2 can increase by over 50%, normalized to wafer starts and normalized to equipment spend.
These will come in from three different areas. The first one is increasing OCD metrology sampling, expanding our applications into planar films and multilayer films, as well as increasing macro inspection with accelerated and increasing adoption of EUV process steps, as well as wafers going out for outgoing quality into advanced packaging. I'll walk you through how we play a role in each of these. This is a blow-up of a Gate-All-Around transistor. The inset there points to a very detailed gate structure with multiple films, which are all printed at 10s of nanometers, as I mentioned before. Our customers want to make dozens of parametric measurements to control these devices.
These can be thicknesses, shapes of the sidewalls, material properties, composition analysis of the starting gate material, and that increases the amount of process control intensity that's driving these measurements. For Onto Innovation, we introduced our flagship Atlas V OCD platform, which is precisely designed to address the metrology problems in these areas. At the heart of it, we rely on our technology supremacy. The amount of unique signatures that we derive with our eyes on the wafer, so to speak, allow us to make precise measurements at the atomic level, allow us to remove extraneous information, and provide information that's critical to our customers for process control, and at the same time, do this in a high-volume, fast production environment, right?
At the same time, as Mike pointed out, all of our technologies are driven through our AI-enabled modeling engines and solutions, which significantly reduce the cost to ramp and the time to ramp for our customers. We believe just in OCD metrology, as the industry transitions from FinFET to Gate-All-Around, there could be up to a 30% increase in sampling rates, which clearly benefits Onto Innovation. We are working very closely with all of our R&D partners in several of these applications for Gate-All-Around. As our optical technologies have evolved, we now see an opportunity to expand our application space beyond traditional OCD methods to planar films. This market is equally very large, as several of you know.
Why this is important for Onto Innovation is really how our customers are driving the efficiencies and the utilization of their fabs. For our customers, it's extremely important to take the capital investment that they're making in metrology equipment and really find efficiencies and utilization across the lifetime of the fab. Where in early R&D cycles you may do a lot of process tuning, and you may require certain type of metrology, like planar films metrology, as the yields improve and you go to high volume production yields, you would still want to do parametric control with OCD metrology. At Onto Innovation, our philosophy has been to come up with a common platform framework, both on the system side as well as on software and analytics. That ties together our OCD solutions, our integrated metrology solutions, and our planar film solutions.
With this approach, as an example, our planar film solution has the same foundational elements as our OCD platform. This allows our customers to bring in our solution for planar films measurements, and as their sampling requirements change, they can migrate some of those applications into OCD, right? With this approach of having a unified common platform approach, the Our case studies have shown that the upfront capital outlay that our customers have to build in could be as low as 30% relative to picking and selecting technologies with different, dissimilar technologies. We have seen the benefits of this. As we announced last quarter, we have seen a top three manufacturer who clearly identified with this value proposition, and we were able to gain a footprint with our thin films metrology solution. Next up, moving on to all-surface inspection.
I mentioned briefly that as EUV layers are getting adopted, there is increasing need to carefully control the defectivity signatures in the hundreds of nanometers along the edge of the wafer and the backside of the wafer. Okay? ASML estimates that the number of EUV passes that they will have to go through between logic and memory devices could double over the next several years. As the lithography depth of field tolerances become very tight, it's extremely important to identify the hot spots, find these particulates before they end up with yield killers during lithography. These are not only important at the incoming silicon stage, the wafers, the prime wafers that come into production manufacturing, but also during wafering process itself. That item, that brings up two separate opportunities for Onto Innovation. We have two solutions.
We work with the silicon manufacturers, where we have bare wafer, silicon edge, and wafer backside measurements. We have some unique differentiation with our capability to measure submicron defectivity, not just along the bevel of the wafer, but also the recessed region of the wafer, which gives our customers 100% coverage of defect sources that could impact yield. At the same time. We, with our Dragonfly Edge and backside solution, we're now able to take that learning and use it for in-process edge and backside, along with all surface requirements. Mike will talk later on about advanced packaging needs. The outgoing quality requirements are also becoming very stringent, so the need for reduced defectivity, not just in the pattern area, but the edge and backside, is becoming important for outgoing quality as well. This also significantly increases our opportunity for macro inspection.
In summary, increasing sampling in OCD, right. Applications extension and SAM growth in films, and then macro inspection. These are the three drivers that we believe will propel our opportunity as the industry migrates to Gate-All-Around. Next, I'll move to memory. Over the last several years, bit growth density requirements have really developed, have really resulted in, especially for 3D NAND solutions, a design architecture which required the bits to be laid up on top of each other. Now the transistors are growing in the vertical direction. The current generation devices already are at over 20 microns tall. The etch processes that create these deep trenches create aspect ratios that are north of 40 to one.
These are extremely complicated structures to manufacture, and the reliability, and the design, and the device yield properties require a lot of very carefully monitored applications and measurements. A lot of these require multi-stacking processes, which also mean that the error budgets from one process step to the next process step, they accumulate, and that results in additional opportunities for process control. In the lower half, 2D DRAM continues with its incremental scaling. We see incremental acceleration of EUV layer adoption into 2D DRAM. We also see migration of 3D stacking in the packaging area, where DRAM, DRAMs, dies are laid on top of each other, connected to a GPU or a CPU base. Then the longer-term roadmap for DRAM probably is it, is in a similar direction with 3D DRAM. Once again, structure is going tall.
This is where the roadmaps converge. Now we have to measure properties of materials along very tall structures, and we also have to still have the sensitivity to make very fine measurements all the way along the structure, and that opens up opportunities for Onto Innovation. Similar to our assessment of the opportunity in logic, memory roadmap also increases our intensity. Our estimation is up to 30% growth can be achieved in terms of our entitlement, based on making high aspect ratio metrology as the vertical scaling continues. Looking at complex thin films and amorphous films, these are materials that are very hard to measure. At the same time, looking at submicron defect inspection requirements coming in from both wafering as well as packaging applications. Here's an example of a blow-up of a 3D NAND structure.
As you can see, our customers are interested in several parameters across the entire channel hole processing group, in this case, looking at very thick films, elemental composition, thinner films, individual films, looking at geometries, monitoring these geometries along the structure, along the sidewall. All of these are extremely complicated measurements. At Onto Innovation, over the last several years, we've been working with our customers diligently to identify and come up with targeted solutions that proactively meet the requirements of the industry. We believe that as the industry scales, several of these would transition into a high volume at graduate and high volume manufacturing.
Next up, with the interest and the rejuvenation in AI architecture, there's processes in the wafer backend, specifically tied to high-bandwidth memory, and the scaling roadmap, which allows us to also look at opportunities in both metrology and inspection and packaging. Today, the majority of the architectures are relating to stacking of 8 DRAM dies on top of each other, with a roadmap to go up to 12 DRAMs. Along with decrease in the die pitch and some of the die parameters, the value that we provide to the industry there is in having a solution that incorporates unique metrology sensors with our macro inspection system. Now our customers are able to look at defectivity sources through stack and defectivity sources, looking at local uniformity variation. On the right-hand side is one such customer example.
In a competitive head-to-head, what we demonstrated is that the dispositioning of good die from bad die, using Onto's unique metrology sensors, was much more accurate than our competition, and that allowed our customers to come up with higher yields. In most cases, we've been working with our customers to really tailor our roadmap to their roadmaps. As the slide shows, we've been able to qualify ours across multiple customers with over 30% of emerging high bandwidth memory applications. As this goes into high volume, we expect continued opportunities for Onto Innovation. So far I have talked about our current applications, right? Metrology, CD metrology, films metrology, and inspection, these are applications that we serve today.
The advantage that we have with the breadth of our offerings between acoustic metrology and optical metrology, and pushing those limits, is that now we have an opportunity, with a seat at the table, to really look for additional problems that we can solve for our customers, and in many cases, our customers are coming to us. What's shown here is our two specific areas that we are very excited about. One is around compositional metrology, and one is for material property characterization. These solutions or these requirements are currently being solved by Onto's current products. They are in early stages in terms of the initial need being served for process qualification, where our customers are characterizing their new processes. They're looking for composition all the way from bare silicon to NAND and DRAM and logic devices, but then also looking at material properties such as stress, conductivity.
These are elements that we believe as the control requirements get a lot stricter, some of these have the potential to go in line. My key message on this one is that as our technologies continue to solve problems for our customers, even solutions for problems that may or may not require in-line monitoring today, could in future require a higher volume solution as tighter budgets come into play. I think that's my last slide. With that, I will hand off to Dr. Mike Rosa to talk about Onto Innovation's opportunities in specialty and advanced packaging.
Okay, good morning, everyone. Thank you for joining us here today. Today, I'm gonna talk about unique technologies for growth markets. The semiconductor industry, we believe, remains on track to exceed or meet $1 trillion in end device sales through 2030. This continued growth in the industry is underpinned by the growth in new applications. We've certainly heard of electric vehicles, we're hearing about AI, ChatGPT, all these various applications, high-performance computing, AR, VR, hyperscale data centers, and so on. The growth in these applications, we believe, will continue to unlock trillions of dollars end economic value globally. At Onto Innovation, the role we play in the semiconductor industry gives us the privilege and opportunity to support that tremendous growth. All of the growth in end-user applications is gonna be enabled through device technology trends, and with those trends come new challenges.
Srini spoke a little bit about the advanced node technologies in logic and memory, and certainly the advanced CMOS in a relentless pursuit of price, performance, area at the lowest cost possible. They introduce new materials and new unit process technologies, which present metrology and inspection challenges, such as some of the ones that Srini outlined. Today, I'm gonna talk a little bit about the specialty technologies, and specifically within specialty, I'm gonna call out power technologies: silicon carbide, and gallium nitride. We're gonna talk about the new compound semiconductor materials that are introduced in those markets, and some of the process challenges that the metrology and inspection technologies that Onto provides, helps meet in the production environment.
We're gonna talk about advanced packaging, and specifically call out panel-level packaging, and the role that that technology plays in supporting high-performance computing and bringing that to market. Today, we've seen a number of examples in the commercial space for panel-level packaging in gaming or to support AI or server applications. Let's talk more specifically about power. We see the growth of the end device power market to exceed or meet $50 billion by 2030. That market growth, and it's been quite phenomenal in the last recent years, has mainly been spurred on by the global green initiative. Whether it was at first, solar power and the development of inverter technology or wind energy and so on, driving the use of power device technologies, now we hear a lot about electric vehicles, the electrification of heavy machinery and industrial applications.
Mike talked about home use applications, electric lawn mowers, things of this nature, all driving a large semiconductor content around the power segment. In all, we see that segment of the market growing at around 65% through 2030. It's very, very strong and continues to show tremendous growth and provides a lot of opportunity for both Onto and its customers in this space. If we look at silicon technologies, all of the growth that we talk about around compound semiconductors, by no means diminishes the role that silicon plays in power today... Silicon technologies at both the small wafer size and the large wafer size for discrete and power management applications see a lot of application in consumer electronics, as well as commercial and communication applications.
In the consumer space, we call out smartphones and mobility, as well as IoT or edge compute applications for their need for low power device technologies. A lot of those today are delivered on silicon. gallium nitride is a new compound semiconductor material we hear about in those markets, used primarily to deliver fast charge capability. These are the chips that bring your phone charge time down from an hour to five or 10 minutes, and we see tremendous growth potential for these technologies going forward.
In all, one of the points I wanted to make on this slide is that through both the uptake in electric vehicles, heavy industry, and fast charge technologies across a broad range of application segments, the compound semiconductor class of device wafers is set to grow by tenfold between now and 2030. That's not gonna be without its challenges, as I mentioned, new materials and new process challenges. Certainly, we're gonna see a lot of wafer size migration from 150 mm to 200 mm, 200 mm to 300 mm, and so on, across different material sets and different device structures.
All of these present both unique technology trends that we're seeing today play out over several segments of the industry, they also bring to us very unique technology challenges in high-volume manufacturing environments that our metrology and inspection solutions are uniquely positioned to address. All of that growth in the power semiconductor space translates more broadly, and I wanted to call out one example here in the automotive space. Everyone's familiar with the phenomenal growth that the smartphone has seen, I don't know, over the last 20 years, 15 years or so. Today, smartphones by unit volume, outpace passenger cars by around 15 to one. There are around 1.3 billion-1.5 billion smartphones sold every year, to the 90 million-100 million passenger cars sold every year.
A very interesting fact, which I'll walk you through, a little-known fact, which I'll walk you through now, is that while smartphones contain around 150, on average, chips per device, to a passenger car's several hundred, depending on your model and the capabilities, between now and 2030, all of the applications that are coming on board for passenger vehicles in the automotive space, full electrification, battery electric vehicles, ADAS, infotainment, onboard communications, vehicle-to-vehicle communications, vehicle-to-X communications. All of these new end applications in the passenger vehicle are gonna drive silicon content relatively through the roof. We anticipate that by 2030, the passenger car will have roughly 20x the number of ICs per unit compared to the average smartphone.
If you were to conservatively look at those numbers grossed up on their total unit volumes, that gives you roughly a 1.3x to 1.5x the number of ICs or chips used in passenger vehicles compared to smartphones. If you combine that with the new technologies in photonics and MEMS, advanced logic, mature logic, and the mix of those technologies in the passenger car, that's why I've whimsically termed this slide, the potentially enabling the next killer app for the semiconductor industry. If we take a look at the mix of the power device technologies in the automobile today, we see that power represents roughly 20% of the total die count in a car. We see that number increasing to over 50%.
Combined with the mix change between sensors, photonics, as I mentioned earlier, that's gonna create, as I've explained, a significant opportunity, not just for the power device manufacturers, but the industry as a whole. Okay. Right. What are some of the things that Onto Innovation brings to the table via its unique process control technologies to enable the success of these compound semiconductor and silicon materials in the high-volume production supporting the growth in the power market?
One of the very unique things that Onto Innovation brings as a company, not just to power, but across many of the specialty and mature node segments, is this simple fact: a lot of the very unique and specialty technologies that Srini walked us through on the advanced node side, both for memory and logic, migrating those back to the smaller wafer size and making them available in mature and specialty markets, gives us a very unique and distinct advantage in the manufacture of these technologies with our customers. Some of the ones I've called out here include the Atlas family of critical OCD products. These are becoming instrumental in looking at either thin films or CD structures for silicon carbide, gallium nitride, and the very deep silicon structures that we see in the 1,200 and in excess voltage, discrete power MOSFETs.
We have technologies on the inspection side as well, such as Dragonfly and EB40 platforms, that allow us to inspect all sides of the wafer, the front, back, as well as the edge and bevel of all of the wafers, for defect and particle inspection of both bare and patterned wafers. Combining that with our overlay metrology on the IVS platforms, and then also some very unique epitaxial layer metrology technology in the form of our Element family of epi metrology tools. Lastly, I wanted to call out our very unique acousto-optic technology.
This is a technology that allows us to measure the thickness non-uniformity of single or multi-stack layer opaque materials, typically deployed into the power device customer segment, where we see very front, frontside metallization that can be in the range of 4 µ, 5 µ, or 6 µ thick, and very important to be able to trace the non-uniformity of those metal layers throughout the final stages of the fabrication process. Enough on the technology. I wanted to point out the banner across the bottom. We are present at each of the top power device OEMs in the industry. The capabilities of Onto's products, both from a metrology and inspection perspective, allow us to touch 80% or more of the unique process steps in the entire process flow of a power device.
That, coupled with the unique capabilities that we bring to the market, lead us to estimate a greater than 3x opportunity growth for Onto between now and 2030 in the power device segment. Okay, while we, of course, apply a lot of our technologies individually, in different customer scenarios, we also look for opportunities to combine those technologies into portfolios that are able to specifically address the technical challenges of very specific steps in the process flow. Here I've highlighted a few examples. One I will call out is in the final metallization step, where we combine both our Echo acousto-optic technology for measuring the thickness non-uniformity of final front side and backside metal layers with our Dragonfly product to do a final quality inspection of that layer before the wafer is sent to the packaging house.
On the right-hand side of the slide, I call out a detailed example of where we use our Echo family platform, it is the platform that we deploy this technology on, in conjunction with our run-to-run and fault detection and classification software products, which are a key part of our Discover software ecosystem, at the customer site, in order to monitor the variation in thickness non-uniformity during the final fabrication steps of these power devices. This is really critical at a stage where we're seeing, especially in the silicon carbide world, the migration of the wafer technologies from 6 in to 8 in. This is very important to reduce non-uniformities across the wafer and within die, as customers are moving from one wafer size to a larger wafer size in production.
Okay, if I can switch gears a little bit now and talk about the advanced packaging space. Since the mid-1970s or the rough 1970s, advanced packaging was mostly about multi-die modules. We've seen the evolution of that space, where today we refer to advanced packaging in terms of fan-out wafer-level packaging or 3D IC. 3D IC mainly involves the vertical stacking of chips on a large common die or platform. That technology is now evolving into what we call heterogeneous integration, shown here on the slide. Heterogeneous integration is essentially taking your CPU or GPU, taking those functions, taking support functions off those die to allow them to limit the total size, and placing the chips around the core processor on what we call an advanced IC substrate or panel. Okay? Okay, panel level packaging.
This is an example of a typical panel-level packaged device, high-performance device, that you might encounter in the market today. As I mentioned earlier, we've seen these devices in gaming, advanced gaming. We've seen them in support of AI technologies and high-performance computing. We expect this market to grow tremendously. This particular example here, shows a GPU at the center of the advanced IC substrate, surrounded by high-bandwidth memory. In other incarnations, we'll see either GPUs or CPUs combined with high-bandwidth memory or with chiplets around those die, deployed in what we refer to as multi-die architecture. When we talk about chiplets and panel-level packaging, we ask the question: "Okay, that's great, but why now?" Couple of basic reasons. The first one is, as complexity of GPU and CPUs increase, the dies tend to get larger.
As the dies get larger, the manufacturing cost increases and yield comes down, you end up with low-yielding, very expensive die. One of the things that the industry does through chiplet-based architectures to mitigate against that is to introduce node optimization. Essentially, what that means is that we take the large GPU or CPU, any RF or analog or mature node technologies that are also on that same piece of silicon, we remove those, and we place them around the die on the advanced IC substrate, as shown in the image here. That does a couple of things. It allows you to limit the size of your core processor, at the same time allows the yield of that device to increase, which gets the cost under control.
One of the knock-on benefits of that is that we see tremendous capacity adds across the industry in the mature node space. Remember, all of those RF, analog, and ancillary technologies that were once on silicon have now been removed and are placed on the advanced IC substrate or panel in a supporting role. Lastly, we see green processes. A very rough calculation posits that around 3%-5% of the waste that's produced. Or sorry, let me say it this way. Roughly 3%-5% of consumable waste can be reduced by moving from a large monolithic die to a chiplet-based architecture. All that simply means is that through the inherent complexity of making a very large, fully integrated die, right? You're consuming more waste materials, waste chemicals, et cetera, et cetera, right? You're producing more of them, I should say.
Limiting the size of that device and moving to chiplet-based architecture, you're reducing the waste in the monolithic die process, and you're also increasing the capacity adds needed at all of those customer sites that produce those chiplets. Overall, we see between now and 2030, a doubling of panel capacity required in the marketplace. We see through the onboarding and evolution of die and wafer-level integration technologies. We might have heard of hybrid bonding as one of those earlier. We see a tripling or greater than 3x use of hybrid bonded wafers or die integration technologies. By 2030, with all the discussion around chiplets, we're seeing greater than 18x ramp in the number of chiplet-delivered packages.
Consider the impact that that has on the additional green and brownfield sites in mature node segments of the industry, as well as high bandwidth memory, that would be required to support those advanced IC substrate-delivered technologies now. Okay, front and center in Onto's product offering for panel-level packaging are two key product families: the JetStep suite of lithography tools and our Firefly macro inspection products. I won't belabor this slide too much. I'll call out one or two unique technologies for each of these products. The macro inspection technology, similar to even our Dragonfly wafer-based technologies, on panel, use a unique Clearfind technology.
That is very unique to Onto, and it differentiates us in the market because we are able to, during the panel fabrication or lamination process, we're able to detect voids or particles that are trapped under these build-up layers as they're being formed. The technology also allows us to detect residues, organic residues, and particles around the panel during its regular manufacturing process. The JetStep family of lithography tools, very high overlay accuracy, very good depth of focus, which allows us to deal with changing topography on the panel very easily. One of the features which I didn't call out here, was the large field size capability, allowing us to image a greater portion of each panel with every single shot, increasing the overall throughput for our customers.
We can deploy each of these products, as in the previous case, independently, offers tremendous value to our customer. Being able to leverage the Discover software ecosystem and utilize both of these tools in tandem to look at specific examples of where, I've got one image here, the die may be placed already on the substrate, then an epoxy mold applied, and in some cases, the dies may shift. We can track that shift using our Firefly product, feed forward those, the new location information to our JetStep tool, and then we can reduce the field size of our lithography shot to address the rotation on a per-die or cluster basis.
It's a very unique and a very nice example of how we can bring together disparate technologies on the inspection and the lithography side, meld them together with the power of our software to deliver something truly unique to our customers. These type of examples, plus the growth in the market, we believe that there's a 4x opportunity, growth opportunity, so good, I'll say it twice, for Onto between now and 2030. If we go back to the wafer-based side, we talk about the chiplets and the die-bonded and wafer-bonded devices, the 2.5D and 3D multi-die systems that are supporting this growth, we look at our Dragonfly products for macro inspection on wafer....
We can handle 2D and 3D inspection, plus a myriad of the typical defects that you would expect to see on a wafer, from cracks or post-die, sorry, post-SOR die edge inspection, or when you're stacking these devices, we're able to measure the height of the stacked die and to confirm any co-coplanarity issues across the entire wafer. That's the current generation. The next generation of challenges that speak to the onset of hybrid bonding, talk about contact recess metrology, global dishing across the entire wafer due to the CMP process, contact defect inspection on the individual contact, which might be as small as 10, five, or below microns in width, as well as the capability to measure thin films, thin-film dielectrics on the same tool platform.
If I bring everything together, I'll leave with one final message. I would say that Onto has been very successful and continues to be successful in delivering unique, advanced technologies, both on the advanced node front, the mature specialty, and advanced packaging side, evidenced by a broad range of the product families that we have here, whether we're talking about the Element with its dual mode, dual transmission mode capability, the acousto-optic technologies on Echo that I talked about, or our Atlas and Iris products with their 32 channel spectral engines, the Dragonfly and Firefly products use of Clearfind technology, the JetStep family of lithography products with its large field of view, allowing throughput benefits for our customers above and beyond that of our competition. Connected software across the entire product family is essential.
Leveraging the Discover software ecosystem is one of Onto's strengths, leveraging that at the customer site, not just on a per tool basis, but across multiple of our products and third-party products to address the customer's higher value problems in their production environment, positions us extremely well for not just satisfying the customer today, but future growth for the company, as well as riding the wave of the growth of these very exciting device segments. Thank you very much. That's all I had. Let me pass it on to Mark.
Thank you, Mike, and good morning. Obviously, you've heard about a lot of great technology, and certainly the team is well-positioned to drive into these spaces and continue the exciting growth that we have ahead of us. Over the next several minutes, my job is to provide that confidence that we have the infrastructure and the foundation in place to support this growth. As Mr. Sheaffer alluded to earlier in the opening comments, you know, we're committed to get back to our long-term operating model. That model, obviously, we wanna get back to $1 billion starting at $1 billion and in excess of 56% gross margin. In order to do that, we're not waiting for the market to return. We're not waiting for our higher profit margin products to return to get that volume back.
We actually have activities in place. We're executing as we're executing manufacturing cost reductions. We already have some operating cost reductions in place that we're effectively running in 2023. We have the profile within the litho margin improvement program to drive margin accretion within our lithography business. We're also gonna grow our software, our services and software business, which I'll talk about in a few minutes. Again, here, we're committed to get back on plan and continue to drive the accelerated growth that we see and we've heard about. As we talk about the cost reductions, Mike alluded to this in our Q1 earnings call, that we have programs in place.
We're excited to say that we have in excess of $25 million of cost reduction programs specific to our manufacturing and supply chain teams. Those costs will be realized over 2024 and into 2025. Again, that provides the foundation and the support, so when we see that revenue and that growth come back, that we're certainly going to see that margin leverage to get us back to our model. Areas that we're focused on, common supply chain. Right now, with our manufacturing sites, we deal with several suppliers. We're working through supplier synergies, bringing that all onto common platforms and also driving our contract manufacturing strategy. When we working through subassemblies and component parts, we're gonna drive those to contract manufacturers.
What that does is that provides obviously, leverage within our existing infrastructure, but it also provides the drop-through on the profit with the margin. Again, certainly excited about the activity we have in place. Again, just to reaffirm, you know, we're not waiting for the margins to come back with just higher margin products. We actually have plans in place to execute that. Another area that I wanted to focus on was our capacity. Again, just like our cost reduction programs, we have our capacity in place to address our long-term operating model, and we're committing to be even above our long-term operating model from a capacity standpoint. Over the last two years, we've been expanding our clean rooms at three of our four sites to address the capacity needs.
As I mentioned, we'll continue to see efficiencies through as we work with contract manufacturers, we'll provide additional capacity within our plants. One of the areas that we haven't really utilized to its fullest extent is our shift differential. We see a dramatic opportunity as we look to drive second and third shifts through our current facilities that aren't currently in place. A lot of this activity coming out of the merger several years ago, these are activities with the growth that we've seen over the last several years, these are exciting opportunities that we're gonna continue to untap to drive that capacity that we need to address that growth that we expect through our long-term operating model.
Now, we've not just been focused on manufacturing and our supply chain, we've also been focused on our services and software business. Our model, our teams have been evolving the model over a period of time, we wanna move away from a transactional-based model and move towards a more recurring and value-based approach. As the teams have been shifting in that direction, there are certainly opportunities when we look at our service contracts, expanding our service contracts, software contracts, and also moving to sub-subscription-based model. Our goal here is to really drive with partnership with our customers around not just transactional-based metrics, but tool productivity metrics. With that, we hope to untap and continue to drive further expansion within our services and software business to drive that increased profit that we'd expect.
As we've seen, the team's done a great job of driving significant growth and profit through the last several years. We have a great install base. We've continued to see great renewal rates in excess of 90%. With our lithography business, we'll expect to see increased service programs there as well that will increase the profitability in our service business. Again, these are things that when we put it together, there's a lot of, I would say, potential here that we're gonna untap as we continue to see the software and services business continue to grow. Tying it all together, as I commented before, we're committed to getting back to our long-term operating model. We do feel that we have the foundation in place, the infrastructure, the manufacturing capacity to be above our operating model.
You'll hear more about that from Mike in a few minutes. Again, with this, we have the strategies in place right now to get us back on track and to make sure we can recommit to getting back to the $1 billion in excess of 56% gross margins. Again, we also will, you know, from a cash generation standpoint, we've been very focused on our working capital metrics, so we expect this to get back to a strong cash generation over that cycle as well. Again, recommitting to the long-term operating model. Certainly, you've heard a lot of exciting technology that will help propel that growth. Again, thank you very much. With that, I'm gonna turn it over to Mike for his closing comments.
All right. Thank you, Mark. Thank you, Mike and Srini. Great presentation. Before we bring the people up for discussions on... Or for Q&A, I'd like to just summarize with probably one of the first questions that'll come up, which is: Okay, you've shared all this great stuff. What's it mean? What's it mean for revenue? What's it mean for the business? As you know, companies don't like to tell what that all means. It puts us on the spot. I'll attempt, with the caveat of a bunch of variables that we can't predict, I'll attempt to at least build the bridge for what all these exciting transitions mean for Onto. First, we'll start with the $1 billion, last year's revenue, and then we'll look at the incremental improvements that we see from the different categories you heard about.
We'll start with films and Gate-All-Around. If we look at the films and the Gate-All-Around transitions, the projections that we see in the market and our positions in those markets, we expect to be able to add an incremental $200 million-$250 million to our business over the next five years. Specialty market, power and specialty. It's definitely a much smaller market, has very nice growth rates. For us, we see a lot of new product introductions that are moving into the power space, giving us this broader product portfolio to offer power and drive what we think could be another $200 million or so in incremental revenue over our peak in 2022. Similarly, panel packaging.
You heard a lot of great things from Mike about this transition to panel packaging, not only being with AI, not only driving high-performance compute, but also going into other packages, complex packages, that wanna be integrated and take advantage of the benefits, performance benefits from panel-level packaging. Here we see lithography, as you know, has already been growing quite well and very strong. Incrementally, lithography will continue to grow, but we see process control becoming a bigger part of this, and we expect that to add incrementally about $150 million or so. Then, if you add in a little bit of M&A, which we've been very vocal about getting more aggressive with, $200 million in M&A, and you can start to see over the next five years, reaching $2 billion in revenue.
Not quite as fast as doubling in three years, but we think that we're well positioned to achieve this. You'll notice here that I didn't mention this new surface metrology. It's a small component of our organic growth plan, but you, many of you may have seen it out there, and I would encourage you, after I introduce this, to go visit Erik Novak in Technology Row and see what this technology is all about. This is a great example of how, through the breadth of our technologies, we're able to expand and move into some new markets. What is surface metrology? That's not a wafer. That's the James Webb Space Telescope.
James Webb Space Telescope. If you see at the bottom there, you can see the humans, give you an idea of the scale of this telescope and those mirrors, 350 m squared. The surface and the alignment of these panels, so each individual hex panel or mirror, and its alignment to each other, has to be precise down to nanometer scale. That's critical for it to actually see the 8 billion trillion mi of the galaxy, that, or multiple galaxies that it's looking at. How do you measure that surface? Actually, with our technology. With Onto's patented 4Di technology, which is unique in its ability to be vibration immune, we're able to measure nanometer-level precision across a 50-m distance.
To give you an idea what that means, if you used any other technology, any other interferometry technology, just breathing in that space would have rendered the results useless. Because of the technology that we were able to deliver to NASA, we're actually awarded or given an award, yet another technology award from NASA. Again, investors, you're all wondering: Okay, what's the big market here? How many James Webb telescopes are you going to actually support? Not too many. We've been using this technology in optics labs for the last five to 10 years. In the optics labs, I mentioned that the unique capability of this technology is its vibration immunity. Optics labs are a very controlled environment, so they have vibration isolation tables, they have clean environment. They don't take advantage of our vibration immunity.
In this space, we've delivered over 1,500 systems already. The market share, we're number two in this market, with about 10% share. A dominant number one player with about 70% market share, and then the next nearest competitor to us, they all have less than 2%. That gives you an idea of this market. I haven't talked to you about this revenue or this product line for a while or ever, and it's because it hasn't been a big part of our business. What we're seeing is this shift towards digitization of the industrial market, and you see or hear a lot about Industry 4.0, but what I think is really driving this shift, at least in the last four years, is a shortage of labor and the higher and higher cost of labor.
We're seeing more focus on automation in these markets. The markets we're focused on applying the surface metrology to is automotive and then the aerospace, both airlines and the manufacturers of the jet engines. Why this is exciting for us is because these are very dynamic environments. You're not in a clean room, you're in an area where heavy machinery is moving, large, heavy blocks, engine blocks, panels. You're in an area where jet engines are taking off, creating, you know, massive disturbances. This is where the vibration immunity is going to be critical. Surface metrology applications here are going to require our technology. To give you an idea, here you see the Airbus CEO and some comments he made about the importance of this shift for his industry.
I wanted to share with you what this means from a potential market size and a potential growth opportunity for us. I mentioned the optical market already, and I showed you where we were in market share, 10%. It's about a $200 million SAM market with relatively low growth. Again, that's why we haven't really talked a lot about this technology. As we see the application of this technology into these new markets, we see a potential SAM increase of $400 million combined, leveraging all those three circles I talked about. Obviously, a high CAGR, because you can see here, it's a very, very small part of the market today. We're helping to evangelize, we're helping to enable this market. You might be wondering, "Okay, what exactly are you going to do?
What is surface metrology as it means for automotive?" Well, I was surprised to learn that 30% of the cost of a car is paint. That's kinda shocking, I kind of didn't believe it. When I poked, what's happening is there's a tremendous amount of rework, very high percentage of rework. It's not just the cost of the paint, it's actually the rework cost as well. They must have, given this high-value problem, they must have very precise tools to measure the paint quality. There's your precise tool. Right now, you take these panels, you bring them all together into a car, and somebody has to physically look to see if these panels match up, line up, and look right. Chromatically, I think they probably have tools. They're not looking for paint and color per se.
They're looking for the surface finish. Here you can see actual images that we've got courtesy of General Motors, where we've been working and actually co-authored a paper about our technology. Actually, I'll just jump right to it. Leveraging our technology, instead of catching these panels that don't have matching surface capabilities or aren't in spec, we're able to use the 4Di technology that you'll see out there to quantify that surface metrology at the point of painting, not at the point of assembly, where it then has to be disassembled and reworked. It's a significant cost savings. Right now, we're at the stages of proving out the sensor technology.
We're working with several automotive manufacturers that are using our sensors today to qualify and to determine how well they work, and if they do what we believe and what we presented with General Motors, then they would move to an automation step. Aviation. This is the other area I mentioned. There's a couple areas, but I'll just focus on the turbine blades. You can imagine the importance of the quality of those turbines, the blades. Any kind of divot, crack, any deformity under stress, can break up. No one wants that while the thing is 30,000 ft in the air. How do you think these are tested or measured today? Another very precise instrument, fingernails. I know, it's scary, anybody who flew here, but it is fingernails.
Today, and you'll see videos, people are paid to run their fingernails along, looking for cracks, and they're taught certain depth. Thing has to be disassembled. Obviously, here and anybody flying home, you'll be happy to know we're further along with this market and these customers. Here, we're able to bring the 4Di technology and articulating arm to automatically look at and qualitatively verify the performance of or the surface of these turbine blades. Again, you can see here it saves tremendous amount of time, but also I would guess it adds to safety. Here, we're further along, as I mentioned. We're working with two of the jet engine manufacturers. They represent about 85% of the total market. We've delivered one fully automated system, we expect to deliver two more by the end of the year.
These aren't very high dollar value items, so it's not a huge revenue jump right now. Again, we're starting to see this growth and adoption, and we believe it can be a very, more meaningful part of our business in the next five years. Many of you have heard me talk about synergies and the importance of synergies in any kind of business expansion that we focus on. You might be wondering, what the heck are you doing jumping into the industrial market? There are a lot of synergies. First, it's not a market we're not in. We've been selling about $24 million in revenue per year to this market already, so we have channels to this market well-established over years. Just hasn't had the growth that we've expected.
Based on the transitions and the needs you just heard about, and by being able to leverage our automation expertise and the software capabilities we have in the FabWide arena, which actually do apply to this space, our ADC could apply to this space. Our ability to track parts and to trend deformities or defects over time could apply to this space and drive an increased value proposition in the industrial market. At the same time, we're seeing opportunities to leverage this surface metrology capability from the 4Di sensor into our semiconductor market. We're seeing applications for Primarily in 3D, so for die stacking, for high bandwidth memory, and hybrid bonding, that's also potential opportunity for synergy. It's a very nice part of the business. It's accretive to margins. Hopefully, it drives up this growth like we've shown here.
Yeah, so that's what I wanted to share. Now, before we bring people up for questions, the panel, I'd like to just wrap up, kind of summarize everything you just heard. First, we're aligned to multiple high-growth markets. I think if that didn't come out, then we really failed. We are well-aligned to multiple high-growth markets, not just with one or two systems, but with a portfolio of solutions and software that really provide a differentiation that matters to our customers. A lot of this product portfolio is relatively new or added in the last couple of years. That's helped us to expand our SAM, as I said in my opening remarks, from about $1.5 billion-$2 billion to $3 billion, giving us a lot of room to grow. Now, Mark talked about our ability to meet that growth.
If we drive to this $2 billion mark, what's it mean for factories, and output, and expansions, and capital outlays? We believe we have the potential already to meet about a $1.8 billion or so revenue output. Of course, I should have make that greater than $25 million, but platform synergies. Again, we see an opportunity to take advantage of this somewhat of a slowdown, to drive the next wave of synergies that we haven't been able to implement over the last three years, and take another $24 million or so out of the business, out of the cost structure, primarily through supply chain, COGS, platform synergies, not people, not R&D, which will end up driving or resulting in over an $8 per share earnings power.
We've always been a very strong cash generator, feeding that M&A engine and allowing us to invest in additional opportunities. With that, I see I've got zero time on the clock, we should invite up the panel and open it up for your questions.
Did you finish his mic?
Oh, I got the middle chair. You are now.
Okay. Craig,
First of all, thanks for the very informative, presentations, guys. The question I wanted to start with, Mike, as we look at the concluding remarks you had, I think revenues add up to about $1.8 billion± . I think the target model goes to about $1.4 billion.
Yeah.
How do the target model parameters for gross margin, OpEx, operating margins, work from the 1.4%-1.8% mark?
Yeah. Mark had that in, and it looked too good, so I had him remove it. It was just very.
You can have it back. No, we, there is certainly acceleration there, right? There does become a point when we're probably gonna make some additional investments, right? When we get to that $1.7 billion, $1.8 billion. I think we have confidence that from a margin standpoint, we'll still see a relative margin expansion and operating margin expansion. I think in this case, you know, the model does look attractive, and we continue, you'd see the consistency there, you know, the consistent accretion and drop-through that we'd expect. You know, I don't, I don't think at $1.8 billion or $2 billion, we're gonna drop back significantly.
Mike, the follow-up, before I hand it off, you had a number of areas where you expect the business to grow significantly. Admittedly, we're in a low visibility environment, but can you just talk about your confidence in some of the bigger buckets that you're looking at that take us from the $1 billion up towards the $1.8 billion? Could you comment on how reasonable the current environment is for M&A as you do that? Thank you.
I'll start with the first. We have high confidence in the opportunities for growth in those markets. For all the secular drivers we talked about, for all the different investments our customers are making, we're very confident in the growth trajectories. The question is, how confident are we that we're gonna gain the share and grow outsized within those markets? So far, our confidence is actually increasing. You know, we continue to talk about in the earnings calls, new wins, share gain positions that we've, you know, gained each quarter. Those are usually new entry points, where those investments will drive repeat revenue and more volume revenue from those entry points. For us, what we presented, we're fairly confident in.
You know, we can always get disrupted or blindsided, so far, that hasn't happened. We're actually excited about what we see for the future and how we've expanded our position to take advantage of that. Going back to your M&A question, I would say the environment for M&A is maybe getting, you know, better on the lower end, so smaller tuck-in types. The higher interest rates, the higher inflation, investors getting more nervous about funding some of these businesses or continuing to fund them is making, you know, making some things a little better. As far as, you know, bigger, more transformative options, I would say that the environment hasn't changed significantly.
It's still about synergies and, surprisingly, social issues and, you know, overcoming those barriers so that both sides see the benefit of coming together and realize that benefit. Quinn?
Thanks, Mike. Quinn Bolton from Needham. I guess, you know, nearer term and think about the next WFE cycle, we peaked at about $95 billion, ± in 2022. If I make an assumption that WFE gets to something in the range of, say, $115 billion-$120 billion by 2026, just let's use that as a working assumption. It would imply about a 20% growth rate off the 2022 peak. Where do you think you fall? Would you hit your $1.2 billion model? Would you hit the $1.4 billion model? Yeah, would you hit the $1.8 billion model?
Yeah, probably not the 1.8, that I can say. You know, it depends. It depends on what drives that growth. Is it Gate-All-Around transitions? That would be very good for us. We'd see a lot of outsized opportunity there. Is it high bandwidth memory that takes off, and memory really drives some of that? Or does it end up on the more of the specialty legacy and panel? It's a little bit harder to answer. I do think we would blow past our 2022 numbers. If we look at the, you know, a 20% increase over 2022, I think we're positioned now that we would certainly blow past those numbers.
When you say blow past the 20% growth rate that you'd outperformed over the years.
Yeah. If we do the simple math and kind of take everything straight, we've expanded our position in every one of those markets. We've got a track record of outperforming the growth. If the growth was 20%, we would be disappointed at only growing 20%. If we were $1 billion before in 2022, and you grew 20%, we'd be disappointed at $1.2 billion, right?
Got it. certainly comfortable with $1.2 billion, and then hopefully closer to $1.4 billion, would be a reasonable expectation in that WFE sort of environment.
Yeah, definitely. Yeah.
The second question was just, you sort of teased us with some of the new, materials metrology, and I'm just trying to go back to my notes here for a second.
Surface metrology?
It was... Sorry, the, in the materials characterization-
Ah.
Opportunity. I think Srini talked about composition, metrology, materials, property, but that wasn't included in your sort of new opportunity. I was just wondering, could you spend a minute talking about what's the opportunity on some of those new materials, characterization, metrology?
Yeah. I'll just frame it very quickly, and then I'll let Srini answer. It's not included because it's actually part of several of our systems provide both dimensional and materials information, characterization information. We've talked about Atlas, we've talked about Echo. These things can do film thickness, but also provide material information. We have other products that, you know, Srini will talk about, that are exclusively for materials characterization. That tends to be, right now, a bit of a smaller part of the market. That's why you didn't see that in my numbers. Srini can explain a little more details on where he thinks that business can go for us.
That's Mike's correct. There's two aspects to the materials market. One is in the bare wafer silicon composition and concentration measurement. Today, those are dedicated platforms that do specifically that, and they are the sampling rates, depending upon the end user requirements. For us, I think as those requirements get tighter, right, the sampling rate and the attach rate for those systems, it's mostly on the elemental composition, those should increase. The challenge, I mean, or the uncertainty more is as we go in the fab, as Mike mentioned, the solutions that we have in place today for this process qualification, these are solutions that customers are bringing in for a traditional OCD measurement, they're bringing in for other thickness measurements.
For us to drive incremental opportunity, the sampling for those types of measurements, like the stress measurement, the materials acquisition measurement, those sampling rates have to significantly increase, and I think maybe at that point, it can be carved out as a separate bucket. Right now, if I look at the application mix, it's still predominantly the third markets, and these are relatively smaller.
Great. I'll apologize, but I'll take a third question if I could. We've gotten a lot of questions or a lot of excitement in the advanced packaging space around hybrid bonding. Was wondering if you could just explain how Onto may benefit from increased adoption of hybrid bonding. Does that benefit your inspection business? Are there other process control systems that may benefit from growing adoption of hybrid bonding?
Since you guys are always hear from me, I'm gonna let Mike take that.
Oh, okay.
Hybrid bonding.
Yeah. Looking forward, certainly, you know, hybrid bonding is an area that we're certainly looking at for sure. In my presentation, I highlighted a couple of areas where there are opportunities for our inspection technologies, and there are also opportunities for our unique metrology technologies. I won't go into any of the specific approaches that we're considering. I would only say that, yes, there is opportunities, or there are opportunities in both camps, inspection and metrology, at the wafer level, at the die level, and potentially involving multiple platforms.
Hi, Brian Chin from Stifel. Thank you for the presentation. Very, very comprehensive. I guess maybe first for Mike, relative to the baseline of a $2.5 billion-$3 billion served addressable market at the moment, when you get to sort of that super peak of $2 billion, what do you think your SAM looks like, I guess, at that level?
The $2 billion doesn't require any SAM expansion. Obviously, what we presented within the $3 billion would be that. That would be based on the continued work we do, looking for new applications of our technologies in other areas, like you know, we brought up the industrial today. There, I would say we don't have a lot that I can share right now, as far as where else we're looking to apply our technology. From the $2 billion... Are you saying, where does our existing SAM grow during that time period?
Where the SAM would be in five years would encompass growth in those served in the markets that you play in, as well as maybe some additional, yeah, sandboxes that you might.
Yeah. I thought I did understand your question, right. We can't say where else we're going to apply that's gonna go for, you know, some significant SAM. Obviously, the knobs would be M&A. You know, do we open up a new SAM through M&A, acquisition? Do we open up a new SAM by applying our technology into some other markets that we haven't talked about yet, both of which we'd be, you know, not wanting to share right now until we're done.
Yeah, also the incremental revenue expansion from the $1 billion baseline last year, where would Dragonfly sort of fall within some of these various incrementals?
Hmm.
I would think Panel is more like JetStep and Firefly.
Yeah, correct.
Where would Dragonfly kind of fall?
Dragonfly is across the board, so the submicron capability from Dragonfly is opening up opportunities in the advanced nodes in the front end. I think Srini mentioned the OQA side of that, but I think there's more as well, and we have roadmaps to continue to drive our capabilities on the resolution and throughput side for the Dragonfly. In addition, of course, packaging is always where it's been strong, and we continue to see, again, expanding opportunities in things like hybrid bonding for Dragonfly in those markets. Sorry, Panel is also an opportunity for Dragonfly. Firefly, for sure, and full panel inspection, but oftentimes these customers are quartering up the panels, and that's where they're gonna do some bumping and then additional inspection requirements there. We do sell today Dragonfly for quarter panel applications.
Okay, and maybe, Mark?
Mm-hmm.
I think at the $1.2 billion target financial model, I think you tightened up the EPS, the low end of that range, towards the high end. Is that mainly reflecting the $24 million in OpEx center cost synergies?
Yes.
Okay.
Yeah.
Okay.
Yeah.
Thanks.
Questions from this side of the room? All right, Graham.
Thank you, Mike. How are you? Oops! Whoa. Thank you, Mike. Thank you for a great, very extensive presentation. If you were to achieve a 20% CAGR in whatever environment, I'm just wondering if you could sort of break that down into growth. What % would be from new customer applications, new markets? What % would be from new products, and what % would be from increase in average selling price, sort of in the fashion of ASML, who's able to jump up their prices every year with enhancements and all? If you could just sort of break that down. Thank you.
We don't have that broken down, but I'll give you kind of a framework to work within. Most of the growth that we'd be talking about is from the new applications, our entry into new markets and new applications that we've not been participating in before. I'd call that 60%. The films, for instance, the growth we've seen, the adoption of our OCD for Gate-All-Around, panel market, and all the process control requirements that we expect the panel markets will need, and that we're delivering on with Firefly and Dragonfly. That's the bulk. We do see the ability and have been for years, increasing pricing as part of, you know, improving cost of ownership, improving the natural performance of our system, so the natural evolution.
I wouldn't say that's a big part of the growth we're talking about. In fact, it's probably not even in the numbers. We probably just assumed flat ASP. Then you asked, you know, from a point of view of new applications, you know, growth. The other big segment, I mentioned about 60%, probably 20% of that would then be the growth that we talked about. The growth in those markets, how we positioned ourselves in those high-growth markets, that would also be a big contributor. Then the, I think the remaining piece might be new applications, but again, like I told Brian, if there was something new that we haven't talked about yet, we wouldn't, we wouldn't mention it here yet.
A follow-up?
Sure, taking the follow-up. Craig Ellis, ENP, rather. I think this is a Mike Rosa question. Mike, can you talk a little bit about how the company's exposure within power breaks down between IGBTs, silicon carbide, GAN now? As we look out over the model period, how would you expect that to evolve? Said differently, what are the higher growth parts of that market opportunity going to be for you?
Interesting question. At first pass, I would say that, you know, we're pretty evenly spread across silicon, IGBT or super junction, discrete MOSFET technologies, right? Either at 200 mm or 300 mm. The silicon carbide and gallium nitride technologies that I mentioned operating today in the 4 in, 6 in, and 8 in, 200 mm area, right? Our technologies are pretty evenly distributed across those. We're attacking very specific high-value problems on each. However, I will say, I also mentioned in my presentation that we have a unique demonstrated capability in bringing advanced node technologies back, right, into the smaller wafer sizes.
We're seeing more and more of that, both in newer products like the Atlas S that we released recently, but also on some of our, without going into the product details, mid-IR OCD products that we can use to address very specific challenges in the power space on silicon. As you can imagine, right, silicon carbide device maybe can handle a 1,200 V, 1,400 V gate voltage, let's just say. Silicon needs to get much thicker in order to handle a comparable voltage. As you raise the thickness of the material, we run out of gas with certain wavelengths that our OCD tools use.
Our ability to bring advanced node mid-IR from the 300 mm realm to a 200 mm or maybe even 6 in, but in this case, I'm speaking specifically about 200, opens up tremendous opportunities for our customers. In answer to your question, I would say we're very evenly spread across all of them, interestingly, across our entire portfolio. There are pockets of examples of where we're bringing 300 mm technology to address specific problems that aren't addressed today in the industry, quite frankly, on 200 mm silicon. One of the interesting things, without belaboring it, is you will see 200 mm silicon continue to evolve. They haven't given up the fight against the, you know, the silicon carbide and gallium nitrides of the world, and that creates challenge for us.
Not just challenge, but opportunity, because as they grow that silicon, the wheels start to come off the existing mature node technology that has historically been sold into those markets, right? They look to new technology, which we are in a great position to offer.
That's very helpful. Thanks, Mike. Srini, the follow-up's for you. Totally understand the coming architectural change with Gate-All-Around, it's being led by all three manufacturers at once, so it'll be different than FinFET. The question is this, though: Beyond Gate-All-Around, there are things like backside power that are coming, and I don't think I heard you mention that. Is that just a smaller incremental opportunity for the company, one that is outside of the time period that you were looking at? Just help us understand what something like backside power would mean for Onto. Thank you.
That is definitely an opportunity that we currently are engaged with customers in, and in some cases, it's a little bit agnostic. It fits well with FinFET geometries as well as with Gate-All-Around. Both in the metrology and inspection space, we do see new requirements emerging from backside power delivery, and we see overall activity. I don't think we have quantified internally what it could mean in terms of incremental SAM, but maybe, I don't know, Mike, you can comment on that.
I would say that, we have not, at this point, quantified what it would mean in terms of dollar SAM. I would say with confidence that we've got, engagements with the tier ones on both the inspection and the metrology side, there are additional opportunities, high-value problem type opportunities, beyond our current engagements today.
Mm-hmm.
for our products.
Making Amy Shah call away?
Hi. Thanks for taking my question. This is Vee from Jefferies. I want you to double-click on the hybrid bonding opportunity. Currently, most of the packages use, you know, the 3D packages use bumps. As we move into hybrid bonding, how is the competitive landscape changing? Like, we hear a lot of front-end tools being levered into back end. Does that mean this segment gets a lot more competitive than it used to be?
Go ahead, Mike.
Will I jump in? Okay. Yes, and no. A couple of things are happening, right? At the moment, with bump technology, it's a traditional technology, right? There are the incumbents there. Those bump sizes as device densities, node densities continue to scale, you need more contacts, right, on, to pull out the signals, which are going to drive the size of the bumps down.
...At some point, we run out of gas, right? The industry runs out of gas and needs to switch to hybrid bonding, which basically means you remove the bump in the equation, and you just have the contact pad. This is where we might talk about a gray area, front end, back end, something in the middle. What we're starting to see, one of the things that doesn't change. What we're starting to see is that, yes, there are new entrants into this market looking to address hybrid bonding opportunities. However, the cost sensitivities of the back end, even though it's now still may be referred to as this gray area, haven't necessarily gone away, because those cost sensitivities translate into the end application, right?
The trick here is to deliver cutting-edge technology that can address the challenges, some of which we pointed out, even though qualitatively, for hybrid bonding at a very cost competitive price. Just by that definition alone, you start to weed out various alternatives and various players in the market, especially when you look at the front-end players. Yeah.
I have a follow-up. Focusing on the advanced node side of things, you know, your typical exposure, not in 2023, but typical exposure has been a third, a third, a third between DRAM, NAND, and logic. As you know, the next two, three years, you've talked about Gate-All-Around being a bigger opportunity. Does this essentially change that exposure to each end market?
I'll answer that one. It really depends on what happens with the FinFET and how long those nodes stay in play and what expansions may happen there. So is this a transition from one to another, where we just have a growing opportunity because our entitlement is bigger? Or is it both in play? Now, both stay in play and continue to expand, then, of course, we'd see a shift to that model. The other thing is, you know, memory has its own exciting drivers, HBM memory, and you probably have all caught the headlines. You know, bottlenecks for AI and for mighty Jensen Huang, there is memory and DRAM. He needs more memory, and he needs more high bandwidth memory. That's also a very strong area for us and, you know, could drive outsize growth.
It's really hard to model just because of the Gate-All-Around. We have to look at the entire environment.
Thank you.
I think we have time for one more. Graham?
Graham Tanaka Capital Management. A couple of macro questions. One, the AI elephant in the room. What, just in general terms, what looks like a pretty rapid shift to a higher proportion of GPUs versus CPUs on the processor side? What does that do to your total TAM and to your margins going forward? Secondly, I mean, related to that, if you could just sort of estimate what GPUs are as a percent of total processors' value today and what it might be, say, five years from now as a benchmark. A second macro question is sort of the China syndrome.
What do you see happening as potential options down the road, should there be some issues with China, export controls, et cetera, and maybe even, God forbid, you know, some problems with Taiwan? Thank you.
To answer the first part of your question, the shift to GPUs or CPUs, it doesn't really change our opportunity so much. The GPUs, you know, is pretty well-publicized. They're being built by TSMC, obviously a big customer for us. All we're looking for is growth, and growth in any of those devices, whether it's from CPU or GPU. What was the second part of the question? Do you remember?
The mix.
Oh, that I don't know. Do you know how the industry will shift the mix from CPUs to GPUs?
No, I don't have those numbers on top.
I think Jensen answered that question the other day. I think Jensen said he thinks it's double GPU to CPU.
Then the impact of, I mean, the impact of, you know, further escalations with China over Taiwan is obviously significant. I think you see countries and nations trying to mitigate that risk now, by having TSMC set up shop in Japan and the U.S. Germany is now trying to entice them as Germany, Europe, I should say, is trying to entice them as well to build, you know, advanced fabs outside of Taiwan. Yeah, I don't... Well, actually, it's an artificial. For demand, it does nothing. It's artificial.
The whether they fill that demand, they'll build the shells, they'll maybe build a single line or a pilot, and then they're not gonna really, my view, they're not gonna really expand and fill that full capacity for those factories, those shells, unless they see the demand. Otherwise, it's a big waste. I don't expect there to be this big, huge oversupply situation. I would expect there to be the capability, so if anything happens, there's the backup. Then any new capacity demand increases, yeah, I think they would fill it in those other regions. They've made comments about, you know, costs going up, the efficiencies are more challenging outside of Taiwan, et cetera, et cetera. What happens when governments get in the middle of economics?
Hi, Hendi Susanto from Gabelli Funds. A different question on China. There are trade restrictions on China, but there are also some pockets of growth opportunities. Several areas that come to mind is the surface metrology, for example, I think you can still sell to China. I'm wondering about software as well. How do you view the market opportunity in China, aside from the trade restriction for the next few years? Thank you.
Yeah, that's a good question. Certainly, we've taken a hit, and we publicly talked about the trade restrictions wiping out $80 million-$90 million of revenue from us for 2022. The team has shifted gears. Our team in China shifted gears and focused a lot on markets, tier two, other specialty markets, like power, in order to fill that gap. You know, it'll be tough to fill it all in one year, but we do see progress, and we do see opportunities to fill the gap as we exit 2023 and move into 2024 and to 2025. What does fill the gap mean? Historically, we're around 20%-25%, you know, revenue from China, and we think we can get back into those levels. Now, that depends on China's growth as well.
You know, recent numbers have come out pretty sobering.
Hi, thanks for a great presentation today. Can we talk a little bit about the software segment, specifically, how much of the software revenue is recurring today? Are there opportunities to increase recurring software revenue as a percentage of the total? Thanks.
Yeah. Yeah, well, we haven't split out software, specifically, but, you know, when we look at what the opportunities are, the growth in that business, you know, gets to be a larger portion of the total business over the operating model. Certainly from, you know, where we wanna take the model and evolve it to, as I mentioned, kind of the value add and overall, just looking at our, a subscription-based model, you know, in the near term, you know, that does create a higher profitability level as well. When we look at services and software together, they get to become a bigger part of the total pie, and also from a profitability standpoint as well, we see accretion there, so.
Well, thank you, everybody. Lunch is now served in the vestibule, as well as please pick up a free travel mug for hot and cold beverages, and we're gonna give you a bag to put it in to carry it home. Thank you for all for coming, and have a great day.
Thank you.
Thank you very much. All right.