KLA Corporation (KLAC) Earnings Call Transcript & Summary

welcome to the NASDAQ Investors Conference in partnership with Jefferies. Let's go live to London.

Great. Okay. I think we can get started. Very honored to have KLA at our conference this year. We have Oreste Donzella, and he is the EVP of Electronics, Packaging and Components. So Oreste is the -- he's been at KLA for 20 years. And he's held leadership roles in the field of customer engagement, process control solutions, application development, strategic marketing and product development. And before joining KLA, he was at Texas Instruments and Micron, where he held various engineering and management positions and process integration and yield enhancement. So Oreste is going to give a presentation. And if there is time at the end, then we'll have time for a few questions for fireside. So Oreste, welcome.

Thank you, Mark. Okay. Happy to be here in sunny London. I mean, there is a [ meet ] that rain and for gear, I've been living now. I have a flat in London. So it's spending a little bit of time in the U.K. with the acquisition of SPTS, that is a U.K.-based company, and Orbotech that has presence in Europe and Israel. So for me, it was much easier to manage all this organization from here, not from California where my home is. And it was a [ meet ] because, hey, it's sunny weather all the time in London. So why people are complaining about? Okay. Everything started from this. So I'm an engineer in my heart. And transistor is one of the most beautiful product, probably the most beautiful product that human being has invented. And everything started from the transistor. And then we have evolved pretty much in the IC, semiconductor and venture in the old beautiful things that we do every day in our industry. And one thing that happened in the last few years, I've been advocating this for several years. So in my -- as Mark said, I've been working in KLA for 24 years. So before then, I worked in TI and Micron for 7 years. So total of 31 years in semiconductor. All my life has been spent on semiconductor. And I've been advocating the big shift, the secular shift in the semiconductor industry in the last few years, is the fact that we are not driven by a single driver anymore. So if you look at back in the history, of course, semiconductor industry start from military application, military satellite, bringing the man on the moon. And then, of course, it was the PC transition and the Internet and the mobile. But now, it's everything. Everything because really, the digital society is driven by data, by how it transmit data, analyze the data, monetize from data and so on. And the semiconductor is at the core, is the front and the center of this data revolution. So the fact that we have multiple drivers and tightly coupled with AI makes a semiconductor industry, a semiconductor capital equipment industry. Absolutely an exciting place to be even after 31 years of action. So KLA, of course, you may know KLA, I think you know KLA. It's a world-leading company in the semiconductor process control, started 47 or 48 years ago. And we have done one thing for 47, 48 years in one market. That's it. So process control for semiconductor wafer fabrication. Then eventually, in 2019 -- 2018, 2019, we figured it out, we need to diversify. We need to start playing more of a diversified role. Also because we thought about the electronics ecosystem is going to be driven by many, many drivers, as I said, the data, AI and not only one driver per time. So there is a huge potential to grow not only semiconductor, but in entire electronics ecosystem. And then what we did, we bought a few companies, and I'm going to talk a little bit later, and we aggregate this company under EPC, under my group. So KLA is split in 3 groups. It's really 3 companies in one. Each one of us is -- their own P&L. Their own financial statement. We are present to the Board our results. We are fully accounted over our results. Below us, there are divisions. Divisions are split by product type. For example, we have inspection, metrology. We have inspection for PCB and so on. And then this is the way pretty much our KLA is structured. Very simple. So we run 3 different business under a kind of a holding the Tiscali corporation. And Rick is my boss, the CEO of the company since 2006. So this is what I was mentioning before. So this is our diversification play. So KLA semiconductor process control, dominant position in wafer fabrication. But we do more. So we started actually in 2008, when we put a company in Belgium called IQOS. IQOS was and is still the market leader for final inspection of packages. And really, it's the final, final step in assembly and test that is done before shipping the package to the integrator, to Foxconn to put inside the smartphone. And we took this company for a long time. We kept standalone. So we kept separate from the rest of KLA. And then in 2019, this was a pivotal moment for KLA. So we bought multiple companies in a single transaction. Everybody knew the Orbotech acquisition. But in reality, with Orbotech, we got multiple companies in a single -- under a single umbrella. So with Orbotech, we got our footprint in the PCB industry, printer circuit board, in display industry, out of Israel. And then we bought a company in the U.K. That's the reason why it's much easier for me to travel from here in Wales. It's called SPTS, it's a crown jewel of EPC. It's an organization I'm very proud of. And it's growing quite a bit. When we bought in 2019 was less than $300 million. We -- last year, we did $530 million revenue. So almost double in the 3 years. So -- and I'll talk a little bit more about SPTS later. And then we bought also inside Orb, there were 2 other joint ventures, 1 in PCB software in Israel, 1 in PCB lithography in Germany. So the aggregation of all these companies now are part of EPC. And the goal of EPC is to provide KLA diversification play and to grow in the markets where frankly seeing, we are not the leaders. So we are partially the leader in some of this market. It's different than the play that we have in semiconductor process control services that are attached to Semiconductor. We are the clear leader of process control in semiconductor front end. In all these markets, we are making a position right now. And one market that I like to talk a little bit more is package. So this is put together, is a rider for an investor conference in New York. And if you look at semi PC, this is the historical KLA. The historical KLA in the mask or we call reticles, in the substrate wafer and the chips in the wafer. That's it. This is what we do. When we started EPC with all the acquisitions of this company, we started getting in the wafer level packaging in the assembly and test in the printed circuit board and display. So we came closer to the end products. So now give us the possibility to talk to our customer. What a bit? Actually, when you talk to the smartphone makers, of course, they know PCB and display much more than the wafers. So they can guide us in what requirements, which kind of requirements they need from the entire industry in the next 3, 5 years. That means our ability to understand what the market needs from a technology point of view in the next 3, 5 years is much higher because we have these prices all across the electronics ecosystem and not only in wafer fabs, and this is very important. The second thing we did was there are -- and I remember, I was in New York City in 2019, when we did the post-acquisition Investor Day. And at that time, I said there are 2 things that make me exciting about EPC and all these acquisitions we did. One is automotive, one is packaging. In fact, these are the 2 slides that I presented at Investor Day in New York, where you see $1 billion package income that is new to KLA. We didn't have a big presence in packaging in 2019 apart from this final inspection made in IQOS in Belgium. And automotive is another targeted strategic initiative. Now today, I want to tell you the rest of the story, what we did from 2019 to today. Let's talk about Advanced Packaging. Actually, I mean, we get something right. We've got something wrong. We cannot predict the future. I'm failing all the time. I try to predict the future. In one of my many jobs at KLA, almost all my job in semiconductor industry was technology job. Only for 3 years, I took a detour. My boss Rick asked me to do the CMO and worked very, very closely with IR and corporate communications and so on. At that time, it was the time that I was building the market forecast and analytics. And I got to tell you, I get wrong all the time. So it's very hard to predict the future. But these 2 things are predicted well. One is the importance of packaging the semiconductor roadmap. The other one is how automotive industry is going to change. And let me go through these trends very quickly. Advanced Packaging, semiconductor roadmap, more low, 50, almost 60 years. 1965 when Gordon Moore said that it's going to be scaling semiconductor. And mostly sending you that to scale for 1 reason, lithography. Because lithography was drawing the lines and the contact and the space to pack more and more chips on a same piece of silicon. So -- but the certain point, although Moore law is still alive. He's not dead. We are still in the semiconductor, wafer front-end fabrication, making a huge innovation. And by the way, the biggest innovation of all comes from Europe in ASML. They are making state-of-the-art tools. I'm going to tell you, I have a lot of respect. When I saw the EUV scanner was state-of-the-art tool absolutely for lithography. And we keep investing. We keep innovating in front end. So Moore law is not dead. For example, EUV scanner is moving to INA UV. And every time there is a technology inflection for KLA, there is a growth opportunity. We are more exposed at KLA. We are more exposed to technology inflection than capacity ramp. Because, of course, when the technology is complex, there is a huge shift in the road map, they need more and more process control. So UV, huge driver for KLA in the process control for front end. Get all around, another huge driver because metrology got to change completely to track the profile of this structure. So we still see innovation. We still see growth of KLA in the core front-end semiconductor. But this is not enough. And this is where packages come into the play. I remember 10 years ago, I was so long in semiconductor, actually, even 25 years ago. I remember the packaging fabs were like real packaging fabs. Box, a box wrapping around the chip, okay? That's it. And then connected with wire bonder, you are done. Now it's a serious technology. It's a serious technology happening in packaging, happening because of data, because of AI. Because these CPU, GPU, FPGA are becoming so complex that you cannot rely only on lithography for front end. And then it started with -- this is simple packages you start integrating logic and memory on top of PCB. Then you did the 2D packaging, the 2.5D chip level, and now everything is becoming vertical. And there are beautiful example of this from our customers. So if you can see AMD, the Epic server processor with the GPU, CPU logic controller and stack HBM memory on top of each other. You can see NVIDIA, of course, the core GPU, the 2 Intel example for PC and the data center. There is a huge technology happening in packaging or what we call Advanced Packaging right now. So we need a different play for packaging. So now this is the time for KLA to play a role in packaging. If packaging didn't have a technology inflection, forget about KLA. It's is not going to enter. But whenever we see the industry needed help with the scaling the technology, this is a fertile ground for KLA. This is what semiconductor used to be. Actually, this is what electronics used to be. We have a front end. We are packaging. We have substrates. 3 several worlds. substrate is done in PCB. Actually, this was the main reason why we bought Orbotech, who was not for convention PC display, but because we understood high-end PCB called IC substrate will become a critical, critical part of semiconductor roadmap to packaging. Packaging, of course, becoming more and more critical, more relevant and front-end scaling. So I put this animation together to show that's different. The way our semiconductor scaling has evolved is not the same as before. Before, we have 3 several worlds. Now everything is connected because the line between these 3 worlds are becoming blurrier because packaging is getting the similar technology complexity of front end as the substrate is getting the same technology complexity of packaging and more and more automation is getting there. So that's the reason why when we bought Orbotech SPTS, we figured it out, I want to be in the entire circle, and that's what KLA is. Of course, KLA is a primary actor in front-end semiconductor, now is becoming a more relevant player in packaging. And because the Orbotech and the latest investment we made in IC substrate, we are also a player there. Let me talk a little bit more about the challenges, what KLA can offer. So let's start from this first circles here. Front-end semiconductor, the 3-nanometer TSMC and 5-nanometer and Samsung memory, and Micron memory and so on. This front end semiconductor. This is packaging. You see the package is getting more and more towards the front end. The geometry is getting smaller. The complexity of vertical stacking is getting very, very critical. So that is on packages getting to the front-end space. And in the front end, you need sophisticated process control, too. This is what KLA makes pretty much a great company. In packaging, when I was asking 5 or 10 years ago to packaging OSATs and say, "Hey, do you need this kind of sensitivity differ detector?" Say, go, oh, I don't need the year? I don't need the spectrum, you need an expensive tool. I make a little inlet margin. So I don't want to spend this money with KLA tools. So by whatever is available, if I need to expect that we do expect. I go all the way up to the final test, and that's it. And then I ship the package to the integrator. Now it's different. Without process control, you don't make any yield from packaging. In the past, they had like 99.99% yield. Now forget about it. And we need to bring some process control technology from front end to packaging. It's not been easy. That is why we form my organization as well EPC because we represent all KLA products in packaging under a single responsibility. So we bring tools, inspection tools from front-end semiconductor. We adapt for packaging because different requirements. For example, the front-end semiconductor is very standardized. So thickness is always the same. Work edge is very, very flat. When you get packaging, it's like a wild jungle. I mean, everything is all over the map. We have thickness from 50-micron to 1.5-millimeter. Work edge is all over the place. So we need to adopt these tools to the packaging requirements. So we need to do some sort of customization, and this is for inspection. And of course, the beauty of KLA is because we produce the best tools of process control to special methods for front end, we can port some of this technology in packaging AI. We have been using AI for 10 years in front end. Packaging AI, they have no idea. So we brought AI into packaging. Now all our tools, especially metrology tools use AI to find the defesne, classify the defense and measure a critical portion of the wafer. Plasma dicing. This is what we do, by the way, mostly in the United States. Plasma dicing done here in the U.K. in Wales in Newport. Plasma dicing all is ready to deal with plasma dicing. When you have a wafer, you need to singulate the wafer. You need to dice the wafer, okay? And then you start having a simulation of the chips. Then you do some other wafer level packaging, assembly and test and package. Now this simulation is mostly done by blades. It's a mechanical type of process or laser this processing use a lot of defense. So you end up cracking these times when you pass pretty much the blade. When you do plasma etching perfectly clean. Perfect. It's done clean. And of course, you don't have cracking of the edge of the die. And actually, the die looks better, aesthetical. Well, I like aesthetic of the as well. and you can improve the yield a lot. Now plasma dicing has been in the work for many, many years. Nobody was buying until KLA connected SPTS in Newport to the big guys: Intel, TSMC, Samsung, Micron. And now people know the value of having a very clean process from front end and also what these guys are the guys also driving the packaging roadmap. That's another reason why KLA is in this market because we wanted to get the collaboration we have in front end, moving to packaging with the relevant players. And now this business is parroting say this is the unit sale of plasma dicing from SPTS last year went up to the roof. So that's another way our packaging is changing. Before I do the blade dicing, I crack a bunch of dice. Now you cannot afford anymore. When you have chiplet with multiple chips, one DRAM is going to be cracked, all the package is gone and these are spends. Now let's talk about the other 2 circles that is probably not well known. So Orbotech, of course, we bought Orbotech. We got SPTS anyways, a great acquisition. We got an enter in the packaging process. We got the specialty semiconductor, I'm going to talk about automotive later. But with Orbotech, we got access to PCB. And frankly speaking, I was not really excited about the conventional PCB. What I was excited when I got this new opportunity in KLA was Substrate because substrate are extremely relevant for the packaging of chiplet integration for CPU, GPU. So also this line is getting blurred because the substrates are becoming much closer to semiconductor than the conventional PCB. So I'll give you an example. If the conventional PCB on a smartphone, the PCB where you land like the processor or the fusion process for the phone is now around 40-60 micro line space, 40, 60. So you got a line -- drive the line at 60 micron. Easy, but substrate, they need to go find mic. And going in 1 or 2 years from 46 to 5, 8 microline, it's a big deal. There are hundreds of PCB fabricators in the world, I believe 1,000, but there are only 6 people who can do IC substrates at the level that AMD TSMC Intel, these other guys require. So this is the reason why we are in substrate. We didn't buy Orbotech because of conventional PCB. We bought because we saw an opportunity to make a difference in scaling this PCB technology and the substrate for packaging. So what do we do? We have a lithography tool. It's probably not very well known. When you think about KLA, you think about process control all the time. Now I told you we have a process organization in Wales. And we are also an lithography organization in Germany that came from the Orbotech acquisition that now pretty much is going to release the very, very advanced de-lift imaging lithography tool by the end of the year. We will get new tools out of the market in late end of the year that will guarantee growth in the next 2, 3 years. So this is done in Germany, where there is pretty much the Optics capital of the world, Vienna. And also, we brought some technology from semiconductor into the PCB world in terms of inspection including AI. Now we have AI also in PCB. Think about it. I mean, you never expect PCBs to be so sophisticated and implement artificial intelligence for different section. But this is what we do. These are carriers. So we don't buy a company because we want to keep the companies they are. With by compares, we want to improve the company by putting our model, operating model across our technology. So what is Advanced Package for KLA? Look at that. So pretty good. 2019, this is 2017-'22. Of course, we got SPTS that. Is the blue. The blue part is SPTS. The process organization in Wales. And then you have -- sorry, the IC subs that is the blue with Orbotech and Vienna lithography, EPCs, SPTS, and then process control is the core KLA packaging. So our numbers look pretty good. Now how sustainable is this? Of course, the market may go down. This year is not a great year overall for electronics, the semi cap and so on. But we expect the long run we should be able to, for sure, to have double-digit growth on a 5-years CAGR basis because we have many products now coming in the life. So I believe we have a prospect of growth and nobody else. And also because we are in the 3 circles. We are the only company using the 3 circles, as I said. And the other thing I'm excited about, I'll stop in 5 minutes, Automotive. That's a completely different play. So packaging, advanced packaging is mostly driven by PC data centers and advanced smartphone. Automotive is a different type of industry. And the industry is changing completely. I have the story to tell you in 7 years ago, the different job at KLA. One carmaker from Germany came to KLA. We don't sell tools to carmakers. We buy cars, but we don't sell to them. And they came with our -- one of our customers. Their supplier. So it was an upward meeting with our CEO of the company as well, having the customer bidding up our customer in front of us, and we don't like that. The reason was this customer or supplier of the carmaker, they were failing the qualification of the chips all the time. And they are failing because of reliability concern. So reliability is paramount in automotive industry. because of the liability. I mean personal liability if the chip fails while you drive the car, you don't want to be in that car. And the second thing is we don't believe the cost associated to recall is huge. So this reasons are our biggest paramount. The problem is automotive industry has been leveraging what we call N minus 2, N minus 3 technology, lagging technology. Technology they were in the smartphone 10 years ago. But now it's different. You want to do autonomous drive. You go to San Francisco right now, and there are plenty of cars driving autonomously. I saw a dog crossing, not under the -- really on the intersection, and he was getting hit by the car, the Waymo car settled it even. I mean, it's dangerous right now. But at the end of the day, they are trying to make this autonomous drive real, and it will take a long time. Electrification, requiring 4x the semiconductor content than a normal gas engine car. So all this connectivity, every driver likes to be GPS and Wi-Fi, and they have all the sensors running the car. He requires advanced chips. But advanced chips are more reliable because they are not mature. So they started telling us, "I need you to help with reliability. You are the master inspection company in the world. "And then we also map and we say, first of all, the semi forge going up to the roof for auto. The second thing is the diversification of devices. So again, in the past, for the autonomous car, we have analog, power, silicon-based, some trailing chips, that's it. Now we have 5-nanometer chips in the car. And you have silicon carbide material that is super immature. So they are afraid. So what we did was we did 2 things. So this is for the methodology. So we developed an AI, artificial intelligence-based methodology to potentially track reliability fails. So what happens? So one, during the fabrication of the wafer, of course, from the beginning to the end of the chip. So we have multiple process steps that you control. At the end, you do the final test. It can be functional, electrical test. And then you do [indiscernible] and all the automotive. So they stress the chip to make sure it is robust. And then they ship to the carmaker to the Tier 2. They put on the board that they put in the car. So sometimes, we escape. So the test doesn't see the defect. Sometimes the defect gets only criticality in the harsh condition that the vehicle operates. For example, when you have under a big heat, 110-degree fahrenheit, you can stress the chips that pass the electrical case, but now they are failing in the car. So that's the reason why we build metrology AI that compile all the information from the massive amount of data that is generating in the test house, in the wafer fabs, in the packaging house, and the carmakers, and then pinpoint the most critical part of the chip that is going to fail or the most clear process step that the venture will lead into this related reliability phases. We developed these methodology is a software algorithm that is applied to our inspection tool. And now it's part of the AEC standards. AEC is automotive electronic council. So every semiconductor in automotive, semiconductor automotive needs to comply with this guideline. And it suggested that they use I-PAT that is our methodology. Second, another reference to our facility in Wales, our organization with Wales, SPTS. Silicon carbide, of course, is everywhere. So people are really getting excited about the electrification of the vehicle. That is based on silicon carbide. You cannot have the proliferation of electrical vehicles based in silicon. You don't have the property of the material driving this efficiency in the electrical vehicle engine and the possibility to scale down the battery size. So what you do is we develop customer solutions in Wales for silicon carbide, plasma and the etching a deposition solution. They are tailored to the silicon carbide to the market, and they have been proven in production already, PVD and edge. And automotive grew quite a bit for KLA. Also automotive, as you can see, the semi-PC, the process control is in down and the lighter color is SPTS in Wales. So you see a pretty stunning number here for both packaging automotive. So in conclusion, and I'm not going to read this. I put there just a final slide. In conclusion, we have very good position in process control for front end semiconductor. This is not going away. Actually, it's going to be more and more important because of technology inflection in semiconductor on the front end, like an [indiscernible]. But on top of that, we have other player right now in electronic system, thanks to EPC, in particular in packaging and automotive are very, very fast-growing market for KLA in the next 3, 5 years. And with that, I finish. Thank you.

Thank you. Thanks for that great presentation.