KLA Corporation (KLAC) Earnings Call Transcript & Summary

Morgan Stanley and I'm happy to have with us today Oreste Donzella, who's EVP and runs the Electronics Packaging Components group for KLA. He's going to give a presentation, and then we'll have time for Q&A at the end. If anybody has questions, please e-mail them through the system, and I can relay those. Just quickly, as a safe harbor. For important disclosures, please see the Morgan Stanley research disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley representative. So with that, Oreste, turn it over to you for the presentation. Thank you.

Sure. Thank you for the introduction. Hello, everyone. I'm happy to be here, at 45th NASDAQ Investor Conference. So forward-looking statement for more details, please refer to KLA IR website. Today, I'm going to talk about a couple of topics. One is the introduction of EPC that is Electronics, Packaging, and Components group that I have the honor to lead. And in particular, where we are with the process of integration. EPC was born as an aggregation of multiple companies that we bought in the last few years. And then I'm going to dive down into packaging market, packaging technology and explain how packaging is becoming an interesting driver for growth at KLA. So let me start with EPC first. It's all about the market expansion, diversification. That's the story of EPC group. So everything started when we acquired Orbotech and SPTS in February 2019. It was a single transaction. We acquired Orbotech. SPTS was an independent subsidiary of Orbotech. So we got pretty much 2 different companies in one. And the purpose of this transaction was twofold, was really, first of all, to extend our reach into electronics ecosystem, in particular, expanding in new markets for KLA, printed circuit board and flat panel display. The second reason was we wanted to extend and expand our portfolio in the semiconductor, specialty semiconductor and packaging. And this was done with the SPTS acquisition. Of course, with the additional process equipment to our process control portfolio in this particular market, especially [indiscernible] packaging. So after the acquisition was completed in February 2019, our job was to integrate all these companies, and we created an EPC group, Electronics, Packaging and Components. February 2020 was right before COVID started to rise in Europe actually and then the rest of the world. Not an easy time for me to aggregate companies that were globally located in particular in Belgium, Germany, U.K. and Israel from the headquarter in the United States, so -- but we were able to face the challenge. I'm really happy to see how these companies are integrated under EPC in KLA. ICOS was a company that we acquired in 2018 to start our presence in packaging and we found a very nice house for ICOS in EPC when we integrated with Orbotech and SPTS. So -- and now EPC is 1 of the 3 business groups at KLA. So Rick Wallace has been our President, CEO since 2006. And there are 3 business groups. Ahmad khan is in charge of Semiconductor Process Control. This is the core KLA inspection and metrology business in semiconductor front-end fabs. Brian Lorig is in charge of service for semiconductor front end. And then we have my team, EPC, that is spacing across multiple -- market multiple segments, including packaging, specialty and printed circuit board, flat panel display. So one of the narrative behind the creation of EPC was to implement a so-called operating model. KLA operating model is a set of process that we have been developing for many, many years at KLA to make sure that we still rigor discipline in the way how we manage our finance, the way how we develop new programs, in the way how we interact with our customers, our suppliers. And I would like to say that is one of the secret sauces of the success of KLA. And my job as a leader of EPC was also to make sure that we integrate the KLA over the model newly acquired companies like Orbotech and SPTS. When you look at the way how these 3 groups are present in the electronics ecosystem, going from the left to the right, you see like all the chain of an electronic device up to like a board in automotive or in a mobile devices, so in 5G infrastructure bid and so on. So everything starts from [indiscernible] cost and then inform the chips at the wafer level in front-end fabs, and then you go for packaging, eventually assembling test, and we have printed circuit board. The components, the packaging are integrating printed circuit board, adding flat panel display and then you get in the final product feeding all these industries in the new digital society. And then when you look at EPC, it's very much in every single part of this electronics ecosystem and in particular, is closer to the end of product. So EPC gave KLA, the possibility to bring the company closer to the end customers because we are much closer when you have to deal with the product like a component in packaging or a populated printed circuit board or the flat panel display. So that's really the expansion that I was explaining in the beginning of my talk, looking at our EPC and the acquisition that we have done, Orbotech and SPTS and previously ICOS, was able really to expand our market, and get to even closer to the end customers of the products. Now packaging is probably the most interesting portion of the new markets we are serving right now. And the packaging is also including some portion of PCB that we call IC subsets. In this case, you can see the product portfolio that we designed in the last few years, even before or after the integration of Orbotech, SPTS to serve the packaging industry. On the top, you can see process tools that come from SPTS. These tools are plasma vacuum-based. They are either etcher or dicer or deposition machines. And then at the bottom, you see process control. These are inspection and metrology tools that are generally coming from our front end product portfolio, they have been customized to tailor the need some packaging -- wafer-level packaging industry. And then on the right side of this page, you can see our portfolio for final assembly and test. When you think about packaging market, you can think about really 2 markets in 1. One market is a wafer level and one market is assembly and test, and we are now well positioned on both. Now let me go a little bit through some slides explaining why packaging is so important, so critical. And I put like in quotes there everything is changing. That is true. So if you look at how the semiconductor technology is advanced in the last 50-plus years was all about Moore's Law. It was Gordon Moore that said the semiconductor is going to scale every 2 years or every 18 months, we double the transistor density. And the biggest, biggest enabler of the semiconductor technology road map was the ability to scale the lateral geometry of transistor and eventually pack more chips in the same real estate is a piece of silicon. So we were able to do this continuously for many, many years, according to this Moore's Law. And of course, lithography was a big driver of the advances in the semiconductor technology. And the certain point though, the Moore's Law started to slow down. A slowdown, of course, because of the complexity of scaling the lateral dimension of the transistor, but also because it was becoming extremely expensive. And of course, Moore's Law is not there, that we are keeping progressing our lateral scaling, in particular with UV lithography in the next -- last couple of years. And we will continue to scale the latter of geometry -- lateral dimension for a long time, but we need something else to augment the front end scaling in a much cheaper and more economical way. And this is where packaging is becoming very, very critical. And it's critically, it's crucial for many reasons. It's, of course, crucial for costs, is crucial for improving bandwidth. I'm going to talk about heterogeneous integration later that is extremely important to really speed up the bandwidth and making sure that there is no much lagging time in the communication between, for example, memory and the CPU or GPU. It's a matter of power, and it's for end application and the form factors. Overall, it's all about cost, cost and performance. And in this case, when I look at back, I've been in semiconductor for 29 years. Of course, most of my experience is in front-end semiconductor. Packaging is a new territory for me as well in the last couple of years. But I can see that the packaging is becoming now an industry going from protecting a device to differentiating the performance of the entire semiconductor package. So there are plenty, plenty of packaging times available in the market, a tailoring needs for every single industry or every single usage of a semiconductor device. Today, I'm going to go deeper in one of these, but we can talk many, many of these applications and many of the innovations that are coming for every single packaging type. But I want to dive deeper in what is happening for high-performing computing chip like CPU and GPU because I believe this is where the most innovation is happening right now. And is happening because there is a secular shift from SoC that is a system on a chip to pretty much aggregating multiple dies in a single package. So in other words, in the past, it was an aggregation of multiple functions in single die. Now this aggregation from a silicon, a reaggregation of multiple dies on a single substrate. And this is what happens, and of course, all the main players of CPU and GPU are following the same trends, how to integrate multiple dies, different dies, memories, logic control, CPU, GPU in a single package that is vertically integrated to create also an efficiency in terms of bandwidth that I said and also form factor. So -- and this is what we call heterogeneous integration. So it's really enabling the scaling of the technology whilst keeping costs down. And disaggregation of the noncore functionalities makes the silicon chip less expensive and eventually the integration comes at the packaging level where is cheaper to do and more efficient to do. So -- and let me say a little bit about the heterogeneous integration and what are the challenges in integrating multiple dies in a single package, and what are the road maps that are driving these changes in the packaging industry. So a lot of challenges. First of all, we are going through smaller features. So the geometry that generally doesn't change much in packaging. Now it's changing more and more with heterogeneous integration, but also with other type of trends in wafer-level packaging, for example. We have no material. The packages are becoming bigger. We have new failure mechanism. We need to deal also with the fact that a packaging that has a multiple dies inside will have even more challenges to it. So now we are facing an integration of a potential yield losses because of multiple die integration in a single package and yielding very complex packaging is becoming a big challenge. So we have a lot of challenges, a lot of opportunities in packaging for KLA really to play a very pivotal role. And one thing that is important to highlight is what I said in the previous slide. So when you aggregate a lot of dies in a single package, the yield is going to be a function of the yield -- every single die plus the yield that you are obtaining aggregating the dies in a single package. And the deal is becoming a key factor, a key areas of concern in the packaging industry much more than in the past. And on top of that, we also have the possibility that some of the reliability concerns will escape the functional and the final testing and then they will become a reliability concern whenever the chips on the boards are, for example, assemble in a car or in a data center. So for that kind of reasons, we need to really to change the way how the packaging industry is approaching yield enhancement. And I have done a lot of work in the automotive market from front end semiconductor fabs, where we enabled some screening methodology for the quality of the chips in front end for automotive that make sure that we have a prewarning of potential reliability failures in line instead of waiting for the chip to be assembled on a board and eventually integrate in the car. So I believe we need to use a similar methodology for a chiplet packaging to track the reliability. And this is what KLA has a long history and long experience where we can really help the industry. Of course, in order to be successful, this is KLA playbook, we need to work with all the suppliers, all the customers, all the key players. So you have plenty of customers are now investing heavily in heterogeneous integration, multi-die packaging, including the top semiconductor companies, the Fabless and eventually also substrate publicators because subset is becoming a key component of the entire package where you go through this heterogeneous integration. And as I said, when you collaborate what we can offer as KLA is multiple areas of collaboration. We can offer collaboration in implementing new architecture of die stacking process with our SPTS process equipment. We can collaborate with our customers and customer [indiscernible] defectivity reduction in metrology process control. Even by the automation that is extremely important because one of the success of the semiconductor front-end manufacturing has been adopting automation for many, many years to generate enough data, relevant data to create a process optimization. This concept is unknown in packaging. And we will see more and more automation entering the packaging industry, and this is where KLA can provide an incredible help to the entire industry. So the next inflection is heterogeneous integration that is today done generate bumps is going to be doing bumpless. And this is what people refer as hybrid bonding. In hybrid bonding, of course, you don't do thermal compression bump-based bonding but you do copper-to-copper bonding, and this is what we call hybrid. Of course, in this case, you have a huge benefits in terms of speed bandwidth of our efficiency. But on the other hand, you are getting a lot of challenges in terms of inspection metrology integration. And we have done a lot of work. We are doing a lot of work with the top customers in the way how we can develop integration process flow with all the inspection and metrology steps are needed to be successful in hybrid bonding. So to just close my speech, I will say, advanced package is a terrific opportunity for KLA. And we are seeing already benefits in our revenue. So I was in Investor Day 2019, KLA Investor Day in New York City. At that time, we started outlining our presence in packaging. Sorry, there is a type we are the second last call is 2020 or 2021. And you see for us in KLA package has been a very small market, growing at 12% from 2014, 2018, but was only in 2019 when we started integrating Orbotech, SPTS, in 2020, eventually with the creation of EPC. And now 2021, you can see the forecast with all these rising complexity or advanced packaging that we see a huge, huge growth explosion for KLA revenue in packaging. So we see 3 per cab in the last 2 years since the acquisition of Orbotech, SPTS and aggregation to EPC group versus the past history of revenue trend at KLA. So all in all, great time for advanced packages, extremely exciting from revenue growth, from financial, but also for technology complexity and from whatever KLA can offer to the industry. So this is the end of my presentation. I hope you enjoy it. Really excited about talking about packaging, I think, is going to be a big driver for growth for KLA in the next future.

Great. Well, thank you very much for that. That's very interesting. I'll start with some questions, and then if there's anyone in the audience, I think you guys know how to submit questions through the system. I'll come to my e-mail, and I'll relay them. But I'll start. So you're nearing now year 3 since the acquisition of Orbotech, and I guess a little longer than that from ICOS, right? Seems from the outside the integration has been pretty seamless. Has anything surprised you as you've sort of combined the companies and anything in the integration process that's kind of upside or downside relative to your expectations?

Yes. Let me say we completed the transaction of the acquisition Orbotech in February 2019, it's almost 3 years. And we integrated this company. We started integrating this company with ICOS, SPTS as well in February 2020. So it's less than 2 years ago. And I cannot be happier than I am right now to say that we are ahead of our revenue and profitability targets. And I would say, there are a few reasons why we have been successful. And now it seems that this integration has been seamless. Of course, there is a lot of work that is going in an integration of a large company like Orbotech and SPTS together. But I would say -- let me say, maybe 3 reasons. First reason is we have been very, very clear what we wanted to achieve with this acquisition. We wanted to expand our SAM reach into new markets and extend our presence in especially semiconductor and packaging with the integration of process equipment tools from SPTS. So we did that. We have been consistent with our deal cases why we wanted to buy Orbotech and SPTS. Second, we wanted to buy leaders in their own industry. We wanted to buy companies that KLA can provide help, either technically or from a market. And let me give you a couple of examples, technically. I'd like to say, for example, the next generation of automated optical inspection in PCB and substrates is coming with the optical modules that signed by KLA. So in other words, you see already the synergy between whatever was Orbotech before and what is KLA because we are starting integrating the sign from KLA into Orbotech products. From a market point of view, we got access to customers for Orbotech and SPTS that they cannot get any access before. So our ability to open the door with a top semiconductor in packaging, for example, is clearly benefiting the former Orbotech and SPTS divisions. The third thing is the ability to implement a KLA operating model. And I think this is essential. So you asking me, are you surprised? I'm not surprised because we know our resilient, how strong our operating model was. And actually, if I'm surprised a little bit, is in the fact that -- now looking at these large companies we buy, we figured out even more how strong our operating model is. For example, the rigor that we have in developing new technology that any time we ship the first tool out the door, the tool is ready for prime time without a scrapping material, redesigning parts and so on. This is what we call PLC, product life cycle. This is our way to develop technology. Orbotech and SPTS didn't have this kind of discipline before. Now we are integrating and they are getting a lot of benefits from it. So again, I didn't get really surprised for all these reasons, and we have been consistent with the cases. And I believe this was good and seem that the integration is simple as a lot of work, a lot of effort, but no major hiccups.

Yes. Good to hear. So -- and I guess, another reason I was excited for your presentation is that there is so much discussion of advanced packaging from your customers. 10 years ago, I don't remember CEOs talking about it. And now when I meet with Pat Gelsinger or Lisa Su, it's a major topic and not just is something we're talking about, but as a way of really building advantage into their business. How does that affect the fit? Because I think it does really elevate the importance of these technologies to your customers?

Yes, absolutely. As I said, it's a road map enabler right now. So it's a combination of performance and cost, driving big, big interest from the top semiconductor companies in advancing the packaging road map. And we see clear, clear benefit to KLA because the complexity is driving the KLA process control business. And as I said in the presentation, putting all these dies together in a single package is uncovering a lot of challenges in terms of yield reliability and the KLA is there to help. Think about, Joe, if packaging was not going to scale as much as is today, maybe we were not going to invest in packaging. KLA was not going to invest in packaging. Maybe we had like ICOS only for assembly and test, dominating their market and really getting more and more presence in their assembly and test for inspection metrology for assembly and test for the [indiscernible] package. But the way our wafer level packaging is progressing right now is really exciting KLA to invest more and more. And one of the reasons why we acquired Orbotech and SPTS is actually because we saw an opportunity with level packaging to play a bigger role. So that's the reason why I'm not surprised why the top semiconductor CEO are mentioning packaging all the times because this is an enabler and is a differentiator for them. And I'm not surprised either to see KLA investing more in packaging because the complexity will drive more process control and process integration challenges that we can help with.

Yes. And you had mentioned in the presentation, chiplet yield and some of these other issues. And it kind of raises the question, outside of packaging, how do these developments affect KLA's overall process control business? Because you have smaller die, more heterogeneous die, like how does that affect the way people pursue overall yield management and process control?

Very important. And I believe -- first of all, of course, the challenges in process control are becoming more complex because what you said is no die, large package, integration, vertical integration, 2.5D, 2.1D integration, there is plenty of challenges. But on top of that, I believe one thing that -- I'm very excited about it because I like to see a big change in packaging industry is the usage of data and automation. I know the top semiconductor companies are looking at that. We are getting a lot of requirements now to do more data automation in the wafer level packaging, even a subset level that is unknown about it. Because think about PCB, which is not automated at all. So trying to transition into EFEM, [indiscernible] protocol, automation, data analytics in a packaging fab or even a subset fab is something really new evolutionary. So I believe this is going to be a big, big factor in the near and long-term future in this industry. And the KLA, of course, can play a huge role because we have been in data analytics. We are a data company. We develop incredibly advanced sophisticated technology, and we have our customers to solve a very critical problems. But at the end, we are a company that deals with data all the time. So we can help with data automation of packaging industry, subset industry as well.

Great. And I guess, as you look at these trends going forward, I mean it's -- we've seen already a pretty big advantage to companies like AMD with maybe 4 die in the package. But some of the complexity going forward seems pretty significant. Intel was telling me that Ponte Vecchio, I think at one point, they're talking about 70 different tiles in Ponte Vecchio. So really, in order of magnitude more complexity than we're seeing today. How does that change the opportunity from your perspective? What opportunities does that open up as you get...

Yes, increasing, increasing process control, increasing data, increase in automation. I want to mention also another point that we are seeing on the process side, not process control, but process side, we are seeing more and more adoption of plasma dicing. So today, the way all the customer single level wafers in multiple chips is with a mechanical dicing or laser dicing. But because the clearness standards for hybrid bonding or in general for heterogeneous integration is so important. Now people are looking more and more to dies wafers to single level wafer with plasma. The top is a completely new opportunity for our SPTS division here U.K. So it's another where you see KLA playing a big role is something that you couldn't expect before SPTS or Orbotech acquisition, KLA was becoming a key player in a plasma vacuum-based type of technology. But this is happening in packaging because many, many things are happening.

Great. And I guess we have one last -- last question for me would be, the intuitive thing to me, as a semiconductor analyst is these kind of CPU opportunities where we've seen the stacking and the benefits of it. But you also talked about areas like antenna and package, those kinds of things. How pervasive do you see that? And how -- do we see areas like smartphones and analog area is becoming a bigger driver for you going forward in these markets?

Joe, thanks for the question. Actually, it gives me the possibility to talk about another driver that is 5G. And you mentioned antenna package with the 5G millimeter wave, now a fully integrated RF front-end module requires also the antenna to be integrated with, of course, RF front end and the switchers and transceivers and the power amplifiers and so on, is packaging a single package generally done in SIP. The reason why you do that is twofold. First of all, signal integrity and second is form factor. And the package is becoming more and more critical. And in fact, we see an increasing demand for our ICOS advanced technology because whenever you want to expect the measure, these are very complex SIP packages with all these functionality integrated, you need a more advanced technology. This is what we see our component inspection growing quite a bit this year. And on top of that, in order to make sure that you meet the signaling integrity impedance requirements, you need to go to a next level of PCB board quality. So we see a completely different aspects of our business, taking advantage of this antenna packaging, a 5G millimeter wave either in infrastructure or mobile that is more in the component inspection and PCB. So again, the fact that KLA is so diversified now with EPC that we play in multiple markets, give us the possibility to benefit and leverage all these drivers are happening in the electronic system.

Great. Well, with that, we're out of time, but this is a really interesting look at a new growth area within semiconductor. So I really appreciate your time today.

Thank you, Joe. I appreciate it. Thank you. Thank you all. Goodbye.