Okay, I think we're ready to start. Good morning. My name is Giel Rutten. I'm the CEO of Amkor Technology, and what we'll do this morning, I'll give a short introduction of Amkor, and after that, we do a couple of questions. Okay? Okay, let me give it a start here. Now, Amkor is a pioneer or was a pioneer in the outsourced manufacturing industry. It was established 55 years ago, so early on in the semiconductor industry, and it's, it established over the last 55 years a deep relationship with all leading customers in the industry, both fabless companies as well as the IDM companies in the semiconductor industry.
We build a trusted cooperation with most of these companies, and we serve them from a very diverse manufacturing base around the world, where we have about 2-2.5 million sq ft of manufacturing space in six countries around the world. That global presence is important to offer our customers a resilient manufacturing supply chain. Now, we invest in the future. We are a leading company in advanced packaging. We are considered a technology leader. We're investing in both R&D as well as manufacturing capability and scale. In 2022, we invested close to $1 billion in that. So we have a large R&D base, about 700 engineers in Korea, where we support and work closely together with our lead customers.
Overall, we employ about 30,000 people around the world, and we hold a couple of important leadership positions. I mean, we are number one OSAT in the automotive industry, but besides that, we also hold very strong positions in the premium-tier smartphone industry, as well as in the high-performance computing market, where we offer a complete package portfolio. In 2022, we did $7.1 billion. This year, we are in a declining industry. We declined about 9% year-on-year in an industry that is declined between 15% and 20%. So overall, we're doing better than the industry.
We are an advanced packaging leader, so we're not engaging very much in the, let's say, commodity side of the technology, but we're focusing our our support and services to advanced packaging, and that starts with everything that relates to bumping, flip chip, up to 2.5D chip package technologies, where we offer a complete portfolio to our customers. So we're not on the wire bond side, too much of our business, although that's still part of our portfolio, and we see that as a complementary part to support our customers. Now, where are we in the supply chain? I think we are an integral part of a semiconductor supply chain. On the left-hand side of this slide, you'll see, the, let's say, the design companies.
These could be fabless companies or IDMs, and more and more, we see the OEMs also, let's say, building up design capabilities to design their own, their own products to protect their IP. So then, of course, we have wafer foundries or wafer fabs that companies maintain internally, and then, you know, wafers move into an OSAT space, and some companies do their assembly and test internally. We ship directly, of course, to the contract manufacturers. There are a couple of trends in this, in this supply chain. First of all, I think the, the percentage of outsourced manufacturing is increasing. Certainly for advanced packaging, we see that companies that's, that currently have in-house manufacturing capabilities, when their, their products move to more advanced silicon or to advanced packaging, that they tend to outsource more and not invest internally.
Secondly, we see that a lot of the industry mega trends in the semiconductor industry are supported by advanced silicon as well as advanced packaging. So the growth in the industry is almost exclusively supported with advanced products. And thirdly, I think the geopolitical tensions currently, and also the uncertainties in the supply chain due to COVID, actually result that location and geography becomes more important. Customers are willing to take less risk to ship products around the world, to have them assembled in different locations. I think co-location gets more important, and also customers look for a stable, sustainable supply chain that can maintain supply for multiple years to go. In the automotive industry, for example, you need to support customers for at least 10-15 years with a stable supply chain.
So there are two important interfaces where we work closely together with the partners in the supply chain. One, of course, is on the foundry side. I think we consider companies like TSMC, GlobalFoundries, and others as our partners. And then, of course, it is also to the contract manufacturers, where we offer a seamless drop shipping service, and where needed, we look to modular solutions together. Now, there are three important pillars for Amkor, where we base our strategy upon. You know, it starts with, on the left-hand side here, with our technology leadership position. I think customers knock on our door because they consider us a technology leader, and that's also where we invested in the last 55 years, and we continue to invest in going forward.
Now, on the technology leadership side, I think we built, you know, unique positions, as I already mentioned, in automotive, but also in the premium-tier smartphones, where we co-work very closely with customers there, but also in the high-performance computing area. The innovation model is very much a model where we work with our customers in a very early stage of product development, so it's much more a co-development with customers and not an invention-based innovation. So, co-investment with customers in innovation, in R&D, is important for us in order to bring technology quickly to the market. And then, of course, after that, operational excellence makes the big, the big differentiation. Geographic footprint, I already mentioned that. I think that's important, and it's becoming even more important.
You know, we invested in significant manufacturing capabilities in Asia, and recently, in October, we opened our new facility in Vietnam to offer customers an alternative to an established supply chain. For example, in China, we see that there is a de-risking ongoing, and I think there, Vietnam is definitely emerging as an ecosystem, and that helps us to be there. I think we envision that our Vietnam facility is developing or will develop as a manufacturing campus, and it's actually a significant expansion for Amkor going forward. Of course, recently we announced our manufacturing site in the U.S., where we believe that reshoring of manufacturing capabilities to the U.S. will continue to happen.
It will take time to build up that capability, but for us, this was the right timing to engage both on the customer side as well as on the foundry side, to offer that seamless supply chain there. But also we see a similar thing happening in Europe, specifically supported by the European automotive industry here, where we have a manufacturing location in Portugal, and that's very rapidly expanding based on demand coming from the automotive industry here in Europe. Now, then, the last thing is, alignment with industry mega trends. You know, we are a leader in advanced packaging, and we believe that the mega trend, trends in the industry, be it, in communication, moving from 5G to 6G, be it in automotive, high-performance computing or IoT, is all supported exclusively by advanced packaging and advanced silicon.
I think that's the way forward, and there, our engagement model to work with the leaders in these industries is the way that we drive this, the business growth in line with these market engagements. Now, our performance, I think we grew above the industry average over the last couple of years, both on revenue side, we did about $7 billion in 2022, as well as on the profitability side. I would think our EPS, our earnings per share, actually is up sixfold over the last five years. And also our EBITDA, which is a reflection of cash flow, is up significantly. We're able, even through last year to this year downturn, to maintain a positive free cash flow as well as, let's say, overall earnings performance through the cycle.
Now, investment appeal, you know, we believe, and we're convinced that we can outgrow the industry, based on our position in advanced packaging, as well as our geographic diverse footprint. That's an appeal to our customers. Now, we are heavily leveraged to the industry mega trends. We're not on the commodity side of the manufacturing services. We're actually leveraged to the advanced packaging side in the industry mega trends. And that's a short introduction for Amkor.
Great. Well, thank you for that.
Thank you.
Maybe we can go into some questions and really focus on two of the things that you talked about. One, the kind of reshoring phenomenon, and then second, the advanced packaging. So in terms of reshoring, you know, maybe if you could start with, you know, where are governments in thinking about this? We saw, recently, $3 billion of the CHIPS Act money has been allocated to advanced packaging. You look at this, and a lot of the shortage that we saw in the last two or three years was back and related, right? We saw that, most notably, in some of the cutting-edge markets, but really across all the markets.
So, you know, where do you think we are in terms of recognizing that priority, and how important is it that the U.S. government, the European governments, that the Asian governments are starting to fund the development of the localized packaging capability?
Yeah. Yeah, that's a good question. I think in the U.S., we definitely see that there is an awareness that advanced packaging is an integral part of semiconductor supply chain and requires investment. And, you know, there's definitely interest from government, and we also believe that there will be support in order to establish that basis in the U.S. There's even belief that in the next wave of ramp in the industry, there may be a shortage of advanced packaging because the industry overall is underinvesting in this area. So the awareness is there. There is investment.
The $3 billion that you referred to is actually more on the R&D side to develop a strong basis in the U.S. for advanced packaging R&D, but also on the manufacturing side, we applied for CHIPS Act support, and we believe we are well-positioned to find that support since we're, you know, we are supported very well with partners in the supply chain. In Europe, I think in our factory in Portugal, you know, we see that, with the European support, it is going well at another level than in the U.S., but definitely we see support also from customers.
I think the business model that we work with in Europe is different than in the U.S. because in Europe we see, we go very much in the co-investment model, where customers invest in capabilities and capacities together with Amkor. So overall, I believe that there is a strong awareness.
The very recent decision to build the facility in Arizona, can you talk about what drove you to do that? Obviously, there's a lot of activity there with TSMC building a factory. Was that part of the decision? And, you know, what do you think the economics of building, of doing packaging in Arizona versus, you know, more traditionally in Asia, how that's going to be?
Well, let's start with that last part of the question. I think we believe the economics, and I think we put a lot of work in, in that. I think there are pros and cons. In order to make it economically feasible, it's important to choose the right technology domains that you want to cover. You don't want to support commoditized manufacturing services out of the U.S. You really want to focus to the higher-end, differentiating technologies, where customers are willing to pay a premium for that technology. You also need to build a factory that is, to a large extent, automated. I think we have a fairly high automation level in our Korean factories, and we're going to use that also in a U.S. facility.
And then, of course, you know, the longer-term commitment from customers and the support from government is important. We believe the support from government is there, both state as well as federal government in the U.S. And, within that overall framework, we believe we can build a sustainable manufacturing company in the U.S.
Well, I mean, one of the distinguishing characteristics of this period is that the OEMs are suddenly very interested in where their semiconductors are coming from, and, you know, we've written about this General Motors deal with GlobalFoundries, but it's kind of the same phenomenon of OEMs who probably hadn't heard of GlobalFoundries, haven't heard of Amkor three years ago, are now very focused on making sure that they, they can, they can ensure their supply chains will be, will be, continuing to support them, and they might even be willing to pay more for that to happen. Can you talk about that phenomenon and how has that engagement gone with, with OEM customers-
Yeah
... from your perspective?
Yeah, that's, that's a good point. I mean, we see, did see that OEMs get more involved in their supply chain, and that was based on their experience during the COVID time. You know, take the example of automotive. You know, once these big automotive factories are not able to run because their lack of semiconductor components, I think that highlights a sense of urgency. So we see direct involvement now, but even by policies. I mean, if you take, for example, the U.S. automotive manufacturing companies, they are basically, by policies, directing their supply chain or their procurement departments to procure local components or to have local-manufactured components. And, I think that offers an opportunity, and they also are directly involved with manufacturing companies like Amkor.
So in that sense, you know, there is an awareness at that level, not only in the automotive companies. I think, you know, reshoring manufacturing into the U.S. is also a matter of national security for very critical technologies. Moving into industries like AI, you want to have, you know, besides even national defense, you want, you want to have manufacturing capabilities within your own borders.
You might get some nice pull-through from the OEMs from these commitments.
We believe that will help in the different industries. The model is different. In the automotive industry, it's different than in the high-performance computing or mobile industry, but we believe that end-to-end involvement is there.
Yeah, and that's a good segue to the other questions on the advanced packaging side. When you look at, you know, the various forms of stacked silicon interconnect, Chip on Wafer on Substrate, things like that, it's the technology starts to look a little bit more like wafer fab equipment-oriented technologies. You know, does it put you into competition with TSMC? And obviously, you've been a big part of ramping some of those technologies at a time when the industry dearly needs it. How does that affect your business?
Yeah, You know, does it put us in direct competition with TSMC? I don't believe so. The way that I see TSMC or that we see TSMC is more as a partner, more than as a competitor. Of course, there are some technology areas where we overlap, but the traditional model is that on the foundry side, there is an initial engagement on assembly just to make sure that introduction of advanced silicon is enabled. Once that technology matures and is proliferated further into the different applications, I think that goes into a foundry environment. So, you know, for technologies like 2.5D, or Micro- Bumping, or Wafer Level Fan-Out, I see that gradually moving into the OSAT domain. For 2.5D, we are in our third generation of technology.
We started in 2017. I think we are, we are engaged with all the customers that need that technology currently in the industry. We have to keep in mind that the decision on the supply chain model is with the end customer. It's not with Amkor, or it's not with TSMC. The end customer decides what's the adequate supply chain model. In some cases, they will choose for the vertically integrated one-stop shop, but in many cases, they, they look for a diversified supply chain.
And these technologies, are they more capital-intensive than what you've done in the past? How do you have to think about your capital spending with regards to that?
Well, if we take our... You know, we have a broad portfolio of advanced packaging from, you know, advanced to flip chip, to stacked die, let's say, configuration. Now, if we take the capital intensity for technology like 2.5D, that is in line with our traditional bumping business. We have a large install base of bump capacity. A lot of that capacity and that equipment is what we call fungible, interchangeable within, you know, multiple applications. So we don't install a unique 2.5D manufacturing line. I think that manufacturing equipment is fungible across multiple applications, and as such, in that sense, we can warrant a high degree of utilization, and that, of course, gives us a much better return on these investments.
Great. And, you know, I guess the similar question, as I talked about OEMs pulling through, the awareness of advanced packaging at the level of semiconductor CEOs is higher than I've ever seen. You know, we hear-
Mm-hmm.
When AMD bought Xilinx, they sort of talked about the commonality of their advanced packaging roadmaps and things like that, which I wouldn't really have heard ten years ago. I mean, obviously becoming much more important. You know, how is Amkor positioned to take advantage of and leverage those kinds of opportunities?
Yeah. Amkor started to invest in advanced packaging as a first mover in the industry. We built this position, and currently, 75% of our, let's say, revenue is generated from advanced packaging, and of course, that's a broad portfolio. You know, in that sense, for many of the companies that you mentioned, we act as a one-stop shop. We can offer them solutions for more their, their, let's say, mid to, or mid-range products as well, as their high-end products. So, I feel that, and we believe that Amkor is very well positioned to benefit from this, this need for more advanced packaging, and we also see that. I think if we look to the engagement model, it's also different. I think we're becoming much more a co-development partner for these companies in an early stage of their product development.
Great. And last one more question, and I'll open to the audience. You know, you talked about the cycle environment. You talked about being down, you know, high single- digits this year, potentially. Where do you think we are in terms of clearing out the inventory that's there? And, you know, obviously, that's can be difficult to assess, but does that look different from your standpoint than when we think about the device companies? Are there other layers of inventory to consider around that?
Well, what we observe and what we believe there are, you know, there's still some concern on inventory. You know, if you take, for example, the automotive industry, they were hit very hard in the shortages of supply during the COVID time. They built up some inventory. They now see that the supply gets more normalized, so they go to a more traditional supply chain model, and that means they will reduce their inventory a little bit. Of course, that trickles down the supply chain, but I wouldn't say that that's a structural issue with end market demand. Overall, on the mobile phone or the smartphone industry, you know, specifically on the Android side of that industry, we see some first green shoots of reordering of components.
You know, I cannot say that we're completely in a normalized supply chain there, but there are some first signs of recovery. It may take one quarter or so.
Mm-hmm.
So overall, I think we are at or close to the trough of this correction across different industries. So that should be a good basis to see a year-on-year at least stabilization or probably growth into 2024.
Great. Let me see if we have questions from the audience. If not, I can keep going. Anyone? All right. So you mentioned end markets. Can you talk to the automotive and industrial? Smartphone, you know, we've been correcting for now a couple of years. It does feel to us the same, that you've kind of bottomed out. When you look at industrial and automotive, it's a more recent phenomenon that industrial has started to slow, but we're looking at already for your customers, peak-to-trough declines that are worse than 2019, and really worse than any environment other than 2009 or 2001. So it doesn't seem like there's visibility yet that it improves, but it also seems like we're already discounting a fair amount of weakness.
Can you talk to your prospects in those markets?
... Yeah, we believe that, you know, in the automotive market, there are a couple of dynamics. You know, first of all, the semiconductor content per car is increasing, and we believe will continue to increase. And that goes, you know, across the traditional cars as well as the electrical vehicles. So functionality in some cases becomes mandatory because of governments' legislations, and that drives up semiconductor content per car. Of course, the volumes of cars may fluctuate a little bit, but that's fairly stable over the years.
So for us, if we look to the applications that enter into the car, like driver assistance features, digital dash, dashboards, in-car networking, telematics, you know, we believe that these functionalities will continue to go from the high-end cars into the mid-range and lower end of the cars, and that will increase the semiconductor content. And as I mentioned before, all these new semiconductor features are driven and supported by advanced packaging, whether it's an ADAS controller, which currently already moves into a 7-nm silicon technology and is packaged in a very advanced, either multi-chip modules. And so in that sense, that's good. And on top of that, I think most of these new components entering the automotive industry are actually out of an outsourced supply chain, so not from a vertically integrated supply chain.
Okay, great. Giel, I think we have just one more minute.
Sure.
Just, just as we move into 3D ICs, and we see the bump pitch changing get to sort of nanometer scale bump pitch. So I'm thinking through thermal compression and maybe hybrid as well. Does that present as an opportunity for you guys, or how do you think about that space when it comes?
Well, you know, the way that I see it, when it comes to hybrid bonding, it's very much an extension of front-end manufacturing, where, you know, in order to continue Moore's Law from a monolithic integration to a chiplet integration, you need technologies like hybrid bonding. To me, that's very much a front-end manufacturing process. If you really look to manufacturability, wafers cannot even leave a wafer fab, which is at sub-class one cleanliness, in order to achieve the yields that you need. So this is not, in my view, at least, it's not an OSAT domain. However, if these, let's say, multi-chiplet wafers leave the fab manufacturing environment, they still need to be packaged. They go on either an interposer structure with high-bandwidth memories, or you put them directly in a, in a, on a substrate.
I think that all is within our scope, and that's what we're preparing for. I think this requires a close cooperation between a foundry and an OSAT, and this requires pretty advanced technology. So in that sense, I see this as an opportunity. Also, die sizes continue to increase, thermal management increases, bondability is an important factor. So, you know, we are investing in technologies like thermal management, materials, laser-assisted bonding technologies, and I think just to prepare ourselves to take these hybrid bonded chiplets.
Great. Well, that puts us up against our allotted time. So Giel, thank you very much for your-
Okay.