Good day, ladies and gentlemen, and welcome to the Amkor Technology Second Quarter 2025 Earnings Conference Call. My name is Diego, and I will be your conference facilitator today. [Operator Instructions] As a reminder, this conference is being recorded. I would now like to turn the call over to Jennifer Jue, Head of Investor Relations. Ms. Jue, please go ahead.

Good afternoon, and welcome to Amkor's Second Quarter 2025 Earnings Conference Call. Joining me today are CEO, Giel Rutten; and CFO, Megan Faust. Our earnings press release was filed with the SEC this afternoon and is available on the Investor Relations page of our website, along with the presentation slides that accompany today's call. During this presentation, we will use non-GAAP financial measures, and you can find the reconciliation to the comparable GAAP financial measures in the slides. We will make forward-looking statements today based on our current beliefs, assumptions and expectations. Such statements are subject to risks and uncertainties that may cause actual results to differ materially. Please refer to our press release and SEC filings for a discussion on the risk factors and uncertainties that may affect our future results. We assume no obligation to update any forward- looking statements to reflect events or circumstances occurring after the date of this presentation, except as may be required by applicable law. With that, I will now turn the call over to Giel.

Thank you, Jennifer. Good afternoon, everyone, and thank you for joining the call today. Amkor delivered second quarter revenue of $1.51 billion, an increase of 14% sequentially, well above expectations and with all end markets showing double-digit sequential growth. I'm pleased with the way the team adapted quickly to support customers amid complex global conditions. They showed the agility and flexibility by accelerating product transfers and volume increases, delivering strong revenue growth. At the same time, the environment in which we operate remains dynamic. We continue to monitor export controls and trade policies working closely with customers and suppliers to address potential impacts. Our factory operations, which are largely situated in the free trade zones remain resilient while supporting dynamic supply needs. Within our end markets, communications grew 15% sequentially driven by the iOS ecosystem. Android revenue remained flat sequentially and increased 7% year-on-year. We expect a strong third quarter, driven by the launch of the next-generation premium tier smartphones. As AI expands into edge devices, we are working closely with lead customers to assess future device needs and develop advanced packaging solutions to support these innovations. In computing, revenue increased 16% from the first quarter driven by new product ramps in personal computing as well as growth in memory. Looking ahead to the third quarter, we expect sequential revenue growth across data center, infrastructure and personal computing. Demand for AI and high-performance computing applications continues to expand, and our project pipeline remains robust. Export control dynamics led to several supply challenges in the quarter. However, by working closely with our customers, we were able to rapidly respond to the changing demand environment. A major milestone in the quarter was the launch of the first High- Density Fan-Out product in high-volume production for our lead customer. Besides our proven 2.5D…

Thank you, Giel, and good afternoon, everyone. Second quarter revenue was $1.51 billion, exceeding the high end of our guidance range. We achieved strong sequential revenue growth of 14% and year-on-year growth of 3%. Our team excelled achieving strong revenue growth by supporting our customers while navigating a dynamic environment with agility, delivering on critical milestones, and executing focused preparations for a robust ramp in the third quarter. Second quarter gross profit was $182 million, and gross margin was 12%. This includes approximately $25 million in preparation costs for both the robust seasonal increase in Q3, supporting advanced SiP for communications, as well as launching multiple new High-Density Fan-Out products ramping in the coming quarters. We are pleased with the successful ramp of high-volume manufacturing at our Vietnam facility. We are running advanced SiP, supporting the consumer and communications end markets as well as NAND memory products. While we continue to build scale, Vietnam will impact our gross margin. In Q2, the impact was approximately 125 basis points, and we expect this will improve in the second half of 2025 as we optimize utilization. Q2 gross margin was constrained by foreign currency headwinds of approximately 80 basis points as compared to Q1. Second quarter operating income was $92 million and operating income margin was 6.1%, which included a nonroutine $32 million benefit due to a contingent payment related to our 2017 NANIUM acquisition. Net income was $54 million, and EPS was $0.22, including $16 million and $0.07, respectively, attributable to the contingent payment. Second quarter EBITDA was $259 million, including $32 million attributable to the contingent payment. EBITDA margin was 17.1%. We ended the quarter with a stronger balance sheet and greater financial flexibility. Amkor is focused on creating long-term shareholder value through a balanced and disciplined capital allocation strategy…

[Operator Instructions] And our first question comes from Joe Moore with Morgan Stanley.

I wonder if you could talk to -- I'll just ask both of my questions at once. The gross margin in the September quarter, I guess there's a little less fall-through than I might have thought, given the strong revenue. And you mentioned the mix shift. Can you just give us a little bit more color on the short-term gross margin. And then long term, you mentioned some consolidation with more legacy facilities. Can you talk about where gross margins can get to as you kind of resolve some of those issues?

Let me start with some introductory remarks, and then Megan can give more detail on the specifics. Well, first of all, second quarter, third quarter, a few elements impacted our margin. These elements relate very much through a transition in our manufacturing base. Let me start with Vietnam first and reiterate the strategic rationale on Vietnam, where we offer our customers an alternative to a China supply chain. And on top of that, Vietnam is a cost-effective location for SiP that we transfer out of Korea. Now it's important location, and we accelerated the transition and the ramp-up in Vietnam. Originally, we started with two lead customers in Vietnam, but currently already 5 of our top 10 customers are in Vietnam, and we are qualifying these customers. That led some increased cost in the quarter because our initial ramp-up is focused very much on execution and operational excellence and not per se on cost efficiency. We believe that Vietnam long term is a very cost-efficient location but in this stage, I think we're basically focusing very much on operational performance. Secondly, Korea. Korea we're also a bit in the transition. I already mentioned that we're offloading some of the SiP business in Vietnam, and we're doing that successfully. And next to that, we are ramping up our advanced products in Korea. Korea is our Advanced Products Center of Excellence. We have our R&D location in Korea and the new ramp-up of, specifically, High-Density Fan-Out products in Korea increased our cost base to some extent because we had start-up costs for these products. Now next to that, Megan already indicated, Japan. We have underutilization in Japan that takes longer than we anticipated, and that leads to reevaluating that manufacturing base. Next to that, we had some elements and Megan will talk to that with respect to exchange rates, et cetera. So these were, let's say, the top level topics. Megan, can you take it over to give some of the details.

Sure, Giel. Joe, just coming back to the gross margin for Q3. You mentioned the flow-through. That flow-through is actually a little bit better than last year's Q3, but the story is really the same. It's an unfavorable product mix in Q3. And that's primarily related to a very high concentration of advanced SiP, supporting the seasonal communications ramp, but also the continuing underutilization of our mainstream business. So while we are seeing that stabilize that divergent landscape is really accentuating that unfavorable product mix to be quite similar to last year's Q3. However, there's good leverage in that financial model. So while that gross margin flow through is constrained on that 27% increase in revenue, you'll note that the operating income and EPS are going to more than double on the bottom line. So as it relates to your second part of your question, more on the long-term outlook for margins, including Japan, Giel went through our strategic priorities and how those are playing out through this transition. And we see those strategic investments, not only in Vietnam and as well as bringing up High-Density Fan-Out that those are actually enhancing our competitive advantage and our long- term growth potential. So bottom line, our efforts are really paying off with key wins and lead customers and the disciplined execution is translating to very strong growth. So our profitability is impacted during this transition, and we will improve as we optimize over the midterm. As it relates to Japan, we can share more details at a later time once we have the specifics. I did want to remind you though that we had executed on a restructuring activity in Japan several years ago during the prior downturn. And over a multiyear period, we were able to successfully close three locations and consolidate business reducing head count and manufacturing costs. So we are working on the plans to rationalize that footprint, and we'll be identifying our primary sites such as Kumamoto and designing a consolidation plan. In addition, we're also looking to raise prices for those customers that might be running very low volumes in order to manage profitability.

And your next question comes from Jim Schneider with Goldman Sachs.

I was wondering if you could maybe comment on sort of the ramp that you expect for your 2.5D program, say, in terms of -- when volume terms this year? And then if you look a little bit further out, what kind of capacity are you directionally preparing for next year?

Yes, let me give some color on the 2.5D programs. I mean 2.5D is an important program for us. And I think we're working with our lead customer there. It was impacted in the earlier part of the year because of some trade restrictions. Trade restrictions are dynamic and are changing, and that will give us the opportunity to further ramp 2.5D. We are an entrance part of that supply chain and we believe that the direction currently being taken of having more open, let's say, trade approach will create increased opportunities not only for Amkor, but for the broader compute ecosystem. So 2.5D is important. Next-generation technology is High-Density Fan-Out. We consider that a critical technology. We are proud that we have launched the first products this year. And in the course of the second half of this year, we launched a second product with multiple products to come in the early part of next year. It's a foundational technology, not only in compute but it will also be utilized in next-generation communication devices. So an important milestone, and this investment will have significant spin-off going forward.

And your next question comes from Peter Peng with JPMorgan Chase & Company.

It seems like the quarter went really well. For the third quarter, you're guiding above seasonality. Wondering if you're seeing -- or any of your customers talking about any pull-in of demand or anything of that or is it business as usual? Maybe if you can provide some comment on that.

Peter, let me give you a few comments on, let's say, potential strength in the third quarter and in the second quarter related to pull-ins. Now it's difficult for us to see whether customers are really pulling in, but we didn't get signs that there was active pull-in ongoing. There may have been some pull-ins related to inventory that's available or dies that are available, but we didn't see active pull-ins. What we're currently seeing across the market is some last minute spot orders that come in because inventory reaches a critical low point, and that happens, for example, in the automotive market, that's the first indication that inventory comes close to the equilibrium there, which is a good sign. You could label that pull-in in the communication market. Last year, we saw some pull-ins in the second quarter, but we don't see that occurring this year.

Perfect. And then I have a follow-up on just your overall computing. It sounds like things are getting better. You mentioned about the easing of some of the regulations at your marquee customer, and then you're also ramping your second 2.5D networking customer. So how should we think about revenue growth this year relative to 2024? Is this -- do you view this as a growth segment? Or is this kind of a muted growth profile?

Yes, the way that we see it, and I see it is that the compute market, there are a couple of trends in the compute market. First, we see accelerated growth and also the rate of innovation is increasing. That means that changing from one generation products to the next- generation product is accelerating. And that also implies that new technology is adopted faster. So 2.5D is, I would say, previous generation products. The lifetime of this previous generation is extended because of the export controls on technologies, and that could give 2.5D a fairly long life with exports and the new export rulings and we look forward to support that. Next-generation devices, we're participating in a -- in these areas in multiple ways. I think one is High-Density Fan-Out. That's a preferred technology versus the 2.5D that users interposer, High-Density Fan-Out users RDL, high-density interconnect, and that's the more flexible more cost-effective technologies. We expect the lifetime of that technology to be longer, and therefore, launching the first products is very important. And then, of course, we see on the high bandwidth memory integration, multiple technology steps. So we are participating with our lead customers but we see continued innovation going forward.

And Peter, just to add a quantitative reference as well. We did hit a record in 2024 for our broader compute segment and we're on track to do that again in 2025. The first half of 2025 compute was up 18% year-over-year, and it is our fastest-growing end market.

Your next question comes from Ben Reitzes with Melius Research.

Can we just talk a little bit more about your comps guidance versus normal seasonality in the 3Q? And then if you think that, that's sustainable into 4Q, and what is it about -- and can you just confirm that you rewon all your business on the iOS side that was previously an issue?

Ben, yes, let me start with answering that last part. The footprint in next-generation phones and the recovery of that socket is going as planned and the ramp-up of that socket is included in our third quarter outlook and will be included in our fourth quarter outlook. So in that sense, we are optimistic. Understanding what's happening in the fourth quarter is difficult. There's still quite significant uncertainty certainly in the comps market, what the volumes will be. We see conflicting forecast for volumes in the later part of the year. Third quarter is confirmed. I believe that we cannot use normal seasonality, which would be a few percent up or a few percent down of third quarter because of these uncertainties. So we don't guide for the fourth quarter then.

And Ben, just to give you a little more color in the markets in Q3. So the communications [done] itself will be up very strong, probably more similar to the 2023 Q3 comps on 2022 comps sequential increases. And what's offsetting that is our consumer will be mainly flat in Q3 this year because we had that very significant ramp last year Q3.

Okay. And then -- and Megan, can we just confirm the payment on the SG&A and what the normalized SG&A is going forward? $32 million, we just simply add that back and then expect that kind of run rate for the next few quarters?

Yes, sure. So if you were to exclude that, we were right in line with guidance with respect to OpEx and that would be a similar run rate for the second half of the year.

And your next question comes from Randy Abrams with UBS.

Yes. I wanted to ask a follow-up question. Just on the communications view. I think in your prepared remarks, you talked about evaluating content gains, say, for the on-device AI to communications. Could you talk about that technology trend you're seeing there? And I think you mentioned potential on Fan-Out for that application. Also, there's a discussion about multichip module. So if you could go to on some of the midterm content opportunities from that.

Yes, that's a good point, Randy. I didn't call out specifics on technology choices. What we see customers doing with AI functionality moving to the edge, they are looking to the implications of the semiconductor solutions, what does it mean related to increased memory content, Higher-Density interconnect, physical dimensions and there, we work with multiple customers that operate in that market on next-generation solutions. That's also why I reiterate that a Fan-Out solution, certainly for communication and a High-Density Fan-Out solution, similar to what we apply in compute could be one of the solutions for communication also. Nothing definitely decided, but for Amkor, it's a very important market, and we believe we have a very close cooperation with lead customers in that market to evaluate future solutions.

Yes. And my second question actually is two separate financial questions. One on the material costs. Just curious to what you're seeing in terms of cost from some rising material prices, and there's even a bit of talk about the high end, things like [T-Glass]. If there's any constraint you're nervous about on high-end substrates, that's one part. And the second one, it's more tied to the investment. If we could be heading into a little bit bigger CapEx cycle, just -- I think you've noted advanced capacity tight and trying to go after RDL puts to start construction of the Arizona factory.

Okay. Well, let me take the first part of the question, and then definitely Megan can do the second part of the question. With respect to material cost structures and material prices in the market, certainly when it comes to substrates. I mean we don't observe it yet. But on the other hand, we see some capacity constraints potentially happening because of the pull from the compute market on the high-end substrates. So we proactively work with the critical suppliers to secure our capacity needs and to pre-discussed the pricing arrangements, that's ongoing. We put in place a strategic procurement team specifically for this area because we see indeed that there's a potential for constraints. So far, not really some first indications that it's coming up. And of course, that will potentially lead to increased pricing.

And then, Randy, just to pick up on the comment about the significant investment in this particular area. It is quite fungible across the different landscapes of 2.5D as well as High-Density Fan-Out and Flip Chip BGA. So all of those investments where we had really started to ramp up in '24, continuing into '25 and we are planning to continue to invest in this area in '26 are fungible. As it relates to the broader CapEx cycle, including investments here, but also facilities we'll have a more specific outlook on our '26 CapEx at our year-end call. Just as a reminder, with respect to the U.S., we are planning to break ground in the second half of '25. We've reserved around 5% to 10% of our current CapEx guide of $850 million to address our investment in Arizona here in 2025. We do have significant grants and investment tax credits associated with that. However, as you may know, those will be a bit on a lag. So there may be some increased CapEx in 2026 and 2027 ahead of when those grants and incentives are received.

And your next question comes from Tom Diffely with D.A. Davidson.

Just curious if your view -- your long-term view of the auto industrial market has changed with keeping track of the potential closures you're talking about in Japan?

I mean our long-term view has not changed. To give you a couple of critical elements, we see that in the automotive market, although there were a few difficult years, certainly, when it comes to volume and inventory consumption, the trend clearly is towards advanced packaging and automotive where we are well positioned with respect to next-generation ADAS devices and that was also the highlight in last quarter and will be highlighted next quarter also because that is delivering the growth. The areas where we see consolidation is on the more mature part, mature microcontrollers and the traditional analog products going into automotive. The Japan market in itself on the automotive side is going through a difficult period. And that means that the supply chain feeding into that market is also going through a prolonged critical period. Our view basically has not changed on a global basis, but Japan has a slightly difficult or different local dynamics and I think we need to respond to that because we have a significant manufacturing base in Japan. Megan already highlighted what we did in the past, we rightsized the operation in 2019 in Phase 1, and we will go through the next phase in this year and early next year.

So do you comments on the softer mainstream side? Also include power.

Well, softer to the previous very high expectations. Yes, I think silicon carbide is very much tuned towards the EV market, and there were high expectations for ramp in the EV market that is moderating a bit. And in that sense, the silicon carbide market is not as fast- growing as expected. However, the fundamentals are still there. I think that technology together with gallium nitride are the foundation not only for electrical vehicles, but also for the overall energy transition and we see a pickup even into data centers for these technologies. We are well positioned there. And as you know, we're the leader in this market, we have a joint investment in Portugal for advanced their power modules that facility is ready. We are now facilitating that and we're going to ramp that into the next 2 to 3 years. So the basics are still there. The scale and the timing may be slightly delayed versus originally -- original high expectations.

And your next question comes from Ross Cole with Needham & Company.

I was wondering since the U.S. has lifted the restrictions on age 20, have you gotten any signal from NVIDIA to resume production there?

Well, I mean we don't call out details on specific programs for this. In general, I commented already that we believe that this decision was a good decision. It will create more opportunities that will create a bigger -- several markets, not only for companies like Amkor because we are in that supply chain but certainly also for the end customers that you just mentioned. So we are an integral part of that supply chain, and we don't comment to individual customers' forecast changes.

Great. And then my second question, have you received any additional clarity on TSMC's Arizona plan for advanced packaging? And do you expect Amkor to play a role in completing the TSMC supply chain in Arizona?

Yes. I think we are in constant communication with TSMC. We consider TSMC a very important partner, and we also announced an MOU we did that last year on technology sharing agreement for Arizona specifically. Advanced packaging has multiple dimensions. If you move into 2-nanometer technology, that also will enter Arizona a little bit later than expected. But that would mean that you need technology that's foundry specific, and that's also labeled advanced packaging. So our approach is unchanged. We're working with customers as well as with foundry partners to align on our road map, and we are making good progress there. Again, our strategy is to be complementary to TSMC, we're not competing for design slots with TSMC.

And your next question comes from Steve Barger with KeyBanc Capital Markets.

I just wanted to go back to the auto and industrial. And listening to some of the analog companies that have reported, it's kind of a mixed bag on whether inflation is here or when that could happen. So can you just talk about how your customer conversations changed for the back half? And is anyone talking differently about capacity requirements in a way that seems much more promising than it has been?

Good question, Steve. I mean, we see in automotive two sides. One is the advanced packaging side with applications like infotainment, ADAS, but also high-end sensor technologies. There are in high demand because the proliferation of these technologies into the car ranges is actually accelerating. I already made some comments with respect to silicon carbide. We believe that long term, that market is still a very strong market, we are investing in it, we're partnering with leaders in that industry. On the more mature parts, I think you refer to the more analog players. If we go through the customer inputs that we're getting, I would say the -- overall, we believe that the supply chain and the inventory levels, both in distribution as well as at the OEMs are getting more in balance. There are here some pockets of early indications of some shortages, and there are still indications of some high inventory levels. So what we see from customers is they are actually more optimistic going forward, that is also reflected by more short-term orders in factories for short-term delivery. And that's how we see it. We definitely don't see a [V curve] where you see a rapid acceleration in the analog domain or in the more traditional microcontroller domain. So stabilization and slow improvement and what we see for the second half of this year, there will be low single-digit growth in the automotive market and that part...

Yes, great detail. And then on the tester fleet investments, how long will that take to accomplish the upgrades that you need? And are those upgrades to existing units for those increased requirements? Or is that investment really for new equipment?

Well, it's a combination of both. I mean we have a very advanced tester fleet already. The biggest advanced tester fleet is in Korea, but we also have a fairly sizable tester fleet in Japan, a more mature tester fleet in -- sorry, in Taiwan and a more mature tester fleet in Japan and China. So there are upgrades of existing platforms, and there are new testers coming in, depends on the application, whether it's for compute or for communications. Now as you know, there are basically a limited number of tester supplies worldwide. The tester platforms are very well defined, and we're working closely both with customers as well as with suppliers to have a part on upgrading or renewing our tester fleets. When did it start? It already started, I would say, with the last introduction of our latest let's say, communication processors and computing process. So it's ongoing. As we mentioned, the first phase of expansion in Korea will end this year, we're adding another 4,000 to 6,000 square meter of test floor. And then we add another 40% in 2026, ready for equipment move in end of 2026. So it's a significant expansion, which fits very well in our strategy of offering a full Turnkey support to customers and that's needed not only for the compute domain, but also for the communication demand. One thing to mention is that the upgrades that we're making now and the engagements with lead customers certainly in compute, we will mimic these capabilities in our U.S. facility to offer the same or similar Turnkey solutions in the U.S.

And at this time, I'm showing no further questions. I would like to turn the call back over to Giel for closing remarks.

Thank you. Let me recap the key messages. Amkor delivered second quarter revenue of $1.51 billion, an increase of 14% sequentially and 3% year-on-year. Third quarter revenue is expected to grow 27% sequentially at the midpoint of guidance to $1.925 billion (sic) [$1.975 billion] driven by the seasonal increase in communications and continued momentum in high performance and AI applications. We are focused on enhancing operational efficiency, including ramping production in Vietnam and optimizing our mainstream manufacturing footprint in Japan. With a clear focus on our strategic pillars and a strong balance sheet, we are well positioned to deliver sustainable growth and long-term value creation. Thank you for joining the call today.

Thank you. Ladies and gentlemen, this concludes today's conference call. You may now disconnect.