AMKR 2024Q2

Good day, ladies and gentlemen, and welcome to the Amkor Technology Second Quarter 2024 Earnings Conference Call. My name is Diego, and I will be your conference facilitator today. Please note, we are having some system issues with our webcast and slide technology. So if you're connected via the webcast, you may not see the slide presentation, but you will hear the audio of this webcast via the webcast link. And a reminder at this time, all participants are in a listen-only mode. After the speakers' remarks, we will conduct a question-and-answer session. 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.

Thank you, operator. Good afternoon, everyone, and thank you for joining us for Amkor's Second Quarter 2024 Earnings Conference Call. Joining me today are Giel Rutten, our Chief Executive Officer; and Megan Faust, our Chief Financial Officer. 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 US GAAP equivalent on our website. We will make forward-looking statements about our expectations for Amkor's future performance based on the environment as we currently see it. Of course, actual results could differ. Please refer to our press release and SEC filings for information on risk factors, uncertainties, and exceptions that could cause actual results to differ materially from these expectations. Please note that the financial results discussed today are preliminary and final data will be included in our Form 10-Q. And now I'll turn the call over to Giel.

Thank you, Jennifer. Good afternoon, everyone, and thank you for joining the call today. Amkor delivered second quarter performance in line with expectations with revenue of $1.46 billion and EPS of $0.27. Total revenue increased 7% sequentially, driven by demand for advanced packaging, notably for premium tier smartphones and 2.5D technology for AI solutions. During the quarter, we maintained focus on our strategic pillars to elevate our leadership position. We successfully brought online additional capacity for 2.5D technology in Korea and qualified advanced SiP and memory technology in Vietnam to support production ramps in the third quarter. Additionally, we are excited that we have reached a significant milestone in establishing a US manufacturing presence for advanced packaging. We aligned on a nonbinding preliminary memorandum of terms with the US Department of Commerce for up to $400 million in grants under the CHIPS and Science Act. These funds will support building a new facility in Arizona, enabling advanced packaging and test for high-performance computing, AI, communications, and automotive markets. We look forward to being part of a strong ecosystem of front-end fabs, IDMs, and suppliers in establishing a resilient US semiconductor supply chain. Now let me review the current dynamics in each of our end markets. Revenue in the communications end market increased 10% sequentially. Within the iOS ecosystem, we experienced a larger than seasonal increase driven by bills for the full launch of premium tier smartphones. Revenue within the Android supply chain declined slightly sequentially, but still showed a strong 20% year-on-year growth. Our advanced packaging technology supports a wide range of applications and functionality throughout the phone. With our advanced SiP technologies for heterogeneous integration together with our proprietary flip chip package-on-package technology, we support the full range of applications from RF and camera to the latest AI-enabled apps processors that require high-speed and high-density interconnect with fine pitch bonding. Revenue from our automotive and industrial end market was down 2% sequentially. The coverage in this market is taking longer than anticipated due to weak demand and ongoing inventory corrections. Despite these near-term dynamics, we believe the long-term drivers for growth remain intact. Semiconductor content per car is expected to continue to increase, driven by the proliferation of ADAS, electrification, infotainment, and telematics all requiring advanced packaging technology. Amkor is the leading automotive OSAT and has multiple decades of experience meeting the stringent requirements of the automotive industry. With a broad portfolio of advanced and mainstream technologies and established large-scale manufacturing base in critical regions like Europe and Japan and trusted relationships with key customers in the automotive supply chain, Amkor is well positioned to support the secular growth in this market when it exits the current cycle. Revenue from the computing end market increased 20% sequentially, driven by strength in AI devices and several new product introductions for ARM-based PCs. We executed on our planned expansion for 2.5D capacity for AI devices more than tripling our capacity versus second quarter of 2023. In the third quarter, we expect constraints and high bandwidth memory supply to limit revenue growth. Amkor is leading the OSAT supply chain with the deployment of 2.5D technology. And with the robust demand and additional capacities, we now expect the full year 2.5D revenue to quadruple versus 2023 levels. We continue to partner with multiple customers on next-generation technology utilizing organic interposers and expect those solutions to be brought to market in the first half of 2025. Revenue from the consumer end market decreased 6% sequentially, driven by the wind down of legacy IoT devices ahead of the expected ramp-up of next-generation products. Traditional consumer product demand has been muted, but the high-volume production ramp of a new wearable product utilizing advanced SiP technology is expected to start in the third quarter. Consumer IoT devices require many authorizations with increasing levels of integration. Our advanced SiP technology for heterogeneous integration positions us well to meet these requirements and our new Vietnam facility enables us to continue to drive manufacturing scale and innovation. During the second quarter, our manufacturing organization had to manage multiple challenges. On one hand, several factories still showed low utilization where focus was on cost control while maintaining high-quality standards. On the other hand, we executed the capacity ramp for 2.5D technology in Korea, qualified advanced SiP and memory technology in Vietnam, and prepared for the steep seasonal ramp in the third quarter. Additionally, the team further progressed with our advanced packaging and test facility in Arizona by advancing factory design and construction planning and working with customers to develop the technology road map and capacity loading scenarios. During the second half of the year, cost and quality, together with managing a steep seasonal ramp, will remain top priorities. Now let me turn to our third quarter outlook. Considering current market conditions, we expect third quarter revenue of $1.835 billion at the midpoint of guidance. This represents sequential growth of 26%, driven by advanced packaging in support of the seasonal launch of premium tier smartphones, the ramp of a new consumer wearable device, and continued strong demand in high-performance computing and ARM-based PCs. The slower-than-expected recovery in the automotive and industrial markets, together with the continued weak demand in traditional data centers, has dampened anticipated growth in the third quarter. Looking forward, we remain confident that the secular growth drivers for the industry remain in place. With our strong technology leadership in advanced packaging, a uniquely diversified global footprint, and partnerships with lead customers, we are well positioned to accelerate while exiting this cycle.

Thank you, Giel, and good afternoon, everyone. Amkor delivered second quarter revenue of $1.46 billion. This was in line with guidance and represents a 7% sequential increase driven by strength in advanced SiP and 2.5D technology. Second quarter gross profit was $212 million, and gross margin was 14.5%. While gross margin declined modestly from Q1 due to higher material content, gross profit dollars expanded 5%. Operating expenses for the quarter came in lower than expected at $131 million. We expect Q2 to be the peak for operating expenses in 2024 due to preparation costs for our factory in Vietnam. We are on track to begin production in Vietnam in Q3, and we expect a portion of the cost to move to cost of goods sold when production begins. Operating income was $82 million and operating income margin was 5.6%. Net income for the second quarter was $67 million, resulting in EPS of $0.27. Second quarter EBITDA was $247 million and EBITDA margin was 16.9%. Our balance sheet remains strong. We ended the quarter with $1.5 billion of cash and short-term investments and total liquidity of $2.2 billion. Our total debt as of the end of the quarter is $1.1 billion and our debt-to-EBITDA ratio is one time. Moving on to our third quarter outlook. We expect Q3 revenue of around $1.835 billion, representing a significant sequential increase of 26%. Our Q3 increase is primarily driven by advanced SiP products, supporting the fall launch of premium tier smartphones as well as a next-generation consumer wearable product. While we have expanded our capacity for 2.5D technology supporting AI devices, high bandwidth memory constraints may limit sequential revenue growth. Given the soft demand and ongoing inventory corrections in the automotive and industrial market, we anticipate this end market may stay fairly flat compared to Q2. We expect gross margin to be between 14% and 16%. We are anticipating a higher than seasonal material content due to a product mix concentrated in advanced SiP. While this does constrain gross margin, absolute profitability will expand at a much higher growth rate than revenue. We expect Q3 operating expenses of around $125 million. We expect our full year effective tax rate to be around 18%. Third quarter net income is expected to be between $105 million and $140 million, resulting in EPS of $0.42 to $0.56, which represents more than an 80% increase in profitability compared to Q2. Our CapEx forecast for the year remains at $750 million. Our investments are primarily focused on increasing advanced packaging capacity for 2.5D and advanced SiP as well as expanding select manufacturing facilities. Last week, Amkor signed a nonbinding preliminary memorandum of terms with the US Department of Commerce to receive up to $400 million in direct funding as a part of the CHIPS and Science Act. We are proud to be the leading advanced packaging and test OSAT headquartered in the US and this milestone underscores our commitment to ensuring US semiconductor manufacturing security and innovation. In closing, Amkor continues to execute on our three strategic pillars: First, technology leadership by providing expanded capacity to support growing demand for 2.5D enabling AI. Second, expanding our broad geographic footprint by hitting a significant milestone with our US manufacturing plans and signing a preliminary memorandum of terms for CHIPS funding. And third, our focus on industry mega trends. We believe that the secular growth drivers for the industry remain in place and that our partnerships with leading semiconductor companies will drive growth as we exit the cycle. With that, we will now open the call up for your questions.

Thank you. We will now begin our question-and-answer session. Our first question comes from Tom Diffely with D.A. Davidson. Please go ahead with your question.

Yes, good afternoon. Thank you for letting me ask a few questions. Megan, first question on the model itself. When you look at the margin guidance versus previous expectations for incremental gross margins as you ramped up revenue. I'm curious, how big of an impact is it moving from the operating expense line to the gross margin line of this new facility? What is the relative kind of quarter-over-quarter impact of that?

Hi, Tom. With respect to the Vietnam production facility going online, that movement from operating expenses, the order of magnitude isn't going to be that much more than what you're seeing in our guide between Q2 and Q3 going down. So that's less than $10 million. That is not the biggest impact to the Q3 margin guidance.

Okay. So when you look at the Q3 margin guidance and what is the biggest impact that's kind of counteracting the just the revenue growth incremental margin?

Yes. So really, what's happening with respect to our Q3 profitability is it's a function of utilization as it relates to certain products and product mix. So while we are anticipating an increase in our utilization in Q3, probably in the low 70s, the way that, that's spreading amongst our factories is very dynamic. We have high utilization in factories such as Korea that are supporting advanced SiP while others are much lower than expected and as what we had estimated previously. And that's really due to some ongoing lower demand in our traditional consumer, traditional data center. And as we had mentioned in our prepared remarks, in the automotive and industrial. So with respect to those dynamics, we are going to see a peak seasonal material content in Q3. So that's the biggest driver that will constrain flow through. That is just a function of demand and product mix. It's not a structural cost. So that's where your first question about the Vietnam cost, it's not structural. In fact, the manufacturing costs estimated for Q3, which we characterize as labor, depreciation and OCOGS that excludes materials. Those are only going to increase in the mid-single-digits as compared to the 26% increase in revenue.

Great. Okay. Now I appreciate all the extra color on the margins. And then Giel, when you look at the PC and the server market, you talked about how some of the new PCs were getting more active, but yet the servers are still pretty weak. Is that different than normal seasonality? Is there seasonality in that or is it really just a cyclical impact that you see in the PC and servers?

Yes, Tom. Let me try to give a little bit of color there. Well, first of all, we see some initial recovery in the general PC market, where we see specific growth for Amkor is in the ARM-based PCs and they were ramping up starting in Q1 and that goes into Q4 of this year with significant ramps and that goes across multiple customers here. So it's the general PC market gradually recovering and very specifically ARM-based PCs, that's where we have a good position, it's ramping up in the course of this year.

Okay. And then just on the service side, what's your expectation for that recovery?

Well, I mean, the server market is a broad market. I mean, if we take the category of AI-based servers, then we expect that to continue to grow into the remaining part of this year and also into 2025. For the conventional servers, we still see weakness. We believe that in data centers, the investments that are being made by the hyperscalers are really tuned towards the AI servers and that means that the more, let's say, conventional servers are still fairly weak. So far I cannot report an improvement there.

Thank you. And our next question comes from Charles Shi with Needham & Company. Please state your question.

Hi, Giel, Megan. I want to ask a couple of questions. First, glad to hear that you guys are getting traction on the organic interposer. You will come to the market in the first half of next year 2025. I wonder, can you provide us a little bit more color because I believe you have three customers on your 2.5D technology at least that was the case 90 days ago. One, if you can give an update. And two, what's the mix in terms of the engagement with your customers? How many are sticking with the silicon interposer and how many are actually looking at organic interposer ramp with you in the first half of next year? And if I may just the third part of this question would be the tools you put in place for silicon interposer base at 2.5D versus organic interposer base at 2.5D are they fungible? Thank you.

Yes. Hi, Charles. Good to hear you. Let me start with the last part of your question with respect to the tools that we are investing in. A bit more specifics there. I think we had a first tranche of investments finalized in the second quarter, and we went for a second tranche of the same order of magnitude of investment that will become available in the early part of 2025, late this year, early part of 2025. Most of these tools are fungible across wafer-based technologies. So that's both 2.5D with interposer as well as what you call organic interposer or high-density fan-out technology. Now with respect to the transition from one technology to another for the organic interposer, we are engaged with a broad range of customers, let's say, between 5 and 10 customers. We're currently running pilot production initial runs, and we expect to ramp into mass production early 2025. That engagement is broader from a customer engagement perspective and also from a product portfolio perspective. If we take the traditional 2.5D with interposer, then we indeed have in the order of magnitude of three to five customers where we run, let's say, a variable, let's say, volume production currently, and we expect that to continue into 2025 also. So although initially, there was expected a faster changeover from one technology to another, we now see that 2.5D will continue well into 2025.

Yes, that's great. May I ask a second question maybe to Megan. Hi, Megan. I think you previously expect 30% half-over-half revenue growth. Understandably, there's a little bit of softness still in auto, in traditional consumer, 2.5 DRAM, but it looks like there is some upstream constraints on the HBM side. Does that change that 30% half-over-half growth outlook you originally thought was going to be the case, let's say, 90 days ago?

Hi, Charles. Yes. So while there may have been some mix shift changes and we did perform slightly better in the second quarter than expected. Our full year top line expectations have not fundamentally changed. We are expecting a muted first half followed by a stronger than seasonal significant growth in the second half. And what's driving that is also consistent with respect to our strong seasonal iOS launch cycle, our new consumer wearable program. We have installed and have incremental capacity for our 2.5D. However, that will be limited somewhat with some of the high-bandwidth memory constraints and really the only change then is with respect to the ongoing automotive and industrial continuing to be weak. However, for automotive and industrial, we do expect Q2 to be the trough and we are expecting some mild recovery and increase in Q3. So overall with respect to the shape of the year with Q1 and Q2 outperforming somewhat, that might have adjusted the shape, but the full year expectations are the same.

Thank you. And our next question comes from Randy Abrams with UBS. Please state your question.

Okay. Giel, Thank you. I wanted to follow up just on the prior comments you made on outlook. Just a couple of clarifications. One on the HBM constraints. Is that a view? Is it a production or yield issue that's limiting the ramp up? And is there visibility or how your feeling is how long that constraint holds. And then I wanted to ask a follow-up. I think your full year relatively unchanged despite a little bit of impacts on third quarter. So if you could give an initial view just the tail end of the year fourth quarter, are you seeing above seasonal? And who do you see as kind of drivers continuing into Q4?

Okay. Let me try to give a little bit color on the high-bandwidth memory constraints. Currently, we're working with all three suppliers for high-bandwidth memory. So we're fully qualified for all three. The constraint is not related to yield. It's related to supply from high-bandwidth memories, mostly from the biggest supplier there. It started by the end of the second quarter, and it goes into, let's say, this month, and we expect that the next two months of the third quarter, it will normalize. But overall, it has some impact on further revenue growth in the 2.5D business. With respect to your second part of the question, what are the drivers in second half growth? There are in line what we expected in the, let's say, at the end of the first quarter, the main drivers are, as Megan already highlighted, the ramp of seasonal launch of the iOS system together with Android recovering. So although that in the second quarter, Android was sequentially slightly down, we still see a 20% year-on-year in the second quarter and we expect that strength to continue in the remaining part of the year. So definitely, in the communications side, we expect strength in the second half. Also, the ramp of the IoT hearable device that will be a meaningful contribution in the second half. We executed that qualification, and we expect that ramp to go as planned, as is the 2.5D capacity expansion that we put in place. And there, the outlook, the forecast is in line with expectation and that will grow. Megan already mentioned the rebalancing in the automotive and industrial markets that takes slightly longer than expected, although we believe that the second quarter of 2024 is the trough and that from here on, we see a much more balanced supply chain. The inventory situation on most of the subsegments in automotive is much more balanced than it was when we started this year. And the feedback from our customers is that going into Q3, there will be a slight recovery and then going into Q4 and 2025, the automotive market will go back to original seasonality.

And Randy just to add to your question about Q4. Our historical seasonality for Q4 is usually sequentially flattish to up or down plus or minus 1%.

Okay, just to clarify, I have a couple of follow-up questions. The mention of being flat plus or minus suggests that your expectations, as everything is coming together, appear to be somewhat improved at this stage. Additionally, regarding iOS, I noticed your competitor also referred to an earlier build. Is that what you're observing? I recall that in the second quarter, you had some variation. Do you think it was slightly earlier, which could have affected that, or do you believe it is genuinely strong? It seems to indicate that the fourth quarter may perform better than just being flat plus or minus.

Yes. We believe it's for the iOS ramp, we saw second quarter revenue better than expected. In my view, there are two elements to this. One is indeed an early build for the second half ramp. And also we saw a very large correction in the first quarter. So I believe that correction was probably too significant, and therefore, it needs to be corrected a little bit in the second quarter. So two main effects. The net effect is more strength in the second quarter, but we also expect that to continue in the third and fourth quarter.

Thank you. And our next question comes from Craig Ellis with B. Riley Securities. Please state your question.

Yes, thanks for taking the question. I wanted to follow up on some comments that seem to have been a theme through a number of prior questions and it relates to the changes in views versus three months ago in auto and industrial, the traditional server part of the PC market. And some of the upstream constraints in high-bandwidth memory. I'm wondering if it's possible to one just rank those in terms of how they're impacting the business in the third quarter versus what you would have expected three months ago? And then in aggregate, can you quantify what that impact is versus what you were thinking back in the April timeframe?

Megan, can you comment on that?

Yes. So Craig, I'll give some color there. One, we didn't guide Q3 a quarter ago. So just want to establish that there, that there's really not a baseline for us to compare to. But as far as it does really, we did indicate the margin profile, which would have had a product mix associated with that. The auto and industrial, we had expected that to start to recover faster than it is. So that's probably the largest of the three that you mentioned, followed by the traditional data center. That one also has continued to decline. It's more difficult to see that within our computing segment because that's being offset by the strength in ARM-based PCs as well as the really significant 2.5D. And last, that's not, that is impacting Q3 is, as Giel mentioned, a bit of a slower start on the incremental capacity that we've installed given some of that high-bandwidth memory constraints that we are seeing start to come back through. So from that perspective, that's the order of magnitude on those three items.

That's really helpful. Thanks for clarifying, Megan. The second thing I wanted to do is just understand the A&I market dynamic a little bit better. So it's encouraging to see that the business should be up a little bit in 3Q. The question may be more for Giel regarding what you're hearing from some of your bigger customers in that area. And admittedly there might be a lot of smaller ones, but what are they indicating about the timing of when that business really starts to come back? Would it gain materially in the fourth quarter or are they setting an expectation that it really isn't coming back to something resembling prior health until 2025?

Yes, that's a good question, Craig. Based on customer feedback, I can provide some insights. First, we’ve noted that the second quarter of this year has been identified as the low point for automotive across our customer base, although there are some exceptions. Our exposure to the automotive sector shows that our customers consistently signal that the second quarter is the trough. Looking at Amkor's year-on-year revenue, we have seen about a 20% decline in automotive and industrial. Moving forward, we anticipate a gradual improvement. In the third quarter, we expect a moderate increase, likely in the low to mid-single digits compared to the second quarter, with ongoing improvements projected into 2025 before we return to our normal run rate. Additionally, it's crucial for us to assess and confirm our market position, and we believe we are still gaining market share in the automotive sector. You may have noticed our recent announcement in Europe with a major power company in automotive for silicon carbide, where we are co-investing in our factory in Portugal. This project is progressing as planned, and we are even accelerating some aspects because demand is expected to rise in the latter half of 2025 and into 2026. Overall, we foresee gradual improvements into the third and fourth quarters, continuing into 2025. However, predicting exactly when we will be back at full operational capacity is challenging. Feedback from customers also suggests that Tier 1 suppliers and OEMs may be overreacting to the inventory situation, as some Tier 1s seem to be reverting to a two-week inventory level. This might indicate a shift out of this correction by 2025. For the time being, this is our outlook.

Thank you. And our next question comes from Ben Reitzes with Melius Research. Please state your question.

Yes, hi. Thanks for the question. I wanted to talk about gross margin and utilization rates. So if you could just discuss the puts and takes on gross margin as we head into the second half. It looks like about a 0.5 point improvement on the midpoint in the 3Q and I was wondering if you were going to get more leverage in the 4Q, how material that could be, given the pace of sales? And then I know you had a depreciation benefit in the prior quarter. I just was wondering if there's any other puts and takes we need to be thinking about with regard to what we just saw in your outlook. Thanks.

Hi, Ben. Okay. So with respect to our Q3 guide with 15% at the midpoint for gross margin, utilization is what is constraining that, which has been, I would say, had an impact on the product mix based on the nature of the products that are underutilized or the factories that are underutilized. So with respect to our automotive and industrial being weaker than expected. That is having an impact on what we would say are highly profitable products as well as the 2.5D where we're not seeing the magnitude of sequential growth that we would have anticipated. That being said, we are seeing really strong growth as it relates to communications and consumers that is primarily supported by our advanced SiP technology, and that has a very high material content. So while that is constraining the gross margin to around 15%, we are seeing very significant bottom line earnings expansion. So you'll note that our bottom line EPS is growing more than 80%. So there continues to be significant leverage in our financial model. You also asked about potential Q4 puts and takes while we're not guiding Q4 at this time. We do anticipate some of those product mix items to have some growth, as Giel has just mentioned with respect to automotive as well as our 2.5D technology specifically. And then you also asked about the depreciation change in useful life. That was a $0.05 benefit and had an impact of around 100 basis points in Q1. That's a fairly similar impact for Q2. That will also be similar albeit it will somewhat decline as we get into the second half and some of those assets roll off. Did that address your question, Ben?

Yes. So the depreciation was still a $0.05 benefit in the quarter you just had and it will diminish as you go throughout the year. Is that right?

Correct. Yes, comparable to Q1.

Okay. And then just with regard to the interposer comment that Giel made that for the first half of '25, is there anything that we need to be aware of with regard to seasonality or impact on margin as you make that transition and help those customers move to that? Is there something that with regard to the model, when it comes out that we should be aware of?

Well, a few comments there, Ben. With respect to the organic interposer versus the 2.5D with interposer, the value add that we deliver in organic interposer is more significant, is larger than for 2.5D because there we procure the interposer. That interposer is not present there, but we delivered basically an organic interposer ourselves. So we expect some uplift in March in that transition. How that transition exactly pans out depends on the product changeover and what we're currently seeing is that products in all the technologies seem to last longer and that some of the newer products may be launched a little bit later than expected. But that's the impact that we foresee.

It's safe to say that communications are performing better than expected and that trend is continuing throughout the year. Is the same true for the AI PC, or has it been somewhat muted? There have been reports suggesting that the AI PC is not performing as well, but you mentioned it will increase sequentially throughout the year. So, are you experiencing strength in the AI PC segment similar to what you're seeing in smartphones, or is it falling short of expectations? That's my final question. Thank you.

Overall, the absolute revenue in AI PC is still small. However, I mean, the transition from x86 to ARM architecture is definitely an upside for Amkor. And we believe that we have a good market position there. We started with our lead customers three generations ago, and they are currently ramping up. They're also expanding their own silicon in that same PC. So that gives us more traction. But also a second customer that launched their products that seem to be very attractive for the PC makers. And they're now launching new products in the market with that architecture. So we believe it will have lacks and it will further grow. Some specifics over the growth during the year or in the year, we see that over the quarters, there is a step-up quarter-on-quarter, and from what we see now, that will continue into 2025.

Thank you. And our next question comes from Toshiya Hari with Goldman Sachs. Please state your question.

Hi. Good afternoon. Thank you for taking the question. My first one is on Q3. Your Q3 revenue guide. Megan, you mentioned auto industrial should be up modestly if I caught that right. Curious how you're thinking about comms, consumer and computing. Comms from a seasonality perspective. I'm guessing strong double-digits. But computing in particular, with some of the HBM constraints that you spoke to, can that business grow double-digits or is it more like single-digits? Any sort of quantitative context there would be super helpful.

Hi, Toshiya. Yes, I can give you some color there. So as mentioned, automotive, we said would have mild growth sequentially. As it relates to computing, we aren't seeing that double-digit growth that was possible given some of the dynamics with the high-bandwidth memory constraints. So that's also going to be in the low to mid-single-digits for computing. Communications is going to be very strong. Historically, we've seen very, very strong sequential communications ramp. That being said, we did have a stronger than seasonal Q2, so that Q3 seasonality will probably be a little bit softer. But overall it will be very significant. Consumer, however, with our new launch of a IoT wearable device that will have the largest percentage increase sequentially for the quarter and that can be up to 75% or more sequential growth given the scale of that launch.

Okay. Let me start and then Megan can help on the financial part, Toshi. The facility that we're planning to build in the US is a sizable facility. I think we started off with a 40,000 square meter clean room. And currently we're working with the design companies to look at closer to a 60,000 square meter clean room. So it's sizable. How much will it contribute to Amkor total? I think it will be less than 10%, significantly less than 10% of our total, but that's the order of magnitude that we will grow to in the future. The rate of growth and which customers and which technologies that we will ramp up. I think that's currently under evaluation. We are working with customers on the road map and on the ramp scenarios for that facility. And there is definitely a high interest in the high-performance computing part, but also in the communication part. So early to tell, but it's a significant sizable facility in the size of what we're currently having in our K5 facility in Korea. And the same technology will run in the US as what we're currently running in Korea.

I can add a few comments on the investment. As far as it relates to '24, this won't have any material impact on our 2024 investment. We'll start to see some of that CapEx come into play in 2025 as we're planning to break ground in the second half of 2025. But the most significant investments, which will include our machinery and equipment those will happen in '26 and '27. So as it relates to CHIPS funding, it's a bit too early to tell. Typically, those will lag the investments themselves. So as we get closer, we'll update you. With respect to '25, we do anticipate being able to continue in our low teens capital intensity.

Very helpful. Thank you so much.

Thank you for taking my question. My first question is regarding the strength in AI ARM-based compute, which I understand is still quite small. Can you provide any quantification of whether this is compute revenue or total revenue at this point? Additionally, where do you see it heading into 2024, and what are your thoughts on the long-term growth trajectory?

Hi, Krish. The line is not very clear. But the part of your question, the second part was, as I understood it well, the revenue contribution of AI currently. What I can say there is that we shared the capacity ramp that we tripled our capacity by the second quarter of this year and we executed upon that. What we can share on top of that is that we share for 2024, we will quadruple our revenue versus 2023. So it's a significant ramp for the year. The exact percentage of total revenue for Amkor is still in the single-digits, but it's growing rapidly.

Got it. And then a follow-up for Megan. Thanks for the color on the $400 million in P&T from CHIPS Act. But if I understand as well the $400 million in P&T you get from CHIPS Act it implies you'll spend a total of $2.7 billion in CapEx or probably $2.3 billion out of your own profit. Is that a fair assumption? And you mentioned that the majority is coming in calendar '26, '27. So should we assume this $2.3 billion that we spent from calendar '25 to calendar '27 or calendar '28?

So, Krish, again, we're trying to make sure we heard your question appropriately. I think you're referring to the timing and magnitude of our US investment and how that counterplays with the CHIPS funding. For the full project, we're expecting to spend about $2 billion. We are going to do that in phases, which is what this particular CHIPS program is aligned to. So as it relates to Phase 1, we expect that to, within the next three years, be up and running. So we will have had an investment in the order of magnitude of $1.5 billion in order to get that phase up and running to its fullest. The timing of that investment will be concentrated in '26 and '27, and the CHIPS reimbursement traditionally will lag that investment as it relates to that. We can't share further details on that at this time.

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

Okay. Thank you. Let me recap the key messages. Amkor delivered second quarter results in line with our expectations with revenue of $1.46 billion and EPS of $0.27. We successfully brought online additional capacity for 2.5D technology and qualified our Vietnam facility to support production ramps in the third quarter. We are expecting third quarter revenue to grow significantly with revenue of $1.835 billion at the midpoint of guidance, reflecting a 26% sequential increase. We reached a significant milestone with the US Department of Commerce and aligned on the preliminary memorandum of terms for up to $400 million in direct funding under the CHIPS and Science Act to support our planned advanced packaging and test facility in Arizona. Looking forward, we remain confident that the secular growth drivers for the industry remain in place. And with our strong technology leadership in advanced packaging, a uniquely diversified global footprint, and partnerships with lead customers, we are well positioned to accelerate while exiting this cycle. Thank you for joining the call today.

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