Magnachip Semiconductor Corporation (MX) Earnings Call Transcript & Summary

Good day, and thank you for standing by. Welcome to the Magnachip Semiconductor Corporation sell-side analyst briefing and webcast. [Operator Instructions] Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, YJ Kim, CEO of Magnachip.

Thank you for making this time to come here in NYC this morning to listen to our analyst briefing. So let me go ahead and get started about the new Magnachip. Let's power up the Magnachip. Forward-looking statements. So today, myself is here, the CEO; and then Shin Young Park, CFO. And then we have a CTO of the Power Group, Woo Hyuk, he joined us September last year; and we have a Head of Strategy, [indiscernible] and [ Mira ] who is the Head of the Strategic Marketing for Power, right here. So I'll give you a background on the business while we're doing this in our product strategy and how we're going to grow about. And then Woo will give you some of the technology and the next generation of our advantages, followed by Shin Young giving the financial road map. And we can do a Q&A after each section. So that you can ask a question on each of three section as well as the additional question at the wrap-up. So Magnachip today, we have about 880 people and $221 million last year. And we do have 6 different sales sites where we have active offices or [ liaisons ] And we do have 3 key sites in Korea, 2 design and administration center is Seoul in Cheongju, and Gumi is our fab making the power. Our power IC is a pure fabless. So we actually make it using external foundry. Last year in revenue and the product, Power accounted above 84% of total revenue, and we actually sold into consumer communication and computing where all these segments grew double digit last year. And then the industrial automotive was weak in the power market, but we only declined only a couple of percentage points where the industry peers, I think so a more than 20% decline in those sectors. I think you know our history very well, so I'll go very briefly. So we are listing NYSE in 2011. And then we sold our foundry business in 2020 because of pure-play products company focusing on display and power. And now we are exiting the Display business. So we'll focus on Power that's Power Discrete and Power IC. I will go more detail later on and any questions. So today, you saw our announcement that we are exiting the display driver IC. The rationale is you've seen some of the diminishing return on revenue due to multiple factors: a, we had the global supply shortage issue in the 28-nanometer. So that has impacted us in 2021 to 2022, and then followed by trying to penetrate into several new markets, it's taking longer than what we anticipated, it's taking longer R&D. So we thought that it's not going to be a quick turnaround to provide profitability in Display. So we decided to focus on the Power, where the market is much broader. We have a good traction and more than 200 customers in Asia and we are penetrating to automotive, and we just introduced new generative products, which are much compelling and we can address higher value-add markets and so that's why we are focusing to be a power company. We have 3 transition as you saw from foundry display power to product only, now pure power. About 5 years ago, Display -- the Power accounted about 1/3 of revenue. So when we exit the Display in the next few months, then the company going forward will be 100% power company. And today -- last year, it was 80% and it will be 100% in probably 6 months or so. Okay. Let's go into the power and what it brings to our future at energy. The -- one of the key reasons is that it has a huge SAM, not just TAM alone, SAM. Previously, last year, the SAM was about $18 billion with our product portfolios. With the launch of these new products that you saw today, 27 new products that can address 100,000-watt power ranges. We can address about $27 billion of market by 2028. So that's a huge market that's between the MOSFET, IGBT and the Power IC? And if you look at the market itself, it's about 10x larger than OLED drive IC. So -- and it's not dominated by the 1 customer or 2 customers, it's all over. And the key also point is that 70% of the power market is in Asia. So 5 being Korea with the manufacturing site in Korea as well as back-end manufacturing in China and also design centers both in Korea and China, we provide proximity to our customers. And with the introduction of the compelling next-generation products, which we'll go in more detail, we can address much better markets and customers. And in order to support these new generation products, we are upgrading our facility. And as we said today, we will convert 70% of capacity to support the new generation products by 2027. Now these are some trends going on, and on electrification, the cars become electrified, not only EV, hybrid plug-in. So typically, in the competition of the other engine bays, the cars may consume 20% semiconductors. Now it's up to like 60% semiconductor consumed in the EV. Also, the AI data center and green energy, so forth. I mean it is driving a much robust ruggedness and faster switching time and at the same time, most power-efficient environment on the power system. So that is really good because the -- our new generation, it's all about its high performance and high reliability, ruggedness as well as fast switching speed and low power construction and compact solutions side. Now last year, we addressed only 37% of revenue in AI from industrial and automotive. Our goal by 2028, 62% of revenue comes from AI industry automotive, but if you look at the real market out there, it's 80%. What it means is that if you're a Tier 1 today, you are selling more than 80% to the automotive, industrial and AI or server markets. So we -- 62% is not quite 82%, but we're getting there with some new generation products. And that's all -- it is, it's all about the product you make and can address those markets effectively. If you look at it from a geography perspective, this is market breakdown by region. So as you can see, Asia is 69% of the total power consumption and Rest of Asia of 33%. And our goal, today, we provide more than 95% in Asia, by 2028 to provide about 85%. So why is this important data, we are as the proximity of the Asian customer? And B, for example, like automotive, today, we have less than 2% last year of revenue. But if you look at Korea itself, an automotive power market last year is more than $300 million. So by our backyard, we have a great room to grow -- and so our goal is to grow about $170 million in Asia to another $100 million in the next 3 years or so. So now let's talk about the Pure-Play Power company in more detail. So I think you heard this morning, I'm measuring 3-3-3. This is our new strategy, 3-3-3 strategy. What it means is that in 3 years, we're going to hit $300 million revenue run rate with 30% gross margin. So today is March 12. So 3 years from now, March 12 in 2028, we want to hit these objectives. Now we do have a more milestone that we want to hit by excluding this strategy. By end of this year, we want to hit quarterly adjusted EBITDA breakeven. Next year, adjusted operating income comes on list and following year, positive adjusted free cash flow. Shin Young will talk more detail in our section today. So we mentioned about Phase 3 today in the call. So let me give you a walk history back. So when we started the Power business in 2007 up to 2013, we had a first-generation product, which we call Phase 1. At that time, our product was starting and we were barely touching application like mobile phones. And with the Gen 2 product coming in, then we get into TV, computing, smartphone, e-bike, industrial automotive. Now with Phase 3, Gen 3 product, we are getting to AI phone, to servers, traction in automotive and energy storage systems. So this is the evolution. And there's beyond Phase 3 where we get even more. So we are just now launching the Phase 3 product line, where it's more competitive and we can address up to higher-value markets. Okay. So let's look at our road map for future success. So what makes a great power company? New generation products need to be high performing, low power consuming and fast switching speed. You need to have a very good geographical support. In our case, we are at the center of Asia, where 70% of our market is there. We have both design and manufacturing sites as well as sales offices. And you need to be able to really address the growing high-end market. That's automotive, AI and industrial. And our new products enable that. And we have in-house fab, it's automotive grade proven. We've been shipping the automotive products for the last few years now, and we have a very reliable supply chain and addressing our customers, both internally as well as using our sites. Now let's look at the catalyst that is driving our growth and profitability. First, as I said, it's about the product. We just introduced 27 new generation products today 2 IGBT and 25 super junction products. We now announced that we're upgrading our fab to make more new products. The new products have 30% to 40% better performance and we have more than 30% dies per wafer translate to 30% more revenue per wafer, although there's a back-end cost so you have to cut it separately, but it provides performance and cost advantage. And we have enough fab. And then on 1 plus 2 enables better market opportunities and the TAM is huge, and I told you our SAM is growing from $18 billion to $27 billion. So it's a big SAM that we are addressing with our new generation product. So our goal ultimately becomes a very leading power technology concerning in Asia. Now let's look at market and graphically how we are doing this. Between the Gen 1 and Gen 2, we are addressing applications in up to 10 kilowatts. So at the low end, is mobile 100-watt to like e-bike, some actuator and TV and then 1 kilowatt e-bike motors and then 10 kilowatt is like industrial motor, right. So with the new generation, wow, we get up 200-kilowatt and more. So we can get to much better solar power system, automotive traction, EV charger with super junction, Gen 6, on board charger and the servers and a better computing market. So the SAM from current $18 billion grows to $27 billion with our new products. Now not only we are just introducing. We're accelerating the new product introduction. As you can see, last year we did about 7. This year, we expect to be more than 40 with 27 just announced its fully qualified commercial samples. We expect production in second half this year. Next year, we're going to increase more to 55, okay? Now Mr. Woo will explain later in his section, how we can do this and why we can do this. So if you look at the history, the Gen 2 that were shipping until last year '24. It had a better than 15% performance per unit area than our Gen 1. Our latest Gen, the new Gen, Gen 6 super junction, Gen 6 IGBT, Gen 8, LV and Medium Voltage provide more than 30% performance per unit area. So this is how we can be competitive and address new market, the product coverage from all the way from 12-volt to 1,200 volts with high current and addressing 100-kilowatt applications. The new generation product, as I said before, buying more die for wafer anywhere between 25 to 35 more dies per wafers, representing more revenue per wafer, more than 30% revenue per wafer enabled and with the new generation, we'll be able to effectively address better markets where better margins like from going TV consumer to server, e-bike to motor control from solar to automotive traction, okay? And we expect to fill the fab with the kind of product thing. So by 2027, our goal is to have more than 50% of revenue coming from new generation, for the full year. Now let's look at the revenue per wafer portions. Our current forecast, as you can see, 25 -- 27 products that we introduced will contribute some but not too much this year, right, next year and more. So as I said before, by end of '26 this run rate of this should be about 50% of the total revenue. And then obviously, '27, it's much more than half. Now that would drive a better gross margin because we are talking about low-cost products but a high performance. And each product has much better performance than previous generation. And today, Power IC we are #1 provider of Power IC to OLED market in Korea, and we are also a key manufacturer of the TV PMIC and Power IC in Korea. We expect to expand to our adjacent markets and neighbors. Our fab is already automative grade. We've been shipping for the last few years. It's in Korea, it's ready to support the high-performance new gen production with the $65 million to $70 million CapEx that we mentioned about upgrading our facility. So this is actual capacity. So by upgrading that $65 million, $70 million, we can convert the new generation portion equipment capability to grow from about 10% to about 70% by 2027. And if we ramp more in the revenue that I showed you previous slide. And obviously, our gross margin at the company will go up. This is how we go to make 30% gross margin in 3 years. In terms of market opportunity, similar to what I showed you, we started with mobile, then move up to more consumer devices and low-end industrial and move up to little high end industrial here, and now with the new generation, we are really going to high value AI and data center to traction, to automotive, to energy storage system like that. Again, this is example. So as you can see, we have more than a dozen design wins in automotive and many of them already shifted in production and more to come this year and next year. So -- the one that is not a star is our target and the EV traction, HVAC all that, with the new generation product, we'll be able to address that market, and we look forward to bringing new announcement in the future. The AI. Everyone talked about AI, but the AI is all over. It's not just in the server, it's in the smartphone, right? Samsung Galaxy S24, 25 to now the laptops introducing is AI chip onboard, then obviously moving to data center server. So all of them requires very heavy computing, whether it's on the phone, on the servers, heavy computing. What is that? It means this takes a lot of power. So what it means that you need to be able to provide power efficiency. Our new generation provide high power product with fast-switching and also much better thermal performance. So that's why we have mass production on the phone AI to Power IC in the IT already, and these generations will get into here more effectively. In terms of supply chain, reliable to Asian customers. We have fab in Korea, and we are packaging back in China, Korea and Malaysia and so we can provide to customers in China versus out of the China customer very effectively as well as design to manufacturing and the Power IC, we do both in Korea and China, so we can provide to China or elsewhere. So this is how we can service Asia for Asia much better. And this is market data from 2023. We came to rank 14 within 13 years, starting power business with the Gen 1 and Gen 2. Now with our Gen 3, I think we can move up serving better markets, and hopefully, we can move up a year in the future. So to wrap up my section. So first, we said the new generation products is enabling a profitability. And we don't upgrade so that we can ship more of these products. And with these both the product and the upgraded fab, we can address the market much better. And we would provide stable supply chain to our customer. And execute to hit 3-3-3. So that's our business section. Now I'm going to introduce this, Mr. Woo Hyuk, but before that, any question and answer before we go through the more product technology session.

So you talked about different frequencies and different power levels. How does traditional silicon play in versus GaN and SiC in your strategy? Or are they going to be blended together into the products? Is there a separate road map?

Very good questions. So let's go back to the slide. It is much better to explain. So let's look at this slide, right? So silicon carbide is in this side and GaN in this stack. Now let's look at the silicon market is $55 million by 2028. Silicon carbide [ $5 billion ]. So you're talking 10x larger market. Silicon carbide is only a portion of the market. So we have a lot of room to grow in the silicon. And also, maybe this is premature, but our next-gen super junction will have a very good performance. I don't want to prelude already. So I think that it's a matter of going out to market and customers, but also which depends being on the application that you can address the market. Now for example, in the auto traction, the EV car, they want the purest long range, right? In that case, they will use silicon carbides despite it's much, much more expensive, right? Now let's look at hybrid car. Are you going to use silicon carbide in the hybrid car where hybrid car is very cost-sensitive and volume. No, you're going to use IGBT. Same thing, for the GaN, the quick power charger is GaN, but the GaN reliability drops off after year, right? So again, in the traction motor, things like that, reliability is very important. So the GaN probably cannot be used until they really increase their reliability. So we have a big market with silicon. And there are new silicon products, which hopefully in 2 year, we'll tell you about that can also address some of these markets. And then -- we will maybe also look at -- because now silicon carbide is so cheap, I mean, not cheap, but in terms of the foundries are offering a lot of services, so you can do things like that. But in terms of the real price performance, silicon carbide cannot come to the IGBT or super junction. So it really depends on the market and the price you want to -- any other questions?

Yes. I think my main question kind of keeps coming back to what gives you confidence to get that 50% number by, I think, 2027, 2028. Martin asked a question about the go-to-market strategy -- but I guess -- and this slide on the market opportunities kind of looks at the customers. I think my question would be what -- can you talk about customer engagements or pipeline or backlog, anything like that? And it also seems like auto would be kind of the biggest portion of that growth.

Yes, very good. So let's look at this slide -- well, okay. So this slide here, again. This is happening within the last 2, 3 years. So the momentum is getting there, right? And we're actually working on the traction motor product. So hopefully, you will hear from us in the near future, okay. So once you have this thing traction motor and here one more detail what we've been doing. And you can actually branch out and win a lot of application because this is one of the most demanding power applications in the planet today. And the other thing is that with our Gen 1, Gen 2 which is not the world's leading technology, we were able to get #1 position in these places, right? So I think that with the new product and the pipeline we have and the design wins we have, we think we have confidence to get to that revenue portion. And many of their products also within the same customer base, we're moving up the food chain from, let's say, from a single motor to dual motor to a power adapter to more main motherboard things like that, even in the computing -- and we are getting more design win with those computing space as well as automotive. Any -- any other question, Martin?

Yes. So a question regarding Slide 15 on revenue exposure by geography, when you look at each growth by geography, is there any certain product exposure or industry segments. For example, how does Korea compared to China, we don't look at where the growth is coming from our products or by end markets?

Yes. So if you look at a high level, Taiwan is a leader in the computing. China is industrial and some computing. And so the China, they have, I think, 6 -- 10 -- of the top 10 solar inverter companies in the world. We actually sell into 6 out of top 10. And we actually also sell the #1 player in China. And so as I said before, we may be doing some -- the inverter now, but we are going to move out to the energy storage service system with new gen. And China also has the e-bike and tools. Korea and Japan -- Korea has more -- a lot of consumer stuff and then Japan by 75% of their market is automotive. Korea, about, I would say, 30% is automotive and 70% is consumer communication and segment. That's how -- how it looks in the high level in Asia. And I assure that we -- the customers that we went into, we took #1 shares in those applications. Okay. I think with that, I will introduce Mr. Woo. And you will see why you can't accelerate and deliver the products. Okay. Interesting enough, Mr. Woo now first time a manager, okay? He actually developed a Super Junction Gen 1 and Gen 2 product. So if you look at the history from Gen 1 to Gen 3, it only took 3 years. So he has proven capability to deliver, but he left for about 10 years. And he came back. But while he was gone, he worked on automotive and IGBT, silicon carbide all that. So he brings in very wealth of knowledge on the automotive applications. And he came in last September. Now at the 7 months, he has accelerated our new generation bring. So the 25 new super junction products were accelerated when he onboard for 7 months. With that, I'll turn over to Mr. Woo and his English is more average. So if there's a question I should be translating.

Thank you, YJ. I'm very glad to present the advantages of our new generation product and high confidence to deliver that. I'm going to start with technological [indiscernible] from above picture [indiscernible] in this application -- emerging application is that they required significantly increase high power and high reliability. In terms of design -- device point of view, the requirements are becoming more challenging as -- so expertise in high-power system is crucial for success in these markets. And the requirement is not only for electrical and -- but also mechanical, including fabrication know-how of the production of very large size of high-quality. And automotive traction inverter is one of the most challenging application because the efficiency and safety and utilization of electric energy is the key performance of the binary system and negatively that have accumulated this kind of know-how on our fab through the R&D on the device for the automotive traction inverter since 2015. In addition to the technological leverage, we have invested in the R&D resources. We -- from 2020 to 2024, that we doubled our device R&D head count, and we will -- we are hiring additional 20% in this year, and increased the resources we contribute to the release of many products -- many new generation products in 2025 and 2026. As well as the increase of our R&D resource, we changed the methodology of R&D project from serial to parallel. The above feature shows the concept of the running R&D project, and the small gray -- each small gray box is the product to develop the existing generation and green small box is -- that means the project to develop the new generation product. The vertically aligned 3 boxes means that 3 projects running and simultaneously parallely, but the number of the box to 3 is not necessarily a real number. The picture -- from the picture below, the height of the box means that a number of the launched products that we will dramatically increase the number of the new products from this year. And through the 27 already launched the new generation product, I think it was demonstrated our R&D [indiscernible] proven.

He did that within 7 months.

Magnachip Power Discrete products consist of 3 types of devices. The first one is Super Junction MOSFET, that covers from 400 volts up to 900 volts and the second one is the medium voltage MOSFETs. It covers less than 400 volts And last one is the IGBT that covers the high voltage and high current simultaneously. We will release our new generation product or all of these type of device. In the evaluation compared to the previous version, all of our device group showed that much improved performance to the -- than the previous version. Our super junction -- new generation super junction showed the 0.5% -- over 0.5% is actually 0.8%. The improved efficiency at the system level, which means that at the device level, the device consumed the 10% less energy -- lesser energy, due to the reduced -- lesser energy consuming the thermal performance -- the temperature of the device decreased by 3 degrees in the system.

Yes. I think [indiscernible] 0.1% power efficiency is considered a great improvement in the system. So here, we're doing 0.8%.

Yes, IGBT also to reduce the switching loss by 20%. And the medium voltage MOSFET also lowered Rds(on) by 20%. Rds(on) is the indicator of the conduction loss. Here is the advantage of our new generation super junction. So we review the number of the advancement at the evaluation -- comparison valuation. So we did get the valuation with the customer applications by Magnachip. And the important thing is that we achieved this performance in advancement with smaller size of die, if they can increase they could increase the gross die by 30%. In terms of IGBT, we also achieved the technological advancement to -- with the 30% increase in gross die. The 20% reduced with switching loss that resulted in 3 degrees decrease in temperature in the system. And in terms of medium voltage new generation MOFSET and we expected 25% smaller chip size and 20% reduced conduction loss, so we can create new value by applying this chip into the more compact packages with 30% reduced foot print and 50% reduced height. I conclude then our new generation product, we'll provide technological advantages with high probability. My team has high confidence and well prepared to deliver this product. And we are all excited for the new generation era. I think, I believe, this is -- this is a point -- this is a turning point of Magnachip to become a great power company. Thank you.

Open for questions.

Yes. You showed the development process going from cereal to parallel. What are the key underlying elements that allow you to take on in parallel product development?

[Foreign Language]

[Interpreted] So one of the key factor is the resource. And as you saw, we doubled the reserve from 2020. And this year, we're hiring 20% more already higher 10% more . And that creates a foundation to multiple projects at the same time. Question?

So in terms of investments, you talked about R&D, if you also invest more in business development and sales.

Yes. Very good question. So not only in R&D, the business development. So for last few months, I think we have increased the biz dev and sales and [indiscernible] power by, I would say, about 10% to 20% more. So you're absolutely correct. And we now have a strategic account managers in China and Taiwan. So that's a change that it occurred in the last 6 months.

Maybe a question for Woo, why did you leave MagnaChip and them why did you come back?

[Foreign Language]

[Interpreted] So doing each generation is good. But he felt that rather than working on the next gen, he thought he needs to get better experience getting into how to develop automotive products. So that's why he went to the automotive development side. And then with the new knowledge and expertise and the development he has gained, he feel that coming back to MagnaChip makes perfect sense now that we have a fab and all that. So he can take you to the next level for success.

So the concept of being an Asian supplier into the Asia supply chain has been talked about with MagnaChip for years. I just want to know maybe perspective of what the advantage is with the customers in Asia when you are based in Asia as a supplier versus an International supplier. If you kind of clarify those advantages that would be helpful to understand.

[Foreign Language]

[Interpreted] So the high-end application requires a lot of know-how, but the customer doesn't know everything. So it's important to work with the supplier and having the close proximity, that really helps to really come up with a better product and solution together. And in our case, it's our backyard, Thank you. Now with that, I'll turn over the call -- the presentation to Shin Young Park.

Thank you, YJ. For the finance the -- I mean, YJ talked about the market and product road map and also Mr. Woo talked about the technology advancement and how we are going to get there. So what does it really mean to me, the finance perspective, like how that we're going to look like in the context of the financials to become a pure-play power company. So we kind of gave out our goal an ambitious goal like 3-3-3 Strategy, $300 million revenue run rate and 30% gross margin target in 3 years' time, YJ talked about that time clock in Q1 of this year. So like we were kind of thinking of this one, if everything executes very well, we -- our goal is to achieve that strategy. So 2025, and I mentioned that during the call, we are kind of calling transitional year, meaning 2025, we're going to set the stage for us to become a pure-play power company. But like with all of those that the new power products are rolling in, but we are still expecting our gross margin in 2024, we're going to be like about 100 basis points lower than 2024 on an apple-to-apple basis. So why that happened? I mean you guys had a question during the call, and that was because of the transitional foundry like first of all, that freed by the end of last year. I mean, although the margin is negative, but actually it had more positive impact on the power's gross margin because they were absorbing the fixed cost in our Gumi Fab. So we used to have like $40 million revenue with the transitional foundry in the Gumi fab. It came down to like $11 million last year. We don't really expect anything in 2025. So that kind of idle capacity created by the fading out of the phaseout of the transitional foundry that we're going to be impacting our Gumi utilization rate for sure. And you're going to see that impact mainly in the first half. I mean the second half as well. But the second half, the impact we're going to be partially offset by the new power product, which we're going to begin production in the second half of 2025. So the 100 basis points lower in 2025, mainly probably lower in the first half. It's going to improve a little bit in the second half. So overall, but still, you're going to see that the transition impact in 2025 in terms of profitability for us. So first one is our Power Discrete and Power IC together the revenue grew -- I mean that has grown 12 months -- I mean, 4 quarters consecutively year-on-year basis. I use the Q1 revenue at midpoint, and you can still see that we are going to grow at about 12% at midpoint from Q1 of last year to Q1 of 2025. I mean I don't really need to reiterate the fact that we've been -- the revenue has been diminishing quite drastically for us for the last couple of years. And if you look at the overall company, but if you look at the display revenue kind of drop from 2020 to 2024, it was almost like [indiscernible] that we used to do that in 2020. And needless to say, if you look at -- I mean, adjusted EBITDA as a profitability measure, you can see our profitability like worsened in the most recent years. So as a result, we kind of made a decision that we're going to become a pure-play power company, and that is kind of how it shows in the graph. So -- and it looks beautiful to me [indiscernible] is execution, right? So the revenue and gross margin, we talked a lot about like 2025 being transition year and how we are going to. But in 3 years target, we said $300 million revenue in 3 years' time, 30% gross margin, the OpEx. So the OpEx probably it really depending on the quarter. But if I were to [indiscernible] look at the previous last years and history for the display portion, probably about 35% to 40% of our OpEx without the compensation charges were -- I mean, that was tied to the display business. But at the same time, our shared service function and the overhead supported both display and power. There's got to be some efficiency between those 2. So my job is core is [indiscernible] how to kind of rationalize our spending and realign our kind of spending level to become a pure-play power company and Magnachip to find the right spending level. So 2025, to achieve quarterly adjusted EBITDA breakeven by the end of Q4 2025, I think the key thing is we're going to find that the cost reduction initiatives and execute them and better align our spending level to become a pure-play power company. I think that's to achieve that milestone in 2025. I think the point of this slide is that even though the revenue growth like from here there and the gross margin expansion with all of those new products rolling out, the new product revenue number over 40 products in 2025, over 55 in 2026. I mean that will all go into play [indiscernible] but this one is, I mean, in my view, like for the control perspective, that's important in this slide. If you look at that, I don't believe that you are going to measure the incremental portion and how much you're going to grow. But if you look at that, that incremental year-over-year is really not proportion to how we are going to grow in the revenue and gross margin. And the way that we can actually do that is pretty much the majority of our OpEx is kind of fixed in nature like the labor. And the power -- usually, it really doesn't -- R&D expenses really doesn't tie to the number of the R&D development. I mean we do spend more, it's not like this. So for the OpEx side, once we find the right alignment of the new kind of direct spending level, the increment collect year-over-year. OPEX cost increase every year, but it's really not going to be as much as a revenue increase or proportion to that expansion in our [indiscernible] metric. And also -- probably I can discuss this on the next page. The stability is we are showing the similar financial metrics here. So in 2025, 2 things that happen, the cost initiative actions and everything, we are likely to incur some onetime restructuring charges. But I guess, I mean, we don't -- we are not in the stage. So like once we know the magnitude and the right timing, like how much we're going to spend to do the restructuring, we're going to report that in a timely manner. So that's going to be -- that can be incurred in 2025. But if you look at this one, rather than 2025 going into just 3 years' time, maybe I should give a little color on the 2026. So looking into 2026, to hit the adjusted operating income breakeven and positive. And I told you the OpEx incremental will not going to be as much as like the revenue growth. But at the same time, the 40-plus new products coming in, rolling -- start rolling in from the second half of this year and 55 next year. So that' going to be -- we're going to be accumulating the contribution to the revenue sequentially. So it's not like we are not really talking about like hockey stick or seasonality. I mean there's going to be some seasonality, but at the same time, in 2026, I can see the revenue expansion and also the margin expansion of course are going to be probably throughout the quarter in 2026 to get to the adjusted operating income breakeven. And again, the OpEx incremental portion will not going to be significant, and that's how we probably can get there. And also the depreciation wise, we don't really expect our whole CapEx of $65 million to $70 million like in 3 years, we're not going to be fully impacting our depreciation until 2027. So probably 6% to 7% of the revenue, that percentage we're going to stay in 2026 now versus 2027. That range will not going to change materially in my opinion. I mean, tax looked clear in the past because we were losing money and we lose money, deferred tax assets and net loss [indiscernible] carry forward and that we're going to kind of trip up your kind of ETR, effective tax rate. Once we start to profitability, I think it's going to be more normalized and our target effective tax rate is between 21% to 23%. So this is our Gumi Fab capacity transformation. We've talked about the capacity only. In 2020, we sold the foundry and we start to invest in our Gumi fab to increase our power capacity. So there was 23,000 wafer starts per month in 2020. We increased that by 70% now. But if you look at 2025, only 10% of the total capacity was for the -- to support the new power products. So we spent for the last 5 years, the total CapEx composed of 2 pieces: one, expansion Fab upgrade and new equipment purchases versus maintenance. But if you look at this one, we are more focusing on the absolute level of the capacity increase from the last 5 years. What we are doing from 2025 in the 3 years is rather than simply increasing our power capacity and converting our idle capacity from the foundry, we want to optimize our Gumi Fab and upgrade our Gumi Fab, meaning we want to increase the capacity to support the new power product from 10% to 70% plus. And that's how we're going to support all of the new power product will come in this year and next year with a higher ASP and the revenue per wafer, the higher revenue per wafer. So we are going to -- with that, we can achieve our revenue growth as well as margin expansion. And of that $65 million to $70 million investment in our Gumi, we already secured $26.5 million of a credit line with a bank. This is supported by the Korean bank and the Korean government and the interest rate was less than 4%, and that's 10-year maturity. So we think -- I mean, that we're going to partially fund with our huge CapEx. And of the $65 million to $70 million range that we are going to invest in the Gumi Fab. I mean that's a little thing I just kind of put it here is kind of telling you we're going to spend more in the first 2 years, 2025 and 2026 and 2027, the CapEx will go to approximately 25%. That's how we are going to manage our cash and looking at our cash level and the right timing to hit the $300 million, 30% gross margin in 3 years. So with the upgraded fab with the capacity and the investment in Gumi -- in our Gumi Fab with the higher composition of -- to support the new power product and also the new generation power product portfolio that YJ and Mr. Woo explain to you guys, like we kind of set our goal in the next 3 years. Firstly, 2025, more from the expense optimization side, we target to achieve quarterly adjusted EBITDA breakeven by end of Q4 2025, adjusted operating income positive in 2026, adjusted free cash flow turn to positive in 2027. The reason why we put adjusted before free cash flow. Usually, free cash flow is a liquidity term, like mainly cash from operations minus CapEx. The reason why we could adjust it is like our unique situation that we are the Korean company operating in Korea. Functional currency is Korean won, we report in the U.S. dollars, and we maintain pretty much our cash in U.S. dollars. Even though you don't do anything, you actually have transaction gain or loss impact going through the P&L. So sometimes, when you look at the cash from operations, you actually get the benefit depending on the move of FX, but sometimes you get penalized by something you cannot really control. So we don't think we should be assessed or judged by something that we cannot control. So we want to set our goal, adjust free cash flow, meaning we are going to adjust for that transaction FX gain or loss we're going to be excluded in [indiscernible] kind of the typical operating cash minus kind of CapEx term. So the ultimate goal is to $300 million revenue run rate and 30% gross margin target in 3 years is what we hope to achieve with all of those that we explained today. Question?

So you talked about capacity for the new products going from 10% to 70%. I'm wondering what the shape of the in between is how long the equipment takes to get in place and then by the end of '25 and '26, how far along are you to those? Because that seems like an important part of your gross margin improvement, right, to be ready for that.

So that's why 2025, 2026, and [indiscernible] and that's a 3-year time frame if I kind of draw the line, I don't expect that from going to 25%, 21% to like 30%, all of a sudden, it has to be incremental. So like not only the equipment -- because the equipment is also for the R&D for the next generation upgrade of fabs like all those purposes. So for us, not only that the equipment placement, also the number of new products coming in rolling and producing in our Gumi Fab in 2025 and 2026. So I think probably the better way is kind of not going to be a hockey stick. It's not that it's just a onetime thing, but it will going to be more linearly over the course of the [indiscernible].

Do you have specific assumptions on foreign exchange on FX in 2025 or rather through the target?

I mean it has been recently very -- the U.S. dollar got a lot stronger than we anticipated before. So I mean, we kind of look at like what the bankers are going to have there for the next couple of years. But usually visibility-wise, I mean, 2 years ahead, like 3 years ahead no one knows. So we are trying to be more conservative. So we are using whatever the most in the industry kind of average. We are not trying to be aggressive on that part, but also we just want to be at par and whatever the people are thinking right now, which is probably U.S. dollar will not going to be that much stronger like in 2 years' time frame, but we kind of look at the outlook and trying to use the average or more a little conservative side of the FX.

Still trying to figure out the -- like the OpEx levels in '25 and '26 just because we talked about kind of this 30%, 40% coming from the display and then -- but you're also talking about hiring a little bit more for R&D. So yes, I'm still trying to figure out.

About the same thing that we talked about, we're hiring a lot more engineers and marketing and sales. It's already kind of baked in like my 2025 model. So that the new hires. So the way that we can kind of offset those impact is we are trying to rationalize and reduce our cost, even though for the shared function and also the overhead. We are trying to kind of -- and we optimize our spending like now supporting only one business line. So last year, for instance, we spent $92.5 million overall for the whole company, including display. And I think probably 35% to 40% of that was for the display. So I mean, the lower side, 35%, the higher end 40%. So we are trying to kind of lower our OpEx and reduce our OpEx. That's our target, 35% to 40%. And again, that new R&D hires and all those incremental expenses are kind of baked in, we are trying to be adjusted EBITDA breakeven quarterly by the end of Q4 2025. And also, as I said, going into 2026, although once we kind of lower them and optimize the expense, incremental like from the just regular labor increase and some fixed cost increase should not be like too high we are going to control that perspective to achieve just operating income breakeven [indiscernible].

Okay. And then last question there. You had a 40% number in kind of gray area. That's -- this is what you're talking about? And then this also -- what's also baked in is all those new product launches.

Yes.

One question on maintenance CapEx on Page 52. So it seems like the annualized CapEx will also grow either '25 to '27 period. Is that right?

We usually talk about $10 million per year for maintenance. And that maintenance is not only for our Gumi Fab, we have like IT, like system upgrade for the ERP system. So there are certain timing that we have to spend to kind of renew that and also some support functions [indiscernible] like CapEx. So I guess like maybe it's probably not the right way to kind of increase. And it is increasing, obviously, we are kind of putting more equipment in our Gumi Fab. But at the same time, it's the timing of when the other departments require certain CapEx. So we generally think probably $10 million per year is a kind of safe assumption for us to spend kind of regularly on a year. So with that, I'm going to turn this to YJ for him to kind of wrap this up and give us a key takeaway.

Thank you. So just to recap our key takeaways. We are becoming a power-only company. And we are expanding into new high-value market as we described today with our new generation products. And that we can get in to penetrate to better an AI data center, industrial and automotive market segment. By upgrading the -- our Gumi facility to enable more production of new generation product that's going to improve our margins and achieve profitability by executing our 3-3-3 strategy. So with that, thank you for coming and any other questions, let us now.

I just ask again on the strategic process, like what kind of companies are involved now that kind of being opened up a little bit broader display companies themselves or the panel makers or private equity? And any specific geography that you're targeting?

We are talking to all the geography as well as multiple options. So we are looking at every angle to maximize shareholder value. And also, we are mindful of the OLED in National Core Technology. And also, we have to be mindful of the [indiscernible] as the Korean government. So we want to be very clean and both legally as well as financially to the maximizing shareholder value.

Anything specific that gives you confidence that a deal could go through this time?

We have many options. And again, it's probability and pros and cons, and we will make a decision by a few months.

Could you talk about the seasonality impacting kind of both businesses, the Power IC versus the PAS. We then talk about seasonality specifically, it was just mentioned.

Yes, typically, because we used to actually have a PAS and Power IC as the same division under Power before. And we have a lot of historical data. But usually, Q3 is the strongest season and then drop up a little bit in Q4 and then drop up a little bit in Q1 and then go up in Q2, and then go Q3 again. And the -- again, with being Asia, you have a Chinese New Year for Q1 tend to be a slower season as well.

Can you talk about the design cycles for the power products.? I know they vary by end market, but the notion that these new products are going to be giving you revenue visibility in 6 to 12 to 18 months. Just give us the time frame that, that would all have to happen for you to have good confidence in '26, '27 revenue opportunity.

Yes. So as I said, if we get the new generation product commercial ready, then it takes about minimum 6 months to about 12 months to ramp up, right? So we will have the 40 products all going to fully qualified by end of this year. And then that will set a very good foundation for '25 -- '26 revenue. And then we will have another 55 coming in next year. This sets a very good tone for '27 right? So that's how we do. And typically, we've been doing about 25 products per year. So we are actually accelerating the new products and the new products, new generation products is 30% better revenue per wafer. So it helps out in many ways.

Can you remind us the channel strategy for power? Is it primarily direct? Or is there heavy distribution use?

Okay. Very good question. So up to now, I would say the 100% in China has been disti and Taiwan has been disti. Korea maybe the 20% direct, rest in disti. So we just created strategic account team a few months ago so that they're going to drive more strategic direct account activities going forward. That was a very good question that Martin asked, it's not just product. So now with the high-value products, we now have a strategic account team. So they will be involved, whether even involved in disti, they will be involved to drive a much better high-value design. So that's what's the change.

Lastly on that, is there a difference in margin direct versus dist that we should think about?

At this moment, not much differences. But I think the -- with the new product, that will also change because you are not really in the commodities.

A question on your confidence level for new product, new gen products win. Is it coming from pricing, performance, customer relationships? Any -- how do you compare to competitors?

Very good question. So, It's not just a simple question, but again, you asked very good questions. So a, we now have a better product, better performance. B, depending on which product line, right? So for example, IGBT, our key competitor could be the supplier A, On the medium voltage, it could be supplier C, the lower voltage supplier D. So we're actually not competing same competitor on our product portfolio and Power IC also definitely. Power IC we more compete with maybe the Taiwanese, let's say. So -- and in the Power IC case in Korea with domestic suppliers, so there is edge, okay? So that also helped. In terms of the super junction. We now have a new gen product that is better than most of the Tier 1s, except maybe 1 or 2, okay? So that sets our performance level higher. IGBT, Gen 6, now we're becoming competitive. I mean, it's not the best, but we are getting there to hit the traction orders. And then we have a new product that's coming for the new traction motor. That will be very good competitive with the Tier 1. So again, so that's how each of the design will play out and it's in the pipeline of either development or design in pipeline.

So data center, just kind of clicking into that opportunity, are your products more on average closer to the wall in terms of incoming power or closer to the GPUs in terms of onboard power? And the -- just understanding kind of that opportunity there. Would you be targeting U.S. hyperscalers? Would it be OEMs in the Asian supply chain into the data center? How would you be approaching the market from a customer perspective?

Initially, we'll be doing on the OEM on the Taiwan side or the China side and not on the GPU side initially.

Do you have a kind of outlook for the solar market? It seems to be kind of rebounding off a bottom in the past couple of quarters?

Yes. The solar was one of the key segment for us. But I think in 2024, [indiscernible]. So last year, because the slowdown in solar and so forth, our revenue took a hit about 20% from '23. But as I said before, we are designing to and supply to the top 6 solar maker in China. So when they rebound, we will rebound together. And then we will be also getting into more high chain value system like energy storage system and so forth rather than just solar [indiscernible].

The move to the ESS, energy storage systems, is that based on one of these products that hasn't rolled out yet? Or is part of the 27 products today that can applies there?

So the it's out of those 27 products we introduced today will be enabling that market.

A question on next 3 years CapEx. So can you confirm that the strategic decision around display whether or not that's a funding source for [indiscernible].

Any further questions?

I have a question on of the fab capacity and how much that can support for power revenue. So your -- I think 2021 was a peak of power at close to $270 million.

We didn't get to $270 million [indiscernible].

In the fab, including foundry -- how much was foundry back then [indiscernible].

[indiscernible] like $40 million. So we used to have the $40 million per year.

Yes. So on an apple-to-apple comparison, then you'll be comparing the $300 million in '22 to $240 million in '21.

I mean probably not on that just one-on-one match because foundry has a different number of layers compared to the foundry and the power products. So -- and also, we are not really producing the legacy existing product in our Gumi Fab. So maybe that revenue should not be translated into exactly like if we do the $40 million more in our Gumi Fab, are we going to get to the same level of the margin expansion. That's not 100% true. I mean, obviously, we have to achieve certain level of revenue increase because we are losing $40 million of the foundry. But at the same time, it's going to be the combination of like what kind of product you're producing in Gumi Fab, like a new product, new generation product with the higher ASP, higher revenue per wafer. So it's going to be a combination of all of those with the fixed cost component is different now versus 2021 or 2022. But I mean, so that year, probably not the direct comparison, but you see our kind of target model like here '26 and '27 in 3 years' time, and there's kind of a linear improvement on the gross margin and the revenue. And so you're probably going to get not too far away from what you're going to see, but it's probably not one more [indiscernible] direct.

Yes, those '21 and first half '22 is huge shortage period with a good ASP uplift. And we are not counting any shorter cycle with the ASP of [indiscernible] today's forecast.

Will you be in a position at some point as the new products ramp with a higher margin that you'll be able to kind of -- what percent of revenue today is kind of commodity competitive legacy revenue? And will you be able to start saying no to lower margin, lower price? When will that phenomenon start to happen? What will become?

Yes. I think today, about 20% to 25% is commodity type of product. So I think the us, we increase the capacity for the new generation and we've start to fill the fab and you don't book the fab filler products as much. That's the plan.

Maybe end of '25, early '26 is when that would start...

I think that's when we'll probably to start to book those kind of it...

Any kind of lead customer for these Phase III next-generation products? I think you mentioned something like a AI IT OEM .

Yes. So the -- our Gen 7 and Gen 8 low-voltage products already inside the latest AI phone in Korea. The last generation, this generation as well as foldable and so forth. So we are already proven there, and that's one of the key reasons we're going to ramp up on the communications sector. And the other thing is our Power IC all the OLED IT panels, which go into the laptops and so forth. And then our next-gen super junction and the [ MD gen 8 ] will be the one that we will be getting into the motherboard and systems for the server type of applications.

So on the IT OLED exposure with Samsung's Gen 8 fabs coming '26, do you think that could be an opportunity for you?

I don't understand.

So Samsung has the Gen 6 OLED fabs coming online in '26. Presumably, do you think that could be an opportunity for you?

The careful [indiscernible] because it maybe -- but anyway, generally speaking, we are the currently #1 share in the OLED IT today. So our job is to at least maintain the share, then if it goes up, then we can benefit.

So that covers both laptop and tablets right?

Laptop and tablets.

What are the applications for the ESS , what are the applications that you think will drive growth there? Where are you seeing deployments?

Our main ESS market is with China.

Is it for infrastructure for a data center, for what the use cases are?

Infrastructure, data center and main industry. Industrial and so on in the main markets. But thank you for coming today, and much appreciated. Thank you. This ends our session.

Thank you. This concludes the conference. Thank you for your participation. You may now disconnect.