SK Chairman Chey Tae-won on SK Hynix Debut, AI Demand and US Plans

[00:00:00] Speaker 1: I think a sensible place to start is why now you've been thinking about a U.S. listing for some time. But why put SK Hynix in front of U.S. [00:00:09] Speaker 2: investors now. Well we've been waiting for that for a long long time. And since I acquired that Hynix about 15 years ago. So this is one of the milestones. Well we have to go to access to the U.S. capital market. So. Now dream comes true. So people asking why now. But well this is kind of earliest moment. Well we can we can have the earliest moment. Yeah. Twenty six point five [00:00:43] Speaker 1: billion dollars is a lot of money. But in the context of SK's commitment in domestically in Korea and internationally not so big. Can we expect multiple returns to A.D.R.'s. The sale of multiple rounds of A.D.R.'s. Well. Yes. [00:01:04] Speaker 2: I expect we can we can deliver those things too. So we try to expand that our A.D.R. size and also battery. That actually requires a better return. So once we have a better return then there is a more demand. So. So first thing is that. Well we have to keep the stock price stable. And well hopefully in a long run that well we can have that the upside potentials. And I see that the world this moment. So this is a A.I. era. So this A.I. era is actually spiked that a lot of demand for the memory. So this is a kind of time for us. [00:01:47] Speaker 1: Is this something near term. Did you with the leadership group and your CFO say we will need to do this with regularity or you haven't discussed that yet. Well to tell you the truth. Yeah. [00:02:01] Speaker 2: We've been discussing that one. But I cannot just saying that the inner detail yet. [00:02:06] Speaker 1: So I appreciate that. Jamin J. The U.S. Listing gives S.K. A new source of equity capital. But what about the broader U.S. Capital markets. Do we see S.K. Group S.K. Hynix as well become an issuer of global bonds. Global debt. Well that's a possible. But I guess [00:02:29] Speaker 2: right now. Well we have enough cash for that. The S.K. Hynix. Some other company could actually issuing that another the bond and. Yeah. Because that the capital market. Chairman Che. Right. Bear with me. [00:02:44] Speaker 1: I'm going to get this out for our audience around the world. This is why. We're all here. Right. This. You can't even see it. Tiny HBM for a high bandwidth memory. 12 layers in this case. Right. The whole world knows about this. How severe and how tight is the supply right now of HBM. And more importantly. How long is that going to last. Do you have a sort of a year where you see the supply [00:03:17] Speaker 2: situation changing. Well if you really understand the all the requirement and the condition for that HBM memory type and you have to understand that the AI side first. So people know set up a large language model. Yes. Yes. But that large language model as well very important things. But I guess in the future. So the more we're using that AI and the AI actually requires more. The inference phase. Yes. But I see that memory. The mean is a key be caching. So while it generated a lot of key be caching the due to that the while keeping that the performance of the LLM. Then you should have that the world largely on some volume of the key be caching. And you have to actually this is kind of a memory in the end of people that is the same thing. We store that the memory in our brain. And yeah I need that the world those kind of key be caching that memory has to store us somewhere. So if you had at the world better size of the memory than a better performance. So it's a simple is that. And that memory cycle is well it's turning to well as a it's a different different cycle. I shouldn't say that the cycle is a different momentum. So. And back then the old days the memory chip is memory chip demand is actually depends on the number of people number of the hardware set. That was a consumer electronics story in part. Yes. Boom and bust. Yeah. Well mostly well you had that your own smartphone but you don't need that the world tens of different type of the smartphone. You only using that to just one like a PC. Right. So that demand is actually the limited by that the number of the people. But in AI era as a different you may use that the your own AI agent in a probably in a few years from from now that you might have that the world tens over the hundreds difference of your own agent. So that every agent needed a lot of memories was at the world. They created a lot of KB caching. So you have to actually store somewhere that actually needs that the most huge amount of the memory. Well that's the that's the AI the era. So I guess that memory business is turning to that. Well it's a different stage. You still called it a cycle though. [00:06:01] Speaker 1: That's the demand side. A reason to do the 80 hours is to raise capital in part to boost capacity. Your industry peers micron had said very recently they have line of sight to supply improving 2028. The other side of 2027 in part because of the capacity that is coming online. What does SK hynix see. [00:06:27] Speaker 2: Well we didn't talk to that our customers. So where we just announced that a few months ago. So we're going to double up the our capacity. Yes. Within a five years. So that means that what we've been doing that the business more than about 25 years. And then while with the next five years we're going to double up the our whole capacity. Yes. Well. People see that that's unrealistic and probably the oversupply and people worry about. But my customers said that that's not enough. We need it more. So I double up. It's not it's not enough actually. Well they want us that the world cause of a pink well. Yeah. Five times six times. So the more is the better cause that well demand will be exponentially grow in the future. [00:07:21] Speaker 1: Based on that customer demand. What do they see then if they are the the crystal ball the lens to the future. They believe that their level of demand will stay where it is 2029 2030. How long does supply tightness respond to that. [00:07:39] Speaker 2: Well nobody expected. I don't know. Well what's going to happen in next years. But I'm saying that well I have some confidence in the world. The demand will be grow. And our supply capacity never going to catch up. So I don't know how long. But well if. I think it's going to be until that our people our society human society has said there was some settlement to it. A. G. I. I. I. Then. Well I guess that the world that's a kind of a normal lives at the older the cycles. But anyhow that [00:08:20] Speaker 1: until then. Well we we need that a lot of memories. You're with us live on Bloomberg television and Bloomberg radio worldwide. And we're in conversation with SK group chairman J. T. One. And we're talking about a record for SK Hynix in terms of U.S. years. Take me back to 2012. SK acquires Hynix. At the time people questioned that decision. What what did they not see. What did they miss that you saw about the potential of that memory market. [00:08:56] Speaker 2: That's the kind of a bold decision. And it's I think that the. Why made it really risky bet. In a sense of the financial. Because that the companies have really suffer from that deep debt and also has to fighting with its own survival. Because at the wall. Fifteen years ago. That company is under control by the banks. So simply it's a chapter eleven. And. Yes. But in Korean sense. Well we we're doing that the banks is a. It's a workout program. So. But they're doing that more than 10 years in other workout programs. That actually means that nobody really wants that the world. This portfolio. Because that is too risky. Reason why as a. Well memory business is a cyclical. And that actually suffer from that. The uncontrollable cyclical business cycle. And it requires that. Every year multi-billion dollar capex. No matter no matter how. Yeah. So. Even though you don't make that the money. But that that business that require that the machines and expand. And that they do technologies and buying that the new equipment. And they have to spend the huge size up to the capex. That's why. And nobody really wants that those kind of danger business. We you did. But yes. I see a little bit different because that. Well if you see that there were a different angle. This memory manufacturing is most sophisticated the manufacturing business. So if you. Really do it. Then you can do anything. Like the technology. The process. Yes. And also the whole process of that manufacturing as a it actually. Once you make that the little chips. But chip inside that 380 billion transistor in it. Yes. So that's the business. So how are we going to create that those kind of things. And it's a small small size and the silicon things. And we actually put it in the huge size of the world. The the transistors. That's a that's a state of art of the technology. So once you're doing that. And I can manage any things. And well I can actually challenge anything. And also. In digital world to create a lot of memory in the future. Right. So that's all that. Yes. Yes. I just better bet on that. And well people need that the memory. [00:11:48] Speaker 1: Yeah. You talked about managing the process and the idea that you know at that time in 2012 debt was a very big consideration in the market. It was boom and bust cyclical. How much is this. You've been asked about this a lot. But the idea of the long term agreement with customers changed that history. [00:12:09] Speaker 2: Right. But. There is a supply shortage and every customers wants to have to secure that the while they'll want supply. And the cover of their own demand. So they asking us that we need some kind of the long term contracts. It's not suggested by all send the world. They actually require that. [00:12:30] Speaker 1: It's them that's making the proposal not SK. Not SK. So they are more willing. [00:12:35] Speaker 2: Well we also doing love to do it because that the it's not cyclical business anymore. If we had some long term agreement is well even though it's a downturn and we still maintain that our volumes and the certain level of the price. So that actually gives us a different moment. And that actually created another opportunity for we can change it to some our business model. [00:13:00] Speaker 1: even. What about SK's commitments. Where are you going to prioritize your investment. You know domestically versus here in America. Not just with the proceeds from from this U.S. share sale. But but generally with your balance sheet. [00:13:14] Speaker 2: Generally outside of the Hynix and the world most important matter is AI data center business. So we announced that almost at the one trillion dollar investment in the next 10 years and to set up with the AI data centers. My target is about 15 gigawatt in Korea. I guess at the five gigawatt in outside of the Korea. So I'm looking for that this whatever area including U.S. side and if there is a good chance. And the problem is that the one main thing is a power. So power and the price of the electric city is a very important matter. With that that is AI data centers. But once we build that AI data centers and then well you can actually synergy between the memory chip and this infrastructure matter. [00:14:10] Speaker 1: That's that's what I'm doing in the future. We're live on Bloomberg television and on Bloomberg radio and this is Bloomberg tech in conversation with SK group chairman. Che day one. Thank you. You talked about the data center. The issues of the data center. This is an SK group approach. Yes. So you're talking about SK group looking more broadly at the AI stack not just memory. Yes. Where else will you focus your attention. [00:14:36] Speaker 2: Well. Well. The company like SK Telecom and well they're doing that they'll study on a there or the print their resources on their own LLM. Also the applications and the physical AI side. Yes. Robotics. Yes. And the health care side. Well there is a world. There is a bunch of opportunities and the one we've been actually trying to proceed that the world all those business opportunities in the future. [00:15:12] Speaker 1: An issue here in America and in Korea is talent. There is a deficit of skilled workers in the chip industry in America and a company like Micron knows that there is a lot of talented people in Korea. How are you going to manage that. And again your investment plans for here in the USA. [00:15:32] Speaker 2: Well. Well. To secure that bring that more talent is a what what a deal is one thing. One good one option. Yeah options. And the world idea gives us all set. There was some some kind of financial play options like even the stock option then while we could actually using that the ADR. Right. Yeah. Right. So it actually attracted a lot of people's a lot of talent to the world to our group. I see the ADR is also about public perception of SK. Right. That's that's another thing. And the perception is matter. So well also. Well. We have to spend a huge money on a while. The recruiting a lot of good talent. And I'm I don't really care about the talent is. So it could have been at the U.S. Japan. Well they could be anywhere. It's everywhere. So I try to recruiting that much more the talent. And that's the kind of important things. And I'm asking that our CEOs to well your job is to secure that talent in the future. So let's just set the doing. And well we're going to you're going to create that a lot of jobs. And well. I've you announced that more several trillion dollar investment plan for in the future. So we need a lot of talent. Right. Chairman Jay. What is the biggest [00:17:00] Speaker 1: technological advantage that SK has built. Hmm. Well one of the memory side or memory side. Okay. Process. Design. Yes. That. Well you cannot just as the same just one things. But what's the most important thing. Well one team. [00:17:23] Speaker 2: So Hynix. Well I really focused on the we could bring that to just one company one team. So this is a simple business model and memory manufacturing. So we can actually emphasizing that we are just one team. And so I think that the one team made it with a really one big historical achievement now. So that's what I'm saying. And that's the most important matter. Yeah. You can say that the world some. Well what design capacities and the process technology. This is right. But that's all working with TSMC on the base die. Yeah. Which the others do not do. Also. Well we don't actually compete with any customers. That's another. You do not compete with any customers. No. And you feel that's important. Yes. That's important. And that actually gives us a lot of room to actually partnership in the future. So that's how we can actually maintain all those partnerships. Bigger picture. Can you catch up with Samsung. Oh. That's a. I don't want to answer those kind of questions. So I don't. Well my main goal is that the what I have to make the our stakeholder happy. Yes. Well in order to that there is a well certain part is a competition. But I'm not [00:18:52] Speaker 1: really wanted to ahead of what. That's my main goal. That's difficult though chairman because the shareholders love the profit. the customers want more capacity more supply. Those seem to be competing ideals. That's what we call that the stakeholder. So it's not only the [00:19:08] Speaker 2: shareholders. So share all the ones that there are great returns here. Right. And the customer wants that the world cheap product and enough supplies. But well I cannot just satisfy every time every stakeholder. But while we had at the certain balance. So you know between that stakeholder and what I'm really see that. Well I also have to make that the well my members and members of the company. Well they have to be the [00:19:38] Speaker 1: happy. So happy company make the happy result. So that's a simple. I need to say chairman Jay the ADR is one stakeholder. The U.S. investor. The ADR is indicated to open 177 dollars per ADR. Of course the ADRs were priced at 149. The memory market is also evolving. So while supply is tight. There's a there's a very broad look at other technologies. I'm very interested about SK's a hynix's approach to NAND and flash memory. [00:20:11] Speaker 2: And like the role you see that technology playing going forward. Right. So it's a different type of the memory storage. So D.R.M. has their own functions. But that's the instant things. And like what people you just remember the certain numbers. And while if you just prepare that your exam then you had a real good memory on of those certain numbers and the certain target. But you also need that a lot of other memories. And when you was young then you had that some of some memories. But well that memory needs that though not just that in a brain you probably saw that somewhere. But like well A.I. A.I. need that there was a lot of storage. So D.R.M. has said the instant response and the instant the response the type of memories and you should store there. But if you had at the more big size and we call that the cold data. So if there's a cold data then well you could actually store in the yeah. [00:21:21] Speaker 1: And where is where sorry to interrupt you chairman. Where is the innovation on that other side. Oh. [00:21:29] Speaker 2: There's a lot of technologies. And the problem is that the world in an end it actually. Well. Responses are well need that a lot of the time. That means the latency. So in order to shooting that the what that. The latency. Well we need that a lot of special technology on that. And while probably well what we call that the KB caching the key value caching. And while that the cash has to store in the somewhere. So that's a kind of the pyramid. The hierarchy of the world. The whole memory storage systems. And we need that all those evolution of the world those the memory things. [00:22:14] Speaker 1: And SK will invest into that technology. Of course that that's our the main business. [00:22:19] Speaker 2: And while we're going to we're going to make it. And we could actually well deliver some other the business model even not just that the memory producer. So we could actually memory services. The memory as a service in a future that that is another area. So we could actually focus on. So talk a little bit more about memory as a service. Well the basics of how that would work. Well it's like right now is the kind of imaginary things. And idea. Yes. Just idea. So what. But I think that the whole AI trend. The direction is goes there because that the world memory is a kind of real bottleneck and we have to solve it. And some of the specialists actually bring that there was certain type of solution to the each customers and the each application. So it's not just that the world. One solution could actually cover that all those problems as it's not going to happen. But as the influencing market and that influencing market has that they own the characteristic that characteristic. And we have to develop that the certain adjustment for that the stack up to the memories and how to do it. And we need that some special software together. But anyhow that well we could actually serve as a kind of services rather than just produce a commodity. It's an interesting look at the future. We are live on Bloomberg television and radio speaking to SPA group [00:23:55] Speaker 1: chairman Jay. In October when you spoke to Bloomberg Sherry on you said everyone asks me about AI bubble. AI bubble. AI bubble. AI bubble. And you said at the time it's about the cost of AI. The cost of AI is too high. Yeah. Has that changed since October. And what is the [00:24:18] Speaker 2: trajectory of the cost of AI. Yeah. Well you could actually say that the token for a dollar. Token per dollar or dollar per token basis. Yeah. Anyhow we have to lower down that the token cost. But do you see it coming. Yes. Actually well within three years actually token cost is about one fifth or the one tenths. So that's the evolutions. And well there is a lot of players just contribute that those kind of the token token cost is a way down. So even my memory company make them some solution for the how how we can actually lower down that the token cost. That's our challenge. And well our direction is goes there. So that's a kind of ecosystem for the AI. And well right now is cost is too high. But somehow that our technology you know down the road probably five years ten years. And while we actually digest that all those problems and it could actually be that well everybody's AI and everybody could actually using the today on the Asian in the future. So that's that's that's that's the direction I see. But no AI bubble. AI bubble. And well when you see that the AI bubble and this real world and the stock market stock market always that I see that there's some overshooting and well as a too hard to cold. So there might be that the world there is a lot of bubble for that the stock price itself. But that just that doesn't actually reflected the whole real world. But the trend wise is a if it's a long long than two years three years down the low then you could actually reflect that the old the real world. That's what I see. And you can say that the AI bubble. But I see that the world and stock market. But this AI technology is real. So you use use it. And well actually the problem is well this AI thing is that right now is a little immature the imperfect yet. It's the early stage. Of course it's the early stage is a kind of I said that I said that it's only four years old baby. But when you become a the well some college boy and something else. And it has that much more the mature. So it could actually help us that a lot of productivity. And we just rely on all those things. And this is a kind of race. And right now we see that there's some AI tried. Chairman Che. It's a historic day. Yeah. And literally making [00:27:10] Speaker 1: history with the ADR's. But the focus on America. S.K.'s relationship with America. What is your relationship like with the Trump administration. And how are you working with them to bring more of S.K. Hynix to to this country. Well not on [00:27:27] Speaker 2: just S.K. Hynix. And we've been already the S.K. already investing at the more than 35 billion dollars in U.S. side. And not many people have noticed those things. But we do have bio business. We do have the battery business. And the Hynix just announced that the new fab in the Indiana side. And also S.K. Telecom made it a lot of investment in the AI startups. So. We are the investing to 35. But I see that. Well my plan is that the much bigger number. So. Bigger than 35. Yeah. Well much much much bigger than 35. So I'm planning to. And this is a one milestone. And we access that the U.S. capital market. And we have a lot of options. And also we have that the responsibility to keeping that the share to share price. And. Well it's one good thing is that we made it the right investment. And so. We keep it at our own value. China. It's an important market. A.I. Yeah. China. A.I. as well. We have to actually notice that the China approach is a way different than the U.S. side. How so. Hmm. They have a lot of disadvantages on the A.I. side. But anyhow that they put it in a lot of other resources. The number one is that the human force. So. Human capital. Yeah. Human capital. And the number of people number of engineers. So annually. Well they had that stem. The student from that. The major in the stem. And. It's more than other. The million. Student. It's come out. And every year. So. They're looking for the jobs. And mostly they apply to the tech area. When you say tech area. Well probably 50 percent is A.I. side. So. they are A.I. as well as a. Yeah. They using that the different stretch. So. I don't see that the. Well. Any. Any kind of. Well. The visible. Yes. But there is some signs. And they making that. Some. The cheaper way. And. They try to create it. The world cheaper token. The token cost comparability. Yes. Yes. That's the kind of important matter. And. The U.S. focus on a world top niche market. And. Well. The quality of tokens. But. China is a focus on other ways. So. I don't know the result. But. Well. I see that. What if. They're using that. The using the different strategy.