All right. Thank you, everybody. I'm Joe Moore from Morgan Stanley Semiconductor Research. Very happy to have with us today, CEO of Texas Instruments, Haviv Ilan. For the first time, actually, I think, at our conference, if I'm not mistaken.
Correct.
So I really appreciate you coming. So maybe if we could just start with a bigger picture of you. You've been in the CEO role now since, I think, April.
Yeah.
It doesn't seem like the corporate priorities have shifted all that much, but maybe you can just kind of give us a general perspective on how you feel after almost a year on the job in this role.
Thanks. Good morning, and thanks for having us, Joe. Yeah, you're right, priorities have not shifted. From a very high level, our ambition is to continue to maximize the growth of our free cash flow per share. And we do that through the strengthening of our competitive advantages, our business model, our capital allocation, and always driving for efficiency, and that will continue this year and beyond. We are in a big phase of execution of a plan we laid out a couple of years ago, and I'm very pleased with where we are right now.
Mm-hmm. Okay. Yeah, I think most people are aware of what you're doing, but, you know, a fairly ambitious plan to increase CapEx, to have a much bigger footprint in the United States. Can you talk about the philosophy underlying that? Obviously, these are very long-lived fabs, and we understand a lot of the incentives that you have. But, you know, the focus on US, specifically, you know, you do reduce the dependence on Taiwanese foundries, but, you know, you saw other people are sort of building fabs in Europe and geographically diversifying in different ways. So, you know, generally, just the strategy of expanding so much in the US.
Yeah, before we maybe talk about investment in capacity, let's talk about the why. What we see is an opportunity, and for us, when we see the opportunity, our job is to prepare for it, and that includes that investment. But why, why is the opportunity so attractive? First, we believe and see evidence that the secular growth of semiconductors, especially in industrial, in automotive, is going to continue at least for the next decade and maybe decades to come. We are pleased with our position. We worked very hard in the last 10, 15 years to get to where we are today. We finished 2023 with close to 75% of our revenue in industrial and automotive.
That revenue grew at a 10% CAGR between 2013 and 2023, and that didn't happen by accident. That's us, you know, diverting our resources, our capital towards these opportunities, expanding our product portfolio, getting more acquainted with customers. We are making invested in the channel to be in that position. So we are now at 75% of a business of two markets that are going to continue to grow. And you add to that, you know, the customer preference, and that's more of the last two, three years of what we call this geopolitical dependable capacity that I'll talk about in a minute, which is part of our investment. So that's what drives our ambition to make the investment. The opportunity is there.
When you think about where we want to invest, we always wanted to build our capacity internally and base our growth on our factories. Mainly because it's, we always said, it's better control and lower cost. The lower cost part, especially with 300mm wafer fab, everybody always understood. I think the control piece is more discussed these days, and not only in these type of conferences, also with our customers. Our customers are more aware as geopolitical tensions are rising, especially between the U.S. and China. There is a more desire, and that's a very high level management levels of our customers, to get a dependable capacity, and this is why we are making the investment, specifically in the U.S., for us.
Now, I believe customers don't want or don't need domestic capacity, but rather dependable.
Mm-hmm.
And our plan, I believe, is very dependable. Investments in Texas and the U.S. for the front end, for the fabs. A big investment that we announced last year in Malaysia, and together with the Philippines, a very good footprint of back end. And as we go and make this investment, by the end of the decade, we will be able to support a higher level of revenue. We quoted a revenue capability of $45 billion. That will allow us to grow at a 10% CAGR versus 2022. We said that more than 90% of our wafers will come from internal fabs, and more than 90% of our assembly and test, or our packages and tests, will come from internal manufacturing.
Which is a great position to be in and very, very important to our customers. Last but not least, why U.S.? Look, we have a footprint in Texas. We got a footprint through the acquisition of a Micron fab in Utah. We love the team, and that's part of the IP of the company. So it's much easier to grow footprint where you already exist, rather than putting a new dot on the map as a greenfield. That's very, very difficult to do.
Great. Thank you for that. I, I think, you know, as you talk to customers, and I know you talk to a lot in the industrial and automotive domain, they're actually really focused on this issue of dependable capacity, and I think where, where a lot of the press attention is on the cutting edge and, you know, what are we gonna do about, about GenAI, around advanced geometries, things like that. But the customers who are actually voicing concerns around this are gonna be automotive and industrial. So, you know, I guess as you've been out talking to people in the automotive and industrial world, how's that message going? Are you getting, d o you think that's turning already into additional commitments from your customers? And just how is the marketing of this idea to your, to your most important customers?
Yeah, maybe let me take two points here, Joe. First, to your first point, you're right. There is a lot of investment in the, that leading edge capacity, you know, FinFET and beyond EUVs. And that as the high performance compute, you know, market moves to the next node, traditionally, many years ago, we could have reused the old capacity towards, you know, our type of parts, the analog and embedded products. That point of time is gone. I mean, due to physics, we can take a 10 nm process, FinFET process, and apply it into a power management device. This is why you see that as people call, you know, legacy capacity, I call it foundational chip for analog and embedded in industrial and automotive.
You need to kind of do ground-up investment. You can't reuse old fabs. So that's also part of the change we are seeing, and that's why we are investing, including the foundries. Now, regarding customers, yes, at the CEO level, which was, you know, traditionally, if you think about OEMs in automotive and even industrial, at our level, we sell an average unit price of a part at less than $0.50, you know, about $0.40 or so. The discussions would not be at that level. They are right now.
It's becoming more and more strategic, and we are getting that direction and help from CEOs, from CPOs, to direct their R&D teams, their engineering teams, their purchasing team, to put, you know, dependable capacity on the board, and this is where we are getting an advantage. Some of these are translated into mutual agreements, and, if you think about automotive, these days, we are talking about, you know, SOP or ship of a product in 2028, 2029. So there is a lot of forward-looking in the type of markets that we serve, and I'm just very encouraged by the progress we are seeing there.
So that's part of the reason you saw us taking our capacity plan to a higher level, as we are convinced that that opportunity is gonna present itself later this decade.
Yeah, that makes a lot of sense. That's good to hear. So as you think about this kind of build internally versus foundry, it sounds like part of the math there is just that foundry is gonna have to get more expensive, 'cause that's gonna have to be fabs that are being purpose-built for these nodes, as opposed to, as you said, aging from older nodes. So you think maybe foundry is not as positive an economic strategy, or is there just more variables in that you wanna deal with?
Look, we have a high belief that this type of capacity that is kind of getting built right now, not from
Yeah. Used equipment
Yeah
Is gonna be more and more valuable.
Yeah.
That's our assumption.
Yeah.
And you kind of even see it with the foundries. I mean, they are making similar investments in new capacity, and, you know, they don't do it for free. If you check the trajectory of their margins, you know, they get paid for that work, rightfully so. So this is where the cost advantage for the company is real, as we do it internally. Of course, when I push my team and I think about our manufacturing team, I don't compare the wafer price for the foundry, I compare the cost, so I can analyze the cost of what the competition does, and we need to be at least as good in what we do. The second point is dependability.
If you look at where these foundries are putting, the capacity, you know, there are some investments in Japan, there are some investments in Europe and the U.S., but if you look at the scale, and we talked about it in the capital management call, last month, the scale is really mainly in China. If you look at the tonnage-
Mm-hmm
Okay, or the number of fabs in a more massive way, it's not, it's not in, in my view, a good answer to many of our customers. So that's where I think we can have a good competitive advantage with our strategy.
Great. And then, you know, the cost benefit to moving internal is pretty clear. You talked about 300 millimeter. You talked about generally, you know, you guys are very good manufacturers, very low-cost manufacturers. At the same time, there is a utilization input to the cost structure, and, I mean, you're underutilized now. You know, if you know, you look at the PP&Es up, like, 60% in 6 quarters, and your revenues are down quite a bit. So it seems like you might have underutilization for a while. Like, how do you feel about that? Like, do you feel like, in some ways, that's a trade-off that you're willing to make for the dependability of the capacity that you have? And, you know, any thoughts of sort of cyclically adjusting those spending plans?
Yeah, you know, the short answer is yes, meaning we are gonna keep a steady hand making our investment through the cycle, simply because there is no other way, in my opinion, to invest for the long term. The cycles are short-lived. You know, building a factory then ramping it to full production takes years.
Mm-hmm.
You always wanna be, you know, I will quote Rich, that used to say, "I prefer to be two years earlier than two months late." Which, that's the way we look at it. So we are gonna go through our investment, and I think we were very clear that through 2026, we wanna put the company in the position, that then we can adjust or build according to what revenue and growth will be. And you don't build half a fab, you have to go through the investment, and that's what we are doing. So you can count on us to continue with our investment plan. We are. This is, like, really a six year plan. We started in 2021 with the acquisition of the fab in Utah.
We are three years in. Now we are in the fourth year, and, you know, kind of almost three to go. I'm very, again, pleased with the execution. These factories are ramping.
Yeah.
They are showing the cost advantage already. As the market recovers, and it will, because it always does, I think the folks will be very attractive.
And you do have some of the longest-lived fabs in the industry, and actually, some of the shortest depreciation times in the industry, 'cause all the memory and cutting-edge guys are now depreciating over longer phases. So when the depreciation's gone, these fabs still have decades of useful life.
I mean, we have just announced, I think a couple of years ago, we have one of our 150 mm fabs in Sherman. This is where we're actually gonna build a mega site in Sherman. It worked for like 50 years, you know, and these products are still running, they are still selling. We are just taking them into more modern capacity right now.
Yeah.
That's all we're doing. But the tail of that revenue and margin is where the secret of our market is. Yeah.
So you had 45 years of depreciation free. Can you talk about the CHIPS Act as a factor here? And, you know, we already know you're getting the tax credit benefits back, but there's also grant money to be delivered.
Yeah.
You're really doing something that's really important to a lot of the big U.S. customers for semiconductors. At the same time, you know, a very profitable, cash-generative company. You know, how do you sort of look at the CHIPS Act as being an influence on all of this?
First, it is, you know, it is a factor in our decision metrics, you know, analysis. And we actually were very open about it. In 2022, we put a plan out there that had a more modest capacity investment, and then CHIPS came out, and we said, "Hey, that's a good time to invest in America." I think it's historic. I think it's very, very important. I think it puts us in a level playing field with the competition. We just talked about the foundries and where they are putting the capacity. I think TI is a great answer. So that kind of informs our decision, and we comprehended them in and taking the capacity to a higher level.
We just announced Lehi 2 , a year ago, and that's a very big investment in Utah, and of course, I think we are able to do that with the anticipated support of CHIPS. So again, we have, we are getting already the benefit of ITC, or we will get it actually, the cash will come actually later this year. And the grants are in process. We think we are an important player in the resiliency of the U.S. economy and also the national security. So we hope to hear good news here.
Yeah. Okay, great. Maybe shifting to the business conditions. You know, you've had a tough correction here. You know, this will be your sixth quarter, I think, of year-on-year declines, and I know you guys have sort of talked about that. That's kind of an average length of a downturn. Is this, you know, dying of old age at this point? And can you, y ou know, it does seem like you see significant inventory reduction at a lot of your customers, but you have areas like autos, where that's probably been less clear. So can you just talk about where we are in this correction?
Yeah. You know, it's been talked for a while now. This cycle is a little special because of the way it went through the market, right? Almost asynchronous way, PE first, automotive last. That's the way I think
Yeah
About it. PE started to correct back, and automotive just started to decline. I mean, we saw after three and a half years, I think, of sequential growth, every quarter, we saw Q4, I think, decline in single digits. And again, that's because of what you said. Customers, at a certain point of time, even if they want to keep the inventory levels at a high level, once they build them, or they stop building inventory levels, that revenue rate decreases on us, right? Because they have enough. And I think we got to that point also on the automotive side, but across all sectors, in industrial as well. That will, as usual, correct itself, and again, our focus is on the opportunity ahead. I think it's still very large. Secular growth is still there.
We can look at the boards, at the platforms, at the architectures of these EVs and these robots and these test and measurement systems towards the end of the decade, and they're exciting. There is more, you know, more parts on the boards. And, you know, the cycle is a cycle, it will play out. Our focus is getting prepared to what's behind it.
Yeah, I mean, we've looked at some of the numbers and, and, you know, revenue has declined. It's the third worst peak-to-trough revenue decline in the 30 years I've been doing this. The two that were worse were 2001 and 2009, which was much different, right? Demand destruction on a large scale. So it's maybe the worst inventory correction that I've seen. At the same time, the shortages were also the most intense that I've seen. I mean, the most. You know, I was spending half my days talking to auto analysts, like, "When are we gonna get these semiconductors?" You know? So the shortage was very severe. I guess, not shocking that the correction was pretty severe. But my perception, at least, is that the underlying demand is still, you know, reasonably solid.
Can you, can you talk to that? And, you know, do you have visibility on where those inventory levels are at this point?
Yeah. I mean, I just. As I gave an example before, and I can give a more specific example of even on EVs, you know. Think about EVs, there is a little bit of, "Hey, we were a little bit ahead of ourselves." There is a little bit of a slowdown and, you know, so maybe they're not going to grow as fast in 2024 versus 2023. But when I talk with the leaders of these OEMs, their engineering managers, all the R&D is going there, okay? So all the new system, new designs, new architectures are towards that, you know, EV vehicle. So maybe the adoption is gonna be a little slower, but it's not gonna go away.
Mm-hmm.
This is why I think these secular growth are persistent. They are not, they are not changing, and that's where the conviction comes from. I'll give you one more example. Even in our factories, you know, assembly and test, we are taking this internally, we are building it more modern, modernly. We are putting them, for example, in Malaysia. We can actually increase our output by 50% in the last three, four years, while taking our labor down. That's all thanks to automation, AGVs, robotic arms, and more and more robots in a factory. This thing is not going away. The return on this investment is usually less than a couple of years. So these are strong secular growth elements that I think maybe are tactically being slowed down with in-
Yeah
Inventory digestion. But, I can go on, on and on, on sectors like renewable energies and others, where, if you look forward, that opportunity is not going away. It's actually stronger than ever, in my, in my opinion.
If you look at, I mean, you guys started focusing on automotive as a major priority over a decade ago, when it felt like, you know, internal combustion cars without dramatic shifts, a very good market, very stable, you know, a good TI kind of market. But, you know, the evolution of this around EV and ADAS, and those kinds of opportunities, has clearly caused a more, you know, steeper growth curve around that. But it's also one where I feel like your customers are trying to figure out where they wanna participate. We've seen slowdowns in ADAS, for sure, relative to the plans of four or five years ago. EV, like you said, there's some bumps in the road. Our auto team is actually quite negative about some of the programs, just how much is being invested in EV programs.
At the same time, it feels like, you know, you probably don't want to be building internal combustion vehicles in 10 years as your only line of business. So, you know, I, I guess, put a lot in perspective, do you see any, you know, i s your five year view really any different from, you know, given what we're seeing this year?
Again, as we said, just a minute ago, the short answer is no.
Yeah.
We have a strong view. By the way, in our case, if you think about our automotive business, in the last 10 years, it grew mainly because of ICE, right?
Yeah.
This is just the body, the lighting system, the infotainment, and the ADAS are unrelated to EV, non-EV. There is just more content in a car. We can see it.
Yeah.
Every generation of car we buy, we can see it in the screens, we can see it in the speakers, we can see it in the moving seats, in the windows, et cetera, et cetera. So that is, has driven the growth in the last 10 years.
Yeah.
And I think that growth will only accelerate as EVs are getting adopted. Now, maybe the S-curve is not as steep, but it's gonna go through it. So in that sense, Joe, again, I'm very optimistic about what that market is gonna do for our industry, for semiconductors, for automotive.
How does it inform, you know, your product planning when you sort of say you're focused on automotive and industrial? You know, for a company that does, you know, more catalog business than anything else, you're sort of thinking about what are the building blocks that we're gonna need, but at the same time, you're seeing these major tectonic shifts in the vertical market, where you're gonna need stuff that's dedicated to that, and they're big enough markets that you need to be dedicated. So how do you think about catalog versus more vertical orientation?
And I call it even on a in automotive, and it's actually true in every market. I don't call it vertical; it's more I call it application-specific. Meaning, when you start to put a more, a higher level of integration in a part, solving a more specific problem, now you could call it more vertical. A radar chip for automotive is very, very different than a radar chip for industrial. So you need to have specialty there. And TI does both. I know people always talk about us that we have catalog, but actually, TI is very strong in both. We usually don't like to mention, and I actually don't like to do it myself, because my team then, you know, concentrates on one thing. I would like my team to build the breadth of the portfolio. And if you look at the automotive.
You know, if I look at automotive, as many end equipment, the amount of boards over there, I think we have more than 1,000 different sockets that we can address today in automotive, and that's growing.
Mm-hmm.
More and more of the architectures of the car is changing, and I want attention to each and every one of them. I want the, as you call it, the catalog, call it building blocks, as, as, you know, the glue logic, as I call it, or the glue analog logic around the the big processors. I also want a very integrated connectivity solutions, I want integrated radar solutions, I want integrated zonal solution, wherever you want to go, okay? And we need to do all of them. So that's the play of TI. This is why we usually don't call out one part, one chip, one even end equipment in the car. The opportunity is very, very broad, and I think that served us well. It, you know, you don't down select, when the opportunity is so vast.
So that's our, our strategy, and that's how we deploy resources accordingly. We have 60 product lines in TI. I think 95% of them are building parts for automotive in different, very different type of products. Very integrated or vertical, as you call them. I call them ASSPs, or more catalogish amplifiers, converters, power management chips , to fill the many sockets out there in automotive.
And I guess some of the same dynamics in the embedded part of the business, where, you know, you've had a couple of years now of revenue under performance in embedded, but underlying that, there seems like there's a shift towards more programmable microcontrollers. We actually did some survey data in China, where end customers said TI was best in class at programmable MCUs. So, like, there's clearly an underlying story here that the kind of core business that you're focused on is doing pretty well while you're kind of divesting some of the other businesses. How do you think about that embedded?
Yeah, and embedded, if you go. I would say in the last 4.5 years ago, we made a change in the strategy. We've adopted it, and we said: Hey, why, what, where do we want to be great in embedded? What should be our strength? And we said, let's align our product portfolio to the assets of the company or the competitive advantages, meaning we want a part that can be built predominantly internally, inside the company. We want a high breadth of product. We don't want to play in one or two sockets. We want to play in many sockets. We want to have our channel advantage help us, meaning that we can sell this part to many, many customers.
And also, we want to build, you know, long-term positions, in terms of diversity and longevity. So you're seeing us, if you just take MCU or low-power MCU, as an area of, investment. We've strengthened the investment. We put a roadmap together, we've put more people on the job, and we are starting to get the parts out. You're saying, I think we have still—I think we have stabilized that business, but still, the opportunity is immense ahead. You can say the same about, you know, wireless connectivity. You can say the same about real-time control. If you think about our old DSP business, now controlling specific systems in automotive, in industrial, whether it's power management or power conversion, whether it's motor drive. And on the processor side, be more selective.
You know, this is where we will use at the lower end our 28 nm that process that we'll be developing, or 45 nm process that is running today. But you know, some of it will be used with foundries. But you are seeing a shift of the embedded business moving inside the company for manufacturing, and relying more and more on our competitive advantages. The results are starting to come in, Joe, but I, my expectation from the team is much higher. I think embedded could be a great contributor to our free cash flow per share growth in the future.
Okay, great. So I just have a couple other cycle-oriented questions, and then I'll turn to the audience. I guess, you know, maybe inventory price in China, because those are three questions I get a lot. So inventory management, I feel like you guys have done a good job of kind of proactively going to your customers, helping them deal with any inventory reduction, not enforcing any kind of long-term agreements, things like that. Can you talk about how that philosophy has resonated with your customers and how that might help you going forward?
No, I think when customers make decisions now preparing for their next opportunity, they look at, you know, who is a good supplier. And I think what we have put together for them, a combination of, first, get capacity back, you know, ahead of demand. That, and they see it, our investment, they are excited about it, they like where it is, so they can bet on us. The second thing is how we behave in terms of, you know, let's not force inventory on customers that they don't need, and our willingness to carry that inventory, simply because the risk of obsolescence of this inventory is much lower now compared to the TI I remember 10 and 15 years ago. Much more diverse product portfolio, many customers, longevity.
So we can just build it according to what we think demand is going to do. And you do it by part, by product, by this level of diversity, according to the cycle time, that you'll need to catch up once the demand comes in. You want to have inventory to serve your immediate needs. So that's what informs our inventory levels on die bank, on finished goods, and customers can see it. It's not only talk, they can see it in our balance sheet. This is not by accident, this is us making an investment in inventory. So I think that plays for our advantage of being a great supplier. We just want to be a better supplier and, and prove that cycle to cycle, always have the parts you need, even before you know you need them.
Mm-hmm.
That's the idea. We are investing in that. It's not only inventory, it's also the way we connect to customers, the way we modernize the channel, the connectivity between our-
Mm.
customer base. And I think as we do that, and we can prove that cycle to cycle, we can be that great supplier, that, at the end of the day, helps you gain market share.
Okay, thank you for that. And then on the pricing side, sort of two parts. One, you know, if we're underutilized as a sector, you know, do you anticipate there's a more aggressive pricing environment? And I know we haven't historically seen that, but we also didn't see the price increases that maybe we saw the last couple of years. So put that into context. And then there's, not necessarily my view, but there's a view that I hear a lot, that TI is an aggressor, right? That TI is being more aggressive on pricing. And I feel like when that comes up, it's usually tied to some, you know, consumer part that's a lower priority business. But you know, maybe talk to that. You have a lot of factory capacity, you have a low-cost structure on 300 millimeter.
Does that put you in the position where you want to be more price aggressive?
Yeah. First, pricing high level. You know, pricing is, was favorable in the last cycle, in up cycle. And, you know, we said, I think we are expecting it to go back to, from this higher level, to a kind of a low single-digit decline. That's kind of the assumption we are making as we, we prepare for the future. So far, it looks like it's doing more or less that, and, and that's, again, a favorable pricing environment. Now, to the way we compete, you know, even on analog, we are, we are less than 20% market share, right? So we are not setting the market price, but we, we want to compete for the sockets out there across the entire market, not only in investment automotive. We have good products that we've built for personal electronics.
We are continuing to invest in that market. We have a great cost structure, so if the market price to win a socket, is the price to win, that, that's how we will price it.
Mm.
What you're seeing maybe, and that, maybe that's the noise, is that we were short a couple of years ago. I mean-
Yeah.
As the cycle was hitting us and customer demand was so high, we had to make some tough decisions. Luckily, we now have capacity inventory ready to sell the entire boards out there. And TI has a great portfolio, as we said before, from kind of vertical, very complicated, integrated CHIPS to, you know, more catalog-ish, and we want to play and win in all of them. I think our competitive advantages allow us to do that, so that's what, maybe you're hearing. As I said, before, we always were fighting and winning a good, profitable business, and we'll continue to do that in the future. There is no news here from our perspective.
Okay, great. Thank you. And then last question before I turn to the audience is, you know, China, obviously, there's a self-sufficiency program in China as well, where there's very heavy investment in equipment, and presumably with the goal of building things, you know, for domestic consumption. You know, and you, you have a lot of consumption in China as well. I tend to think of you as being somewhat insulated from what those guys might do. But just how do you feel about competition from domestic China over the next decade or so?
Yeah, it's been more visible lately, but, you know, we've been watching the China competition evolving in the last, you know, six years or so. I've been watching every quarter, you know, about 20 competitors out there, local, all fabless, all with a very narrow portfolio, to your, to your point. But how they do, and, and we pay attention. I think it's very, very important to compete, respect the competition, and see what they're doing, analyze their parts, and make sure you're better. I give ourselves a good chance to compete. I prefer, instead of, you know, you know, saying, "Oh, this is, like, not a good market," or just, "Let's go," and, and, and, you know, allow that to, to happen. No, TI has a great portfolio.
As long as we have built the right part, we put it on 300mm wafers, we put it in our latest and greatest packages, cost and form factor, we should be competitive. All these companies are fabless companies, you know. I know there is a lot of investment in the China foundries, but they also don't work for free. The OSATs also don't work for free. Our portfolio is broader. Our access to customers with our channel advantages is there. So I tell my team, there is no reason why we shouldn't be able to grow our position in China. And I believe that. I believe that you want to take the fight to the front rather than wait for these people to grow bigger and bigger and then find them elsewhere in the future.
So that's the strategy of TI. I think it's also, you know, from a culture perspective of competitiveness and result-oriented, that we want to be innovation. It serves us well, because if you can go head-to-head on any socket, even in Shenzhen, for a consumer high, you know, volume end equipment, you can do the same, later on, elsewhere. That's the way I look at it.
Okay, very helpful. Do we have any questions from the audience? All right. If not, we can wrap it up there.
Okay. Thank you, Joe.
Thanks so much. Appreciate it.