Okay. We can start the clock. Awesome. Great. Thank you all for coming. I'm actually glad to see so many people here at 11:00 A.M. on a Friday. That's great. I'm Stacey Razga. I cover the US semiconductor and semicap space here at Bernstein. And it's my honor to introduce our guest here today, Haviv Ilan, the President and CEO of Texas Instruments. Before I start, I want to mention if you have questions you'd like to ask during the presentation, you should have a link to the pigeonhole forum. You can submit them, and we will have time for Q&A at the end. TI, you know, TI used to be thought of as sort of the boring semiconductor. I think you were proud of it, actually. They sort of wore that with pride.
It has been a little less boring, I think, lately as they have embarked on a program of significant capacity expansion here in the U.S. that at least temporarily has sidelined some of the cash flow and return that was always part of the thesis. Look, TI, they always think long-term. Now we are kind of approaching the tail end of that investment strategy. The company believes those investment seeds that they have planted will be growing into even more cash and cash flow in the years ahead with a footprint that may leave them very well advantaged in a world that is becoming increasingly decoupled, which I am sure we will be talking all about today. To tell us all about it, it is my great pleasure to welcome Haviv.
Thank you so much for being with us today.
Thanks. Thanks, Stacey. Thanks for having us.
It is a pleasure to see so many people in New York on a Friday in person.
Absolutely. This is fantastic.
Thank you.
You know, maybe just to, just to start there, right? You know, you're into what? Year four of a five-year CapEx cycle, and it looks like.
Year five, but who is counting?
Year five. Okay.
We're counting in 2021. Yeah.
Coming to the end of it. I guess maybe you could talk about, like, just at a high level, maybe for the folks who may not be aware, why did you do what you did, and then how has it gone? Like, what's it supposed to bring for you, and how has it gone versus your expectations, and what should we expect, like, now that we're kind of maybe coming toward the end of it now?
First, thanks for the question, Stacey. Again, you're right. We are at the tail end of six years investment cycle started in 2021 with our acquisition of the Lehi Fab from Micron. Since then, embarked on building a new mega site in Sherman, Texas. We are four and a half years in, right? Very much, 70% in. I'm excited about where we are now. The reason, of course, is to allow ourselves to support a future opportunity. We thought it was the right time for several reasons. I'll start with the fact that, you know, when you look at, in general, semiconductor adoption across markets, we see an acceleration. We call it secular growth of semiconductors are added to more end equipments, especially in industrial and automotive. That is something that we've seen in the previous decade.
I think we think this decade, it, it's accelerating. It's very easy to see when you think about your car today versus your car three years ago and your car ten years ago. You can see the content coming in. If you go into.
Switch to industrial and auto is not new, though. That's, that's been part of the strategy.
Correct. Correct. I think the secular growth is accelerating. We look at this decade with EVs coming in, with robotics, seeing a lot of adoption, factory automation. I think these trends also accelerated by COVID. We saw all the difficulties with labor during COVID, the issues around energy. We are seeing acceleration. We say, "Hey, these markets are going to be probably growing instead of maybe the markets beat single digits. Maybe these markets can grow together faster." The second thing is we have done a lot of work to be better exposed to the market. You are right, the last decade we said, "Hey, these are gonna be fast-growing markets," but our exposure was very low. We had a 40% exposure between these two markets, you know.
Was that low?
That's low. Yeah. That in 2013, actually. The first time we started to look at markets was 2013. That's the most ancient history we have. And it was 40%. We crossed 70% during 2022. Last year, we finished at around 70%. I think it will accelerate this year. In general, we like our position. When you are exposed at a higher level to faster-growing markets, you just wanna make sure you support it. On top of it, you touched upon it in your introduction. We are seeing in the last several years, but more visibly in the last month, that geopolitically dependable capacity is gonna be more important. We've been talking about it for three years. Everybody kind of ignored, I believe, and it came from one ear, went out from the other one.
I think we are now seeing, and I'm seeing it also in discussion with our customers, that capacity coming from a diverse set of manufacturing footprint. In our case, a heavy footprint in the US is unique, and it's going to allow us to support our customers better and maybe pick up some market share. That's the reason for the investment. Please, please be the execution. You know, I'm seeing early evidence that this investment is gonna come to fruition in the coming years.
Got it. Do you wanna remind the audience exactly what you're investing in, like how much, where, how much capacity is gonna be in place, the revenue targets that it can, but I shouldn't say revenue targets, but the supportable revenue, I guess.
The revenue opportunity we could support. Yeah. I touched upon the Lehi. The first fab was an acquisition. As I explained, the excitement seen from customers, we decided to, and we announced it, I think, back in 2023, if I do not, if I am not mistaken, the back end of 2023. We broke ground last year on the Lehi 2 fab. That is gonna be a very large investment.
That was $11 billion?
It's probably gonna be more like 12. Yes.
Oh, okay.
And, it's our first three-level fab. It's gonna go all the way down to 28 nanometer. It could even do another step. And this is going to mainly support our.
It's gonna go to 28?
It's already, we are already having test chips in 28 coming out of this fab. And it can do one more step. It's not gonna do FinFET, but from a bulk transistor, that's where we are going to end there. And it's very good. It's a very good fit for our embedded business. Think about MCUs, DSPs, you know, embedded flash memory on a single die. And also our high-speed, what we call high-speed analog, so high-speed mixed signal connectivity. You can think about some of the parts in analog that need some MCUs embedded inside. It's gonna serve both analog and embedded, but mainly an embedded-centric fab. And that's gonna be a couple of fabs over there. So that's two new 300 mm wafer fabs that are going to support a lot of output.
The other investment, you remember our first RFAB1, it started in 2010. As we remember, during COVID, it got full, unfortunately, and we did not have the clean room in RFAB 2. That fab is now up and running. It is actually utilized at close to 70%. You know, it is connected to RFAB 1. It is not even a new qualification for us. It is just another 300mm wafer fab. Now we have three. A new mega site that I have mentioned before in Sherman, that fab has broken ground in 2022. We are moving wafers today. First products are going to be qualified this year. That is a big thing for us because when you put a new factory, you do have to qualify your parts.
Even if you have them running in RFAB, because it's a new site, there is a qualification phase that we have to go through. A very, very good result. We are happy about the throughput that we're seeing from the fab. It's gonna be actually a, it can support up to four sites. We've built the first two. We've equipped the first one, not the full clean room.
You got two shells.
Two shells, one sub-fab already in, Sherman 1, and some of the equipment to run the pilot line.
Two empty lots, basically.
Two empty lots. By the way, lots we have in Texas. We have options. That is very important because this is where our customers that are now more worried about supply, and especially US supply, are coming over. We are entertaining customers. We had customers coming this week. They see the capacity that we have built. They see the technology that is gonna go there. They are very excited. I will give one example on the technology, Stacey. Analog. It is mainly an analog fab. You know, our next-generation BCD process, 65-nanometer lithography, monolithic power stages, are going to power, for example, data centers and GPUs are gonna come from that. This is a technology that is being ramped right now. This is where we are going to utilize that fab across all of analog, but also embarking a journey into new markets.
Got it. Got it. I guess once, once that's all in the ground, is this, when did you expect it all to be? There was a 2030, or was it maybe it takes longer than that?
No, this is where.
How do you do with the whole thing with support by the time?
Yeah. This is where the revenue support, and we, we've gone through it in the last capital management call, back in August, but another one, refresh in February. At the end of the day, we have flexibility. We finished, this is why we call it a six-year investment plan because when we finish 2026, we are now sitting on, already, five out of the seven built with shells. Now we have clean room that we can build into because we are, we've qualified it. Lehi is already qualified, and Sherman will be qualified this year. Now we have this, what we call this phase three of flexible capacity spend. This is according to revenue. If revenue, now you choose your revenue target, I know that we always talk about what revenue will do.
We'll get there.
Yeah. We'll get there. What revenue will do next year. We have options all the way to above $40 billion in 2030 if we need to. If we do not need to, we'll just slow down our CapEx starting in 2027. Even in 2026, we have some flexibility based on revenue then. We are excited about that because to me, if you want to start right now to support the second half of the decade, you're already too late. It is just not going to be ready.
Mm-hmm.
In that sense, and you think about geopolitical tensions, we are well prepared.
Mm-hmm.
For many scenarios, one of them is an upside scenario that no one wants to talk about, but also on a downside scenario. Okay.
Let's talk about it. You've laid out some of those 2026, and I won't call them targets, rather than scenarios.
Correct. Framework. Yeah.
Framework. Right. And I think it was $20 billion-$26 billion, right?
Correct.
I'll be honest, I'm below the low end of that.
I know. Yeah. I've noticed.
Seventeen or something, right? Fine. We can have the argument. I think the CapEx range was two to five depending on where.
Correct.
Now, it's, it's hard to cut it to think in 2026 or below two. I mean, some of that was you wanna get some of the Chips Act projects started so you can get some of that. I, I what is the tax credit is and ends and the project has to start by 2026 or something, or?
No, no. It's actually not related to ITC. It's related to clean rooms and shells because you can't build half a shell, right? When you think about Lehi 2, it's a big investment. When you build that shell, you have to go through it. We wanna complete it by 2026 to have that flexibility. That's what drives the spend over there. It's also the qualification of Sherman 1 and the pilot line. We wanna get that completed. That's what drives the cost. From an ITC perspective, that's not, I mean.
Okay.
We are gonna be running, we've started this project so that ITC will run into the future. Now, regarding the revenue, you've mentioned the framework. Yes, you know, we are, I'm not, you know, we give you one quarter at a time.
Yeah.
In OTI for years, and we are not gonna change there. You know, with shareholders, they wanna model, in our case, free cash flow per share by, you know, the revenue scenario. We have given four examples.
Mm-hmm.
You are insisting that we need a fifth one. I don't think we do.
I'm not insisting anything. Like, it's.
But in.
Yeah.
Just let's look at what happens right now. We are coming off, probably the largest down cycle ever, which will argue that it competes with the 2000 timeframe.
Yeah.
The economy was in a different state then.
Yeah.
We are seeing a very nice recovery coming in. We are, you know, just Stacey, let me run one number with you, and you can do the math later on, on your what, what 2026 will do. Thinking about the first half at the midpoint of Q2, at the midpoint of Q2, we are talking about 13% first half growth. I think it accelerates every quarter.
Mm-hmm. Like year- over- year.
You know, when we finish, you do the math, you are you gonna have a good chance to get to that number that maybe you forecast for 2026 already in 2025? You have a good chance. Let it play out. We have the responsibility, and this is where we fell behind in the previous cycle, to be ready.
Yeah.
Because it's very easy to say, "Oh, it's never gonna come." It's easy. You don't need to invest. The hard decision is to make the investment so if it comes, you can support it.
Yeah.
I am very proud of our steady hand on a strategic perspective, but also on our execution. The team has executed well. Some of these fabs are built into the biggest shortage of supply ever, not only on semis.
Yeah.
On steel and, you know, high power stations and water treatment equipment. And you had some semi-cab coming in today, thinking about two years ago, and we executed on time. RFAB 2 is on time. LFab 1 is on time. Sherman 1 is actually pulled in probably a quarter. We are very excited about our position. You know, let's let the second half of 2025 play out, and then we can talk about 2026.
All right. Let's talk about that sort of cyclical recovery. You're right. We were, I don't know what it was, eight quarters or 10 quarters of year-.
10.
Particularly in particular.
10.
Particularly in industrial.
Yeah.
It does seem that the, are you seeing recovery everywhere right now, or is it primarily industrial?
The short answer is everywhere in industrial. The other markets have already done it, meaning PE. PE is running very hard, from a year-over-year perspective. I think more than 50%, if I remember well. And comms and enterprise similarly. They just kind of came after PE or, or consumer.
Yeah.
I'll talk about other in a minute, but industrial is a big chunk of our revenue. And it's showing a very broad recovery. Every sector, every geography, every channel, including ti.com. Very good recovery. Started in Q4, actually. We were very, we say, "Hey, it came across probably five points higher than our expectation.
Mm-hmm.
Let's wait another quarter". It happened very nicely in Q1. If you think about our forecast for the 7% sequential growth in Q2 this quarter right now, and we are now two months in, I can just reiterate that we are, it's, it's, you know, industrial continues. We are not very surprised.
Mm-hmm.
Industrial was the fast grower sequentially in Q1, and I think it's gonna continue to do that. Very confident about industrial. It's about time. We are way, way below trend line.
Yeah.
The second thing on automotive, you know, that cycle almost like was quelched. You know, it was a single-digit year-over-year decline at the peak of it.
Yeah.
It is already grown in Q1. It is, I think.
Is it all China though, or is it more than China?
No, it's not only China. I mean, China, of course, is very helpful because there is a lot of EV adoption in China.
Yeah.
You have more content per vehicle. You know, I think, other than Japan and Europe that came in late to the cycle, I think it's looking good. I think the shallow correction, I would say, in automotive, is maybe behind us. Yeah.
Got it. Got it. It felt like you were starting to take utilizations down in Q4 though, even as you were seeing the recovery. Do I have that wrong, or is that what it was?
No, I think Rafael mentioned it in the Q, you know, when we had the January call, I think he mentioned it. In April last month, he said, "Hey, we had an upside quarter, right? And we also look at the orders into Q2." This is all before tariff and everything, right? This is like.
We'll talk about that.
Life is good. We said, "Okay, we better keep these fabs running at a little bit higher utilization." And, you know, that's what we've done. We have to prepare. We can't be in the mode that happened in the previous cycle that you once you fall behind.
Yeah.
It's very tough to catch up.
I totally get that. I think you guys are in a good position in the sense that your parts, as you've said very many times, do not really go obsolete. You are holding lots of inventory, more than you used to hold in the past, much more, much more than most of your peers. To the extent that your peers are holding this much inventory, it is involuntary on their part. With you guys, it is deliberate.
It is deliberate.
Yeah.
Just by the way, we talked about it a minute ago, but think about our business. It used to be 40% industrial and automotive. Now it's 70%-75%. At the end of the day, this is where the low risk is.
Mm-hmm.
These are, there's a, there's diversity and longevity phenomena there that you just, you take no risk. I mean, there is cost of capital, of course, but it's an easier decision. When you're in PE, you, you know, we call it, we don't like our bananas to turn brown. Okay? Because PE moves very fast. If you build it.
Yeah.
If you build it too quickly, you can scrap it. The other thing that, by the way, you know that when you move stuff more internally, you just have more work in progress, inventory, versus, you know, think about us building in Lehi versus, you know, TSMC or something. That also adds to a little bit higher level of inventory. We are well prepared, yes.
Got it. I guess with that, you know, some people would suggest that you guys have a better view of actual demand because you do not really run a backlog model, right? You have, I do not know. It is probably pretty close to 100% of availability of 100% of your products right now. Lead times are probably zero, right? If I go on ti.com right now, could I, I could.
ti.com is zero.
Yeah.
This is, by the way, this is why we see more, we talked about turns.
Yeah.
When you really need it now, you go to ti.com.
Yeah.
Lead times are low, and they are very competitive, probably the best in the industry. You're right. We can't count. Look, semiconductors now are used by so many customers. Some of them, they just don't know how to plan it. They also don't wanna take the risk.
Yeah.
Especially now when cost of capital and cost of inventory is higher with interest rates. We decided we are gonna replenish our inventory based on real consumption. Of course, we'll give you support according to our lead times.
Yeah.
We wanna be prepared for any cycle. We modeled the previous cycle, and this time we don't wanna hit a capacity wall.
Yeah.
Or an inventory wall like we hit last time.
Yeah.
That is part of the strategy.
Is it fair to say that if you can fulfill an order, you ship it, or is that?
Yes. Again, you know, if you forgot to order, you know, I would fulfill it from, I'll fulfill it from ti.com. If you.
Yeah.
If you are respecting our lead time, which are the most competitive in the industry, we'll ship it to you. I think you asked me last time even, you know, is the order coming from, I don't know, Pool Inn or just, I don't know.
Yeah.
You know? So we just fulfill it.
Do you care? I mean.
I don't care.
Yeah.
Of course, I do look at the situation of, and, you know, I'm an engineer, so I like to think about signal and signal-to-noise ratio. I think the signal is a cyclical recovery. There is noise.
Yeah.
Okay? Typically, when you think about a system and you inject noise into the system, you just accelerate the cyclical trend here.
Yeah.
I think if customers were really driving inventory to zero levels.
Yeah.
I think they are thinking about maybe behaving differently right now. Plus, you know, we talked about industrial. They have depleted their inventory. And the beautiful thing about industrial, when you move to your next generation machine and equipment, sensor, whatever, there is just more content.
Yeah.
You see this double acceleration of I'm stopped depleting, I stopped depleting my inventory. I also have my new system that has more content.
Yeah.
By the way, we also hopefully want more share over there.
Yeah. Got it. You know, let's talk about tariffs and geopolitics and decoupling and everything. I guess just the levels that, like, yeah, seriously.
We can spend hours on that, right? I'm just checking time. Okay.
Just the levels, I mean, what are you seeing on tariffs? Or I guess both direct and indirect. My guess is right now not much that you can measure.
Right now, as we know, there was a little bit of chaos in April.
Yeah.
This is why when I talked on the call, Q1, including in China and the U.S., we haven't seen any tariff-related.
Mm-hmm.
You asked me how do you know. We don't know, but that's our assessment.
Mm-hmm.
Just because everything would behave very normally and, like.
Nothing unusual.
Like unusual.
Like ADI called out.
Yeah. ADI is delayed by a month, right?
Yeah. Yeah. Yeah. And with them, but it was only in auto, and they said, "We know what the normal trends are in this for a couple of weeks." It was unusual, so they called it out.
Yeah. Again, the auto tariffs came in in Q1. Look, our parts are mainly, we have our automotive customers in the US using our US-based parts, and it's all consignment-based. I do not think there was any anxiety there, at least from our perspective. You know, April was a little interesting. At the end of the day, if you think about tariffs, and everybody talked about the China tariffs, and now semiconductor tariffs are exempt on both sides of the globe, right?
For now.
For now.
Yeah.
If they want to look, I want openness and, you know, everything to be as normal as it used to be maybe. I think it's not going to. If you just look at the, forget about the noisy weekly news. Just look at the trend from the last five to eight years.
Mm-hmm.
This has not been a one administration or one, this has been a trend.
Yeah.
Okay? If you kind of average it, it's a trend of, I think, decoupling at least on the semi side. Even if you listen to the Secretary of Treasury two, three weeks back, I think he mentioned it. We do not want decoupling from China. In some areas, and I think semis alongside pharmaceuticals and aluminum and steel were mentioned. We have to be prepared for that. We have built our capacity in the US not for that reason, but we are in a good position. Okay? That is really where we are. I think our customers, and especially those who spend more time in Washington rather than less, understand that very well.
Mm-hmm. How much time are you spending in Washington these days?
More than I thought two years ago when I took the job. To be fair, this is not related to one administration or the other. I've seen a very continuity on the objective to have the US be controlling its destiny in terms of semis, you know, different, different maybe ways to get there. But it's been very coherent. I think this is like the one bipartisan area where there is totally agreement on.
Yeah.
Look at COVID. I do not think for a bad reason. I mean, you do want to control your destiny. And semis are now everywhere. They are in, they are in energy. They are in medical. They are in aerospace and defense. They are in data centers. They are in EVs and robotics. It is a fabric of our economy. I think you want it to come from dependable sources.
Yeah. Yeah. How much of your business actually goes into China these days?
You know, more or less aligned with our share of GDP. It is 20%, I think. It was 19% last year and 20% in Q1.
That headquartered companies.
Headquartered. That-that's really where.
Yeah.
To me, what matters is where, what's the end consumption of the semis that we ship into China. For the headquartered company, it's a good proxy for what's being consumed in China. They still do have some export business, but it's not very large. We think we are at a good scale, meaning we are not over-penetrated, we are not under-penetrated. We wanna play. Look, the China market is important.
Yeah.
It's 20% of world GDP, less than 20%, but higher than 15%. It's very strong on the automotive EV side. There is a good robotics opportunity. TI wants to play the game in China. It's a very similar game that we play worldwide. It's industrial and automotive. It's selling a broad portfolio. It's many, many customers.
Yeah.
We are continuing to compete at a very high level.
Got it. I guess, like, from that standpoint though, is having a primarily U.S.-focused footprint in this kind of a world, does that become, you've talked about it as an advantage. Could it be a detriment as well?
First, I mean, think I we just talked about it. China consumption is around 20% of, of world semis. I am talking about what stays in China. In that sense, we have more than enough on our non-U.S. manufacturing. We have a manufacturing site in China. It is an end-to-end. And.
Is that Chengdu?
In Chengdu, we have a full, full turnkey, almost. That's the only place in the world that can do full turnkey from, you know, from substrates all the way to finished goods.
Is that used primarily to serve, like, China demand or?
Mainly China, but, you know, it can support everything else. We also have two fabs in Japan, a fab in Germany. And we also have our foundry partners. Think about our embedded business. Embedded can be built in.
Mm-hmm.
In Lehi now, but also, of course, in Singapore and Taiwan with our partner. So we have enough tonnage to support China. That's not the issue. Of course, I prefer to build it in a 300mm wafer fab rather than 200.
Mm-hmm.
Our 200s are under-utilized, and we have enough capacity. Now, everybody likes to talk about China as a risk. No one talks about the upside.
Uh-huh.
There is another side that, how did it all start? It all started because there is a determination of U.S. government, previous and current, to bring semi manufacturing, mainly fabs. That is where the focus is, to the U.S., who is currently investing in the U.S. in our area. Of course, there is investing in Arizona and other places for advanced nodes, but we are, I think we are the only game. In that sense.
Some of the others that were doing a little bit less, a lot less, but some were, were dialing it back too, right?
Those who did something, look at their operating cash and investment in the last two, three years, it all went down. I mean, we are 10X, versus the competition.
Yeah.
At the end of the day, it is recognized. We said it many times, but now people care.
Mm-hmm.
There is a lot of, to me, as I said, I want the world to be open, but if it decouples, I see more opportunity than risk.
Okay.
Let's say that.
I guess, what does the competitive environment in China look like? I mean, everybody always, I-it, because I always, it's not even just in analog. It's in analog, and it's in AI, and it's in semi-cap. Just given this, my view is that the Chinese guys, they're not as good, but they will probably take more share than they would ordinarily deserve to take because they have no choice.
Yeah. Be careful on not as good. I mean, I have.
Yeah.
I feel a little bit like a broken record.
Yeah.
You know, I think the third time we talked about it.
I mean, we've had more share across all these different areas.
No. I mean, they have some disadvantages. We'll talk about it in a minute. Look at how they build EVs. Look at how they build, you know, wireless base stations.
Mm-hmm.
They, this is a capable, a capable country and a capable engineering workforce. Okay? Let's start with that. Every time the China competitors choose to make a part, they are able to do it. I will tell you, maybe it takes more iterations, but they iterate quickly. And it's across the board. I like to talk about the portfolio of analog and embedded as general purpose, kind of more application-specific. The general purpose are more like a building block of the analog and embedded market. This is when you start a new application, you start there. Then you have application-specific that are more integrated, so solving a specific problem, maybe in automotive, maybe in solar energy or whatever. They play in both domains, and they play well in both domains. What's hard is the breadth.
The breadth is an issue because in our market, to satisfy a board or to solve a solution, you have to have so many parts. That's why we have 80,000 of them.
Do you still talk about look left, look right?
You told me that last year, and I, I, we do.
Yeah.
Okay? Sell the board and understand every socket. By the way, there is a socket here that we cannot address. Let's invest. Okay?
Yeah.
Some of these sockets are running at 100K a year, you know, across many, many customers. It is just hard to go attack all of it. It takes really decades to build a portfolio. It does take decades. This is where we have a competitive advantage. Not only that, we are cost-competitive because we are vertically integrated.
Mm-hmm.
With both fabs and ATs. We also have a very broad portfolio. I mean, we have positions at customers that they value. On top of it, you know, think about the markets we play in China. It's about 80% industrial and automotive. We are, you know, customers care about longevity. They care about support. They know what happens when there is a failure or, you know, in a car OEM. Can you do my failure analysis? That's heavy lifting for many of the local competition. This is where TI has an advantage. We are competing.
Mm-hmm.
To your question, the competition intensifies, you know, it's just stronger every year.
Mm-hmm.
I'm glad that we decided to pick up the challenge back in, you know, in 2017, 2018 rather than wait for now.
Mm-hmm.
Because if you wait for now, it's kind of too late. It's hard to catch up. I think we are running at the same speed. This is why I like to call it our conditioning room. If you can compete there, if you can run the Shenzhen and the Shanghai treadmill.
Mm-hmm.
You can ride also the Detroit and Stuttgart,
Yeah.
Treadmill. That's the way I look at it.
Yeah. How much of your workforce, like, field engineers, R&D, is actually in China right now?
About 20% of our workforce, you know, according.
Yeah. Kind of aligned. What we have done over the last years is in terms of R&D, in terms of manufacturing, we do not want to be oversubscribed. R&D, I have less talked about it, but we want R&D to come mainly not out of China. It is also very competitive. It is hard to keep a team there.
Mm-hmm.
The second point is on the manufacturing side, we are good on the front end, but backend a little bit oversubscribed because OSATs, which are, you know, the.
Mm-hmm.
Our external manufacturing for the assembly and test are mainly China. We need to own our destiny there. We are working on, by the end of this year, we'll be at the right level. 20% out of China.
Yeah.
80% outside of China. That's where we wanna land on the ATs.
Got it. Yeah. We didn't talk about your backend as, as somebody that's, but you're, you're building a ton in, in within the Philippines. Is it?
Philippines, Malaysia. We have, in North America, in Mexico.
Uh-huh.
and ATs right now, they have to, they're heavy labor, but that's also gonna change one day.
Yeah.
Stacey, if you think about, and it's part of what we discussed about robotics. We are training robots in our ATs right now. And one day, can you run, lights out, AT facilities? You can. I think it's gonna come.
It's not around the corner, but when that comes, you can actually bring it to areas like, like the United States.
Yeah.
Right? It's one of these things that I'm very excited about, that automation and digitization are going to come into our life, including in semi.
Yeah.
Manufacturing, including ATs.
Yeah.
Who would believe, right? That is an exciting, that is an exciting future for TI.
Got it. I'll have to do some work there, I think. Just one more China question for you. Just given everything that's going on, have you seen any change in the competitive arena in China, especially just given everything that's going on right now? Has there been more of a push from the Chinese customers to do more, to use more local supply versus, because I'll be honest, I always sort of joke it's like you got some companies that do business in China, but I always say, like, Texas has Texas right in the name. So I always wonder if that causes an issue.
Our brand in China is very strong. Customers really appreciate it. And we do appreciate them. You know, some of them are leaders in what they do in the world, global leaders, not only in China. We learn a lot. I think they also rely on us. In that sense, customers appreciate our portfolio. They appreciate our support. We have a very strong sales team in China, very urgent. We are doing well. The customers are very proud. I actually talked with them after even tariff days, you know? They say, "Look, we need you. We, you know, the government knows we need you, and we count on you." When I spent time there in Q1, again, pre-tariff, but all I heard is, "Thank you for your support. Thank you for your portfolio.
Thank you for your capacity.
Mm-hmm.
Because China, like everything else, everything comes quicker.
Uh-huh.
From a cycle perspective, we were talking about, you know, how TI solves issues on supply from our competition rather than.
Mm-hmm.
Any tariff discussion? That was very refreshing.
Got it.
I'm not surprised.
Are you working from the U.S.?
We today have all the backend, of course, out of the U.S. The front end, it depends. You know, it depends on the part. What we are doing right now is to prepare in the case of a tariff coming back, also in China, I have to requalify some of our parts that are in the U.S. into Japan, into Singapore, into China.
Uh-huh.
We are doing it. Analog or Embedded is already ready because they are coming out of Singapore and Taiwan, and they are going into the US. They are already ready. On the analog side, I do have to do some requalification because the rules have changed. It used to be based on the COO.
Mm-hmm.
Or country of origin.
Yeah.
Used to be on the back.
Yeah.
Now it's a front end.
Yeah. Yeah.
I have to now take some of our parts that are running in the U.S. and requalify in Japan. We have time. This is something that we are doing. We are also loading our fabs. You know, the color of wafers, where it came from, was not important before. Now it is. We are also offering to our customers flows.
Oh, okay.
This is.
Oh, okay.
Yeah. What flow do you want? You want the U.S. flow? Okay. That acronym. You want or that prefix. You want another flow?
Can you preferentially charge for that?
We think we should.
Uh-huh.
Yeah. I mean, it's not gonna be the 125%, other, but yeah, if you because if you don't run plain vanilla, meaning I decide where the wafer comes from, it's a little bit more burden.
Yeah.
On our side, I have to carry a little bit more inventory.
Mm-hmm.
I have to now remember where the wafers come from. It's an investment in IT. So we are doing it right now, and we are already offering this to customers.
Mm-hmm.
in China and also in the U.S., by the way. It's not only in China.
Yeah. Yeah. Maybe that's a good segue into, like, gross margins. And we all know gross margins have come down. And we understand why. I mean, it's just that the heavy CapEx burden, there was a depreciation burden. There was lower utilization just because of the nature. I don't actually wanna talk about, like, downside. I actually wanna give you a chance to talk about the upside on gross margins. I'm gonna be very, I'm gonna be nice here. And there's a number of drivers. I mean, clearly, if utilizations are getting better, fine. But you also talked about closing older fabs and starting to bring some of that volume into more cost-effective 300 mm.
I guess, you know, where I can't even remember where gross margins are sitting right now, like, like, mid to mid to high 50s, something like that.
I do remember that. No.
I can't remember where the where the.
High 50s, right?
Where the guide was. High 50s?
Yeah.
Maybe you could talk to, even if you don't, I'm not asking you to give us a gross margin target, but, like, qualitatively, can you talk through some of the positive drivers on gross margins from now, now that we are sort of approaching the end of the investment phase?
Thank you. I like the way you asked me today. We run the company on free cash flow per share targets. That's the way we measure ourselves.
I always joke. It's like, it's not even a joke. You guys, you don't run the company for gross margins.
Correct.
It's never been a.
Correct.
Target, right?
No, but I know it's, but I respect it because it's important. To me, it's just, it, there is a little bit of delay. We like free cash flow per share because you see more immediate what's going on. With accounting and everything, margins follow, right? High level, I will say, Stacey, from a tailwind perspective, I see three, month three coming in right now. One is the shutdown of our last two six-inch fabs.
Mm-hmm.
There is one in Dallas, one in Sherman. Same land.
In Texas?
In Texas. Yeah. Yeah.
Okay.
By the way, the Sherman fab has already parts that were moved from Houston fab that was four inches.
Yeah.
I think. It's kind of crazy. We have now probably 90% done. Okay? We've moved these parts either to our 200-millimeter wafer fabs or redesigned them into 300s. You can almost think when you move a part that used to run on six-inch into a 300, the cost of chip goes to quote-unquote zero.
Yeah.
That's nothing.
Yeah.
The gross, you know, the gross cost of running a fab goes away when you shut down the fab. It's gonna be shut down, our last wafer start is, I think, Q4 this year. And we are gonna wind that cost in 2026. That's one tailwind. The second one is Lehi. That's a substantial one when we, when we move the wafers. Because think about the gross cost or think about depreciation. We, we've put the CapEx, the equipment. We have the people because we took it from Lehi One.
Is this in Lehi One or Lehi?
Lehi One. Lehi One.
Okay.
and Lehi One right now, as we ramp it, and right now it's, it's below 50%, but we're gonna leave the year at about 50% utilization in Lehi One. So.
Just from the transition or from the?
On transition.
Okay.
On trans and there is growth in Embedded this year. When you think about that, that is another tailwind that is part of what we call phase one, meaning transitioning of six-inch into 300, transitioning of foundry into Lehi.
Mm-hmm.
The second element is, is really, Stacey, you know, the mix is gonna change. I mean, industrial is coming back. That's always a higher margin. That's gonna be a tailwind to gross margins. If you look longer term, and we've modeled the company towards 2030 and multiple revenue scenarios, when you go to a steady state of.
Mm-hmm.
Of capital intensity related to your growth.
What, what's that steady state on capital intensity, by the way?
1.2x of growth. If you.
Yeah.
If you grow at 7%, you know, long, long term.
15%.
It's gonna be, you know, 10%. If you grow at 10%, it's gonna be like 12, right?
Mm-hmm.
1.2x the growth. That is a long-term capital intensity. When you run that model, you will find margins are very attractive.
Mm-hmm.
I'm talking about the way they used to run, but it was not sustainable. We ran there when we were when.
Yeah.
We were underinvested in capacity. It's gonna go there very structurally.
Yeah. I mean, yeah. I still remember when you did the first 300mm, which must have been 2009, I guess, right?
Yes.
Yeah. I actually remember where I was. Like, I was actually in, it was, I was having lunch with my family in a Japanese restaurant on the Upper West Side here in New York, and I saw this come across my BlackBerry. My first thought was, "These guys are insane," right? My second thought was, "These guys are geniuses." I actually still remember the call that you did, and everybody was asking about utilization and like, "Oh my God, you're gonna be able, what are you doing?" Your results was effectively like, you guys are, you do not understand, like, do you have any idea what the return on these assets are gonna be, right? I remember back then you had aspirational targets to get to 55% gross margins someday. I think you topped out at 70%.
Yes. Yeah.
Right? And so under, and you're right. It's, it's, I don't know if it was un, I wouldn't call it underinvested. You, to me, you bought, pre-bought 10 years' worth of capacity at 10 cents on the dollar, right?
No, I'm not saying it was not sustainable, right?
Yeah.
Because we ran it at too low.
It's a good route.
Of CapEx. Yeah. Again, I think we can do it more structurally moving forward because these new fabs are, the return on investment is very attractive, but you have to wait. Okay? That's the, that's the, because they're very, very efficient. The scale of 300 is so amazing. We love the efficiency of the fabs. I think it's gonna be different this time because our fab, our fab two is ramping, as I said, 70% built and fully utilized.
Mm-hmm.
The reason it's 70% is not because it's empty. It's because that's what we can build there.
Mm-hmm.
This is why we are pulling in SM1. We see a lot of momentum. Okay? On Lehi, it's just a longer process to qualify 45-nanometer, and we'll finish this year. I think you'll see it coming in, Stacey, and we are very proud. If I may, on a personal note, in the last two years, I've seen how hard it is to invest in the US. I think that's part of the reason we see a lot of people stopping.
Yeah.
You know, all this hybrid manufacturing means I used to be making it. I'm not gonna make it in the future, right? So a hybrid is a transition into fabless.
Yeah.
TI is gonna be, you know, controlling its destiny in investing. It is very hard. How many, how much praise you get.
Yeah.
When you invest during a down cycle, of course, you seem to be crazy, or they think you're crazy. I think we are just thinking about it differently and see also the opportunity, and I think it's gonna come. I think we're seeing evidence that it's coming now. I'm excited about that.
How does the CHIPS Act play into all of these decisions?
It's not. To me, the,
Like, if there had never been any CHIPS Act, would you have done the exact same thing that you're doing now?
I mean, the way, the reason we have accelerated our efforts was not because of CHIPS Act. It's because we saw a larger opportunity, and I'm very glad we have done it. It does help. I think it has to level the playing field. Forget about all, I don't, I'm not a big fan of the grant stuff. You know, the ITC part of that is very important because it just levels the field versus, you know, people who are building it in Japan, in Europe. China is getting fully subsidized. Okay? They do crazy stuff, but I'm not expecting us to do that. I think ITC is a good method.
Mm-hmm.
Let's see what they're gonna do with it, with the new bill. I think.
Yeah.
The ITC is a good way to have kind of a down payment for an investment, and hopefully it will continue.
All right. Do you have any thoughts on that? I just, my view was that Trump hates the CHIPS Act, and I, I do not know.
I don't know. I don't know. Of course, the word is not liked, but I think they do see ITC as a good thing.
Okay.
because it's fair. You know, few invest will help you. But grants, grants are, and also, to me, the process was not good. You know, we were asked to do all kinds of stuff that is not related to semiconductor manufacturing, like unions, and I like where we are going.
Yeah.
Let's say that.
Okay. Yeah. I agree with you. I mean, the tax credit at least was vigorous, 25%.
Yeah. That's where the money is. And to me, ITC is where we wanna. And this is when I talk with the new administration, I see openness to ITC. I don't like subsidies as well. I think it's discretionary.
Mm-hmm.
decision of someone, and they pull in all kinds of other stuff to it. There is a lot of bad connotation to it, rightfully so.
Got it. I wanna talk about the businesses a little bit, especially particularly Embedded. Some of the comments you made, particularly around Lehi, kind of resonated with me. Embedded's been a business for you that, you know, it's had its issues over the last several years. You guys have kind of admitted, yeah, there were some things going on, and you need to turn around. It looks like it's kind of bottomed, and maybe it's sort of resuming growth. Your comments on Lehi suggest to me that you at least see potential for considerably more growth and prospects from that business than maybe we've had over the last 5 to 10 years. I guess, what's driving that? What gives you confidence? What have you actually, I guess, what would even to say, like, what was the issue in Embedded?
Like, what did you do to turn it around?
Wow, that's a long topic.
Yeah. I mean, in the six minutes that we have left, I guess.
Again, Embedded, you're right, Stacey. It underperformed. I'm not looking at the last two years. It's a 2017 to 2024.
Yeah.
Very, very tough.
I mean, my math suggests you guys your market share got cut in half in that business.
It's kind of the seven bad years.
Yeah.
Some of that market share is by design. Okay? Us walking away from base station.
The common infrastructure stuff. Yeah.
Yeah. Base station was hundreds of millions. And also infotainment, you know, overall, above a billion dollars that we walked away from. That is structural, decision, strategy. The other part was issues on execution, Stacey. Execution meaning we went under, the strategy was not a good fit for the competitive advantages of the company. We like to build internally. We like to have a breadth of portfolio. We like longevity and diversity. It usually goes with these low, the kind of smaller sockets. Instead of $50 per chip, can we sell those, you know, singles and doubles, one or two bucks, and sometimes 50 cents, MCUs so we can play that game? I think that is what we have done in the last five years. Now, why we have confidence is now we have enough evidence.
Mm-hmm.
That not only we stabilize the business, we are prepared to take share. The reason we want Embedded to be a strong contributor to free cash flow per share, to your comment, maybe even better than Analog because the compare is easy. Okay? When you lost so much market share over the last seven years.
Mm-hmm.
It's time for you to play offense, and that's what we are going to do. We have a new portfolio that is coming to fruition. We have new wins, our final. I hate to talk about design in finals, but it's so substantially different in Embedded versus the Analog. When I do the comparison over the last five years, it grew tremendously. The expectation from Amica and the team is that you grow fast. In the last peak to peak, and I show the slide in, from 2018 to 2022, Embedded revenue, forget about market share, declined.
Yeah.
That's not gonna happen, from '22 to the next cycle. Embedded was flat over here in Q1. It grew sequentially nicely. It's gonna do it again in Q2. The target and the objective of the team is to accelerate from here. I'm talking to you the way I talk to the team. It's the same message. It's time to grow.
Mm-hmm.
We've invested. We wanna see the results in 2025. Please check us in, check us every quarter, you know, every quarter at a time. Remind me if we are behind because I'll use it with the team. I have strong confidence that we are going to make that happen in 2025.
On the Analog side, though, I mean, to drill down onto the market share point there, I mean, lost a decent amount of share over the last several years, maybe since COVID started. In Analog, I mean, you went from 20% market share to 15%?
Yeah.
Or 15?
That's when, yeah.
When was that?
First, it's 4 points. It's not 5. But first, Stacy, let's be fair. Our Analog share was running at 18 for many, many years.
Yeah.
It popped to 20 at the beginning of COVID. Okay? It popped because we had supply. We had supply. We had inventory. Remember, back in 2020, we ran the fab open.
You were the only ones to run.
Yeah.
Yeah. Yes, it dropped, and I'll talk about it in a minute, but I don't think you know, I wanna count the money at the end. I-I'm not conceding on the Analog. I think our position is strong. I think the cycle is not done. Look at our share, the quarterly share even, on a quarterly basis in Q1. Wait for Q2.
Oh, you always tell us not to look at the cycle.
Yeah. Yeah. Look at over the long term. The cycle is not done, Stacey. The cycle, we are seeing a first-in, first-out evidence from TI versus the competition. I think this continues into Q2 and beyond. A long story short, I think the Analog business is in a very strong position. Now, the issue we had, and that's our frustration, and this is why we are seeing us in a very different mode during this down cycle, we had the positions. We had the design needs. We were on board. As a young COO, I had to call, you know, the customers and tell them, "This was your last wafer start. This is your last wafer start." We had to allocate our wafers into markets that we thought are more important.
That's a tough situation to be.
I've never understood it because, again, you were running the fabs. You were making stuff. Did you?
The issue that we have, and I don't wanna go too deep here.
Yeah.
We ran out of clean room. Our fab too took a stumble. You know, remember Rich?
Yeah.
We took a short sabbatical.
Yeah.
You know, remember that? There was a loss of, of the company stumbled before Rich came back.
Oh.
When Rich came back, we drove our fab too in the highest speed, but a little bit too late. This is part of the lessons learned. We will admit first when we've done a mistake.
Mm-hmm.
The good thing, we have the part. We have the position as customers. Even those same customers I called to say, "That was your last wafer start. We are back in, not at 100% share, but we are winning back sockets.
Okay.
They were all in China. They were all PE customers.
Oh.
That's where we had to.
Faster cycles.
Faster cycles. You know, do not go and protect your automotive and industrial market customers, unlike what we have done in 2010 or 2009 after the GFC, support them with your wafers and come back, fall on your sword, which we have done, and try to win back share. This is part of the reason you see us growing in PE. It is across the geographies, including in China. We have work to do. To me, I prefer that problem, that you were not prepared last time because you did not have enough supply. You are now, and we are going to play a different game in this next up cycle. We are very prepared for that, Stacey.
Got it. Got it. We got about one minute left.
Okay.
We've got a room full of investors here. Again, I will, as I always do, I will give you your soapbox.
I thought we'll be alone here on a Friday.
Yeah.
Thank you for coming.
This is all you. They're interested.
Okay.
Why should they buy Texas Instruments stock?
Look, I, I've been in TI for more than 25 years. And, for the last 10 to 15 years, we worked so hard to reposition the machine, I call it, the strategy, Analog and Embedded. We talked about divesting some businesses, stepping out of business. You know, 15 years ago, we had a big business deal at Nokia. You know?
Remember. Remember very well.
And, and BlackBerry. Remember them?
Yeah.
Yeah. We have done a lot of work to position ourselves in the right market. We are well-positioned into industrial and automotive. We have a great technology that can go into more markets like data center and beyond. I am very excited about our position. We do not have any more headaches. We have built the capacity. We have the machine. We have injected higher level of competitiveness. It is time. TI is prepared to play offense. You will see it coming in every quarter, starting this year. I recommend joining us. It is going to be a wonderful journey for the company in the next 10 to 15 years.
Got it. I think we'll close it out there. Thank you so much.
Thank you, Stacey. Always a pleasure. Thanks.