Morning. Morning, we'd like to get started. Thank you for joining us on day two of the conference. My name is Toshiya Hari. I cover the semis and semi-cap equipment space at Goldman Sachs. Very honored and very excited to kick off the semiconductor portion of day two with the team from Texas Instruments. We have Rafael Lizardi, Chief Financial Officer, and we have Dave Pahl from the IR team as well. I'll kick off with a list of questions, but hopefully we can keep this somewhat interactive with questions from the audience as well. First of all, thank you guys for coming here.
Thank you.
Thank you.
Really appreciate the time. So Rafael, I want to spend most of our time, you know, discussing strategic elements of your business, but I do want to hit on a couple of near-term questions. On your recent earnings call, you guided September quarter revenue, essentially flat on a sequential basis, which I think is sub-seasonal, although seasonality has sort of gone out the window.
Mm.
What's embedded in guidance, what are some of the puts and takes from an end market or geographical perspective? And to the extent you've seen any volatility around kind of the central case so far, what have they been in the quarter?
Yeah. So, as we said at the earnings call, the quarter came in about as expected. All end markets were weak, except for automotive that remained strong.
Yeah, and I'd just say, as we looked at, you know, first quarter and second quarter, you know, as we looked into the guide for third, didn't see a significant change to point out, so, so I think the way you described it, you know, what we saw, automotive continued to be strong, and we saw weakness in the other markets.
Okay. So for the September quarter outlook as well, it's, it's auto that sort of stands out in terms of still being resilient. Other parts of the business, from a market standpoint, remain fairly subdued or muted.
We don't break the guidance for the next quarter into pieces. We kind of stick to what happened in the quarter that ended, unless there's something significant to point out, which we didn't.
Okay.
Yeah.
Got it. I realize you're not in the business of predicting cycles.
We're not.
You're not. Which is good and fine, but you have been in the business of semiconductors for a very long time, and you've experienced many, many ups and downs. As you compare and contrast this current cycle with past cycles, any similarities or fundamental differences that you'd kind of point to?
Yeah, sure. So, you know, to me, there's always three things going on. One is the cycle itself, the semiconductor cycle, and that's just a function of inventory, right? Customers build too much or not enough inventory, and that's what drives the cycle. On top of that, you have macro situation, which that can be different, right? And that could exacerbate or mitigate the cycle. And then in addition to that, you have secular trends, right? Which in the case of semiconductor, clearly we have, particularly in auto and industrial, a lot of content growth that has been happening and will continue to happen for the foreseeable future. So, you know, you can think of those three things and how those are playing out in the current environment.
The last thing I would point out that's clearly different this time is the asynchronous nature of the current cycle. Meaning, in the past, as far as I remember, all end markets moved about the same time. This one has been different. So personal electronics kind of peaked in 2Q22 or so, and it started weakening since then. Industrial started weakening a few quarters later, not nearly as steep as personal electronics. 'Cause I think we talked about that, our personal electronics business was down, like, 50%, in this last quarter from the year ago quarter. Industrial, I can't remember the exact figures, but, down, but not nearly as much. And as we said earlier, auto has remained strong.
The asynchronous nature of this current cycle, would you sort of characterize that as a good thing?
Well, it is what it is. I'm not, I don't call it good or bad. It seems that the root cause was the pandemic and a lot of people bought, you know, consumer electronics for so long, and there was a shift in the demand towards that, and at some point, people had enough laptops and TV monitors at home, and they stopped buying. Our customers stopped buying, started draining their inventory, right?
Yeah. Yep. So, during the pandemic, you know, you referred to products where lead times were extended as hotspots, right? Are there still hotspots within the context of your portfolio? If so, where, to the extent you're willing to share? And where are average lead times today, how do they compare with pre-pandemic, and what's considered sort of a good place to be for TI?
Sure, maybe I'll take that. So I think we've made really good progress in addressing the hotspots that, you know, we've mentioned in the past. You know, obviously, in a portfolio of 80,000 products, there's always gonna be, you know, a device here or there that's gonna be a hotspot, but obviously, we've had the opportunity now to build inventory to position us well for, you know, the next side of this cycle. And then our goal is to have excellent levels of customer service for all of our customers, all of our end markets.
It's hard to specifically point to a single, you know, lead time number for the whole portfolio, but, you know, as you look at our website, making sure you've got more availability there, customers can come in and get products, you know, same day or next day and/or second-day ship. You know, we wanna make sure we have as many of those as possible, and we've made a lot of progress on that.
Got it. During the pandemic, many of your peers introduced long-term customer contracts, right? Under different acronyms and names, NCNRs, LTSAs. You guys did not, and I don't think you have plans to do so. Why is that the case, and have you had customers come to you and say, "Hey, TI, we'd love to have long-term supply, you know, guaranteed from you guys?
Yeah, I'll maybe take that one, too. We don't believe those types of programs create demand. Customers taking product that they don't need is not something that we would prefer. If you think about how we wanna make sure we grow our business, it's not just we wanna have one or two customers that do well. We want all of our customers, all of our end markets, to be successful. And so, as we think about the capacity we're putting in place, it's for that entire group. Obviously, we've had the opportunity to share the roadmaps that you all have seen in terms of the fab and the AT roadmaps with our customers. They're very excited about it, but again, that's built for all of our customers, not just for, you know, a few of them.
You know, of course, those conversations happen, but we wanna make sure we service our entire customer base.
In, you know, keep in mind, the majority of our products are what we call catalog, so off-the-shelf. So that means that we're not designing-- we don't design them for any particular customer. It's for a broad set of, customers, applications, even end markets. So, so that anchors our, our view on long-term agreement. I mean, think, and this is not a perfect analogy, but I'll throw it out. I think if you go, go to Home Depot, buy hammers and nails, and they say you gotta sign a long-term supply agreement and take, you know, 10 hammers per month for the next, 10 months, I mean, that's, doesn't make sense, right? For an off-the-shelf product. If it's a custom, product, I understand how the dynamics are different, but, but our strategy, doesn't call for it.
Right. I like to think of your products as being more sophisticated than-
Hammers and nails.
Yeah.
Our products are cheaper than, hammers and nails.
Yeah. Yeah. I guess in terms of the long-term growth drivers of your business, you know, you've consistently called out automotive and industrial as key focal points for the market, and it's been the right call, given how the various markets have played out. I know at TI, you don't focus on one or two or three things, but within the context of automotive and industrial, if you had to sort of point to a couple of applications that you're really excited about, what are they? What would they be going forward?
Yeah, and I would start with, there are great opportunities all day in market. So obviously, we have had a strategic bias toward automotive and industrial in terms of where we invest, but there's great opportunities everywhere. Maybe if I was to— And this is one of the beauties of the type of products that we make, is that they go into everything. But if you look at industrial, just for a moment, you know, you have 13 sectors there. What's driving that? Automation. And what's driving automation? Access, cost, and safety of labor is a piece of that. You look at things like building automation, the grid, build outs, motor drives, reduction in energy use and greenhouse gas is a big part of what's driving those markets.
You look at medical, the fact that you had, probably 20 years ago, instruments that you wish you could even have in a hospital now or in your home. You can go through every one of those 13 sectors and there's a long-term growth driver underneath it, and we play in all of them, which is what's exciting. And that's before you even get into automotive, which, as you know, when you transition to an EV, you remove the transmission, you remove the belts, you remove all the, the hydraulically-most of the hydraulically driven things, and they become electrically driven. And that's more opportunity. So you got, instead of a gas tank, you got a battery. It's gotta be monitored, it's gotta be charged, it's gotta have a traction inverter.
So all those things mean semiconductor opportunity, and I think we're in the early phases of that. So, we are fortunate that because of the portfolio, we don't have to pick a single winner. We can be in all of those things. So those are probably the areas I point to, is a few of the many that we're excited about.
Got it. There isn't a single session that goes by without mentioning the word or term AI. Within TI, again, I know you guys are, you know, a massive company, you're very broad, but do you see AI or data center sort of moving the needle for your business? Or again, is it one of dozens and dozens of applications or end markets?
You know, so enterprise is about 6% of our revenue. It's an important end market for us. Our products don't really care what the workload is, whether traditional workload or it's an AI workload, but if there is a processor there that's doing AI processing, it's gonna need power, it's gonna need telemetry, it's gonna need data communications. Those are all the types of products that we participate in. So we do have exposure there. It's not the only thing we are exposed to, but that's certainly an interesting and exciting market for us as well.
You know, beyond that, obviously inside TI, we have many business processes that we're always looking at for improvement, and AI is another tool that we can plug into those business processes to make those better, and that applies to R&D, sales and marketing, manufacturing, and even some of the support functions, like in finance, for example, looking at ways that we can incorporate AI into our processes and be more efficient.
Since you bring that up, Rafael, are you kind of early in that process in embracing AI and sort of leveraging AI as part of your operations?
When you say early, I don't know, compared to other companies, I don't know. But, we—i t's something that we're doing and we'll, we'll continue to do. I guess maybe the key point is, I don't see— our emphasis is on the business processes, and whatever tool makes sense to put in those processes to, to make them better, we do. AI is one tool. You know, we don't get overly excited about, you know, the hype in one place or another. But, if it makes sense to, to do something along those lines, to, in that particular process, then we'll, we'll deploy it.
Got it. Shifting gears a little bit, a couple of questions on the geopolitical backdrop and how it's affecting your strategy and how you think about it. Obviously, the backdrop is extremely challenging, difficult for everyone. Can you remind us how much of your revenue or how much of your products are being actually consumed in China, what your exposure there is? And again, just given how the relationship is going between the U.S. and China, has your strategy in China, the approach, resource allocation, has any of that changed at all over the past, call it 12 months -18 months?
Yeah. So a couple angles to that. You asked the question, consume. Let me reframe it a little bit, and then I'll try to get to consume. About 20%, and it's all in our 10-K, and thank you. About 20% of our revenue ships to, it's for customers that are headquartered in China, so kind of Chinese customers, right? Now, does that mean that is where the consumption happens? Not necessarily, right? Many of those customers are— They're obviously shipping in China, but they're shipping to many parts of the world. I mean, you take BYD, for example. They're shipping electric EVs in South America, right? And in Europe and in other places. So, as far as consumption, our best, frankly, guess on that is just percent of GDP, right?
So whatever percent of China is of worldwide GDP, that's probably about of the percent of our products that end up being consumed, you know, ultimately in our customers, maybe even customers', customers', product. The second part of your question, I'm sorry-
Your strategy and your approach toward China as a market, has that changed at all? Evolved at all?
Yeah, it has. So we have sales obviously in China. We have some R&D in China. We also have manufacturing in Chengdu, and that's— I mean, pretty much everything we do, you would categorize as legacy. We prefer to categorize it as optimal manufacturing technologies for auto and industrial, analog, and embedded. But we have one fab and— Well, one AT, assembly test in Chengdu, in China. Now, not only a vast majority, I would say, our entire- the entirety of our fab investments over the last couple of years and going forward, that's all in the United States, the vast majority. Clearly— So we have our fab, our fab, which is Richardson fab. It's just north of Dallas. We have two fabs there.
They're next to each other. In Sherman, Texas, which is just south of the border of Oklahoma, we are building two of what will be at least four factories over the next few years. Two of those will be shell-ready, and one of those two will actually equip and have a pilot line there by about 2025 or 2026 or so. And then in Utah, we bought a factory there about a year and a half, two years ago, from Micron, and we've already qualified it, and now we announced plans to build a second factory there. Those are all 300-mm state-of-the-art factories. Four are 45-nm to 90-nm, maybe 130-nm. We include the analog part technologies, optimal for those end markets.
And that provides us control over our manufacturing footprint. It also gives us the best cost structure because it's 300 millimeter, and like we were saying, it's geopolitically dependable, so meaning it's in the United States. It will not be affected by any issues between the United States and China, and potentially Taiwan.
That sort of brings me to my next question. You know, you're probably one of the few beneficiaries of this, if you will, just given your presence in the U.S. And I think you've shared a couple of times, some anecdotes as it pertains to customers coming to you, asking for more, three years out, five years out. Can you share some of those conversations, what they're like?
Yeah, let me make one more comment about the previous question, and, Dave, if you want to address that.
Oh, yeah.
But, I forgot to mention assembly tests. Don't forget, that's an important component of manufacturing, and, we're, we have a really tremendous footprint in Malaysia and the Philippines, and we're expanding in both places. That's also, we consider that geopolitically dependable, capacity. And as we announced at the beginning of this year in capital management, those investments will take us from about 80% internal on the fab side, to 90%-95% internal, and on the assembly test, the AT, from 60% to about 90%-95% internal. That is huge as far as the owning our future, our destiny, but also the cost structure, because now it's internal, you don't have the double, the margin stacking, problem.
And as you bring things from outside to internal, you get immediate return on that investment. So, sorry, I just want to mention that.
Yeah, and in terms of the customer interaction, you know, the last couple of years have been really interesting, I think, eye-opening for a lot of customers, especially their executive levels, where, you know, you think of an industry where it was historically electromechanical, and I'll just throw an example, like construction equipment or tractors. You know, there, very little chip content in one of those 20 years ago, now full of chips. And when you're building a module for a piece of equipment like that, you're not just looking for something that's gonna be designed in for 2 years. It's something you have to be able to supply to your customers for, like, 20 years. And so what we found was that there's a large group of executive leadership that are realizing the importance of these devices.
They have seen our roadmap, as I mentioned before, and are really excited about where it's located, but also that they can, as they design it in, be confident in where they can get it for the long term. You span that across the entire industrial base, the entire automotive base, and that's where some of the confidence you've heard us talk about comes from. And so that, and that's what's really exciting is you think about what we're putting in and that differentiation. Customers aren't looking for more suppliers, they're looking for fewer that can address those challenges. So that's the type of conversations we're hearing.
The intensity or the volume of these conversations, have they stayed elevated? I imagine they really picked up during the pandemic, given the shortages, et cetera, et cetera, but this is more of a permanent dynamic where, I mean, the geopolitical landscape isn't changing anytime soon. So I'm guessing, you know, you're still pretty busy with these conversations?
Yeah, what's interesting is the shift went from—i nitially, it was, you know, I'd say there's a specific product that's needed or something like that, was where the conversation started, and then it evolved to: Okay, what about, ten years out, what's that roadmap look like? What's that footprint look like? And those conversations continue because, of course, you know, our customers, you know, believe in the growth of EV and believe in the growth and, and what you see in those sectors, in industrial, I mentioned before, they're, they're not less excited about the opportunity in front of them, so they're certainly wanting to have those conversations, and they're, they're more strategic and long-term in nature than they probably were two years ago.
Right. Right.
Given what you just said, I'm sure you get the question: you've talked about spending $5 billion in CapEx every year, what would you need to see for that to change? The short answer is probably, you're not gonna change it, given those conversations.
So, yeah, let me address that broadly. As we said, well, earlier this year, in February, during our annual capital management call, we announced that we had a CapEx plan of about $5 billion per year for the next 4 years, so $20 billion. That was up from the previous year; we talked about $3.5 per year, so call it $14 billion, going to $20 billion. And that's to enable a potential revenue growth of about 10% CAGR through 2030. And we have some charts, you know, that talked about that. A few comments on that. Well, one, that enables. That's not a forecast of revenue, but it enables; it could enable that much growth.
That's just, just great optionality, to have there. And keep in mind, that is, capital that lasts a long time. That equipment, even though we depreciate over five years, it lasts for decades. So it's really, what I say? It's a solid investment that doesn't, doesn't go bad, and that's part of the, the dynamic and the reason why we, why we make it, like that. The other angle to that, as we announced, at that point, is, those numbers are growth numbers, so not including the benefits from, the CHIPS Act. We talked about at that point, that we expect, on the ITC, so there's Investment Tax Credit, and there's grants, that's part of the CHIPS Act.
On the ITC, we expect about $1 billion per year for 4 years, so $4 billion of cash that should offset that $20 billion CapEx bill. On the grants, because it's uncertain, you have to apply, and we're still through that application process even today, we don't know what we're gonna get, but we certainly expect to get something. It's comprehended in our numbers, and you know, we believe it's a great legislation that puts the U.S. on an even playing field, and we are just very well positioned to be part of it. And we're, we think, a tremendous candidate to be part of not just the ITC, but the grant program.
Got it. Pricing has been a pretty significant tailwind for many of your peers in the analog space and the microcontroller space. And, you know, we had the NXPI management team here yesterday, and they talked about, you know, low single-digit pricing growth in 2022, mid-teens in 2023, and continued growth, if you will, in pricing for 2024. More cost plus, right? A cost push based inflation. You guys, I don't think, have talked about pricing all that much. You've sort of kept pricing relatively stable. I guess the question is: how do you go about pricing your products? We've heard through industry conversations that, you know, TI is being relatively aggressive in the marketplace.
Has there been a change in your pricing posture, or is that a fair characterization of how you're competing in the marketplace today?
Well, I'd start with, you know, our pricing strategy hasn't changed. You know, that we are competitive where, you know, where we compete, and for a device. We have the benefit in that we have a lower cost structure, that we don't have to sub-select where we compete. But our strategy hasn't changed. And did we benefit from increased pricing in the last, you know, few years? We benefited from it. Similar to how others did, too. Now, the degree to which I don't know, but we did benefit. Now, as we think about the future, what could pricing want to do? You know, it. I think there's a reasonable argument for either scenario that I think I've heard, and either of those will benefit as well. If pricing continues to be inflationary, we'll benefit from that scenario.
If it goes back to kinda how it used to be, we've operated very well there, too, and we've got the internal manufacturing, the cost structure to support it, and you know, positioned well. But the strategy is still the same. You know, price to where the market is and for a given device and you know, maintain that over time is how we've done it.
Let me emphasize a point that Mike just made. You alluded to aggressiveness in our prices. We're competitive. I wouldn't characterize it as aggressive. Our sense is what's really happening, as Dave alluded to, is other competitors have decided to sub-select, so meaning get out of price-sensitive end markets. We welcome that. We have the best cost structure with those 300mm factories that we already have and the ones we're building. State-of-the-art, but 300mm is 40% more cost efficient than 200mm, which is largely what our competitors have. Plus, they rely on foundries, which, you know, those foundries are not doing that for free. So our cost position is the best, and we embrace going after more price sensitive markets, as well as the, as the, you know, the more traditional higher-end double markets.
I guess as a follow-up, so as you sort of, you know, develop a product and you're sort of thinking how to price something, is it—y ou know, are you looking to grow market share over the next five years? Is it free cash flow? I know you don't— you know, optimize for gross margin. But what is the debate internally at TI as you price product? I'm sure there's supply and demand and where the market's at, to Mike's point, but is there one or two metrics that you really, really care about and prioritize?
You know, it's still true that a customer care part has to solve the problem well, perform well, has to be available, and it's got to be reasonably priced.
Yeah.
It is a factor of many.
Yeah.
When you price a device in its release, a lot of factors go into how you decide, and part of that is, where is the market at for a given family of products? And so there's not one, you know, recipe that you do that, but, it is a factor of many now. Our ultimate goal is the growth of free cash flow per share for the long term. Changes in price don't necessarily equate to big changes in share, in a short term. And so, you know, as we think about growth in the long term, it's gonna be across those different dynamics, the gain share through having great products, having a broad portfolio, having an excellent reach of channel, having manufacturing and technology in-house. Those are the things that are gonna help us continue to gain share.
Got it. Correct. Yeah, that makes sense. Maybe I'll pause here and see if t here's a question right there.
Morning.
Good morning.
Thanks for taking the question. Going back to the autos dynamic, I know you don't want to get granular, want to give granularity in terms of guidance and so forth, but one thing I'm curious about is, we're seeing a bunch of sort of higher frequency data that suggests that inventories are building on the lots. We've seen some companies come out and say default rates are going higher for used cars. There was a company that released earnings yesterday that got some attention.
And I'm wondering if you have ever done a back test on your operations and your financial model to say, "Okay, when we see these type of activities happen, inventories go up, default rates go up, there is a lag of X, Y, Z months, quarters, where it starts to impact us on the auto side." I don't know if that question makes sense. Maybe you haven't done the work, but I'm curious. Whatever perspective you have will be helpful. Thank you.
Yeah, no, so I'll take a shot at that. That sounds, you know, I see where you're coming from. It's just that, you know, we sell about 100,000 different parts to 80,000 different customers. Or it's the other way around, 80,000 parts to 100,000 customers. And, you know, the permutations are incredible, right? And the supply chain is incredibly complex, where we sell to a customer that then sells to a customer, to a customer. So, you know, to get into those type of analysis have generally yielded, nothing, you know, nothing that's useful that we can do.
Instead, we focus on being ready for whatever the market may throw at us, and particularly, since our products, as I said earlier, are catalog in nature, so they last, you know, on the shelf, they last about a decade. And our customers, their product life cycles, who are customers, are also decades, right? So we build the part. We have an inventory strategy to have enough buffers at different stages of finish, so chips and finished goods, and then to be ready to support the demand, whatever that may be, and do it on a sustainable basis, as far as inventory goes. And speaking of inventory, I think you have some inventory questions, don't you?
Yeah. You have grown inventory in the past, several quarters. You have a pretty wide range as it pertains to what you threw out on the capital management call, but I think you're at the high end. So how should we think about that going forward? I know you're doing it to, to better serve your customers, which makes sense given your shelf life, but as analysts, we've got to forecast, a little bit forward.
No, I understand.
Yeah.
Yeah, so a few comments on inventory, and the earlier question, I addressed a lot of that, but start with the low risk of obsolescence for those parts. So, you know, the write-ups that we do in any given year, in any given quarter are minimal. So start with that. But of course, it's working capital that's tied up, and there's also the sustainability of how long do you keep building inventory. At some point, you know, how much is enough, right? So I understand those concerns. So just keep in mind— Well, before I answer the question, as I said earlier, the key point, though, is supporting potential revenue upside, right? Once you're behind, you look at a history of not just us, any semiconductor company, just the cycles come and go.
Go back, I refer you to slide 13 of our capital management presentation. Go to our website, look at it. It's a history of the last 20 years or 30 years of cycles. They come and go. They. It's up and down, up and down. It's never been down and stays down. It always comes back, right? So, you know, I'm not about to make a prediction on the cycle but, you know, it's likely that at some point we have an upturn, right?
It goes down, it comes up, yeah.
And when the upturn starts, if you don't have a good inventory position, you are toast. It will take you a year and a half to get just back to par, right? We are very conscious of that. That's why we want to have that inventory. And given our cash flow fall through and our margin fall through, the upside is much greater than the downside, particularly with the low risk dynamic. That's maybe the overarching theme. Now, I will say to your point, that you know, at some point, hey, enough is enough, right? And it's just not sustainable to just build indefinitely.
So that's why, I had alluded to it on the last call, but we are lowering the wafer starts in the factory moderately, you know, thoughtfully, so that the rate of growth of inventory will slow down. That means that it's still gonna grow, it's just gonna grow at a lower pace and a more sustainable pace. And depending on revenue, not just next quarter, but over the next two or three quarters, then, you know, that we'll have a soft landing. Now, of course, what happens when you lower utilization wafer starts? It's a non-cash charge to utilization, right? So that is, that's just math, that's just physics.
Yep.
It's gonna happen, but it's a non-cash charge, so it doesn't affect the free cash flow. And then it just puts us again in a balanced position to be able to support the upturn, whenever that may be, and also build inventory on a sustainable basis.
Got it. Maybe a question on your go-to-market strategy. Obviously, you've transitioned away from distribution, more direct. I think direct, as a percentage of sales, has grown from 35% a couple of years ago, to 70%. And, you know, TI.com has gone from $100 million to $2 billion or something like that. Any sort of positives you've identified throughout the past couple of years as you made that transition, and which, I guess, inning are we in, in that transition?
Yeah, a lot of positives. If you actually go back, this change started actually probably more than a decade ago, when we took control of the design-in process and basically, you know, made our teams responsible for finding every socket on every board. When we did that, we found all these sockets that mysteriously weren't there before, and, you know, through the last decade, have really benefited from that. Now, in 2020, we made the change where we moved to one worldwide distributor and allowed the customers to go direct with us, and went from about 1/3 direct to about 2/3, and I think pushing close to 70% direct, as of our recent capital management presentation.
Having that closer relationship just gives you better insight, first of all, what they need right now, but also, what are they working on? What are they gonna be needing to do in the future? It just gives you better insight into that. So our view and what we have seen is being closer to your customer, that just helps you and helps them, you know, find you a better supplier. And customers have had the option. We basically let them say, "Hey, if you would like to come direct, we've built the piping to do that. So if you wanna be a direct customer, you can. If you wanna go to ti.com and buy, you know, 100,000 parts, you can do that, too, and you can get them the next day or, you know, the day after.
We'll do all the value-added tax, the importer of record, all the shipping stuff, all the messy stuff that you don't typically wanna have to do. We know how to do that now. And so that's just made us an easier supplier to work with. So that's how we think about it. And I think, what inning are we in? I think it's hard to say, and we started it a while ago, but certainly from the order fulfillment side, we're probably in the earlier innings of that.
Got it. Maybe in the last minute, Rafael, just on how you deploy capital. You know, you've got $5 billion in CapEx every year for the next couple of years. Your dividend has grown very steadily for, I forget how many years, 19 years, 20 years, something like that. How should we think about the cadence of buybacks? Does M&A come into kind of the debate internally at TI, or is that sort of out of the, off the table for now?
So capital allocation obviously is very important. That's, you know, that's essentially our most important job, and that's why, you know, we have the capital management call, where we talk about how we deploy capital. Over the last couple of years, and going forward for the next three or four years, CapEx to enable revenue growth is the number one call on capital, and we think that is what it's gonna be the top driver of long-term growth of free cash flow per share. Doesn't mean it's the only one, you know, dividend obviously very important for sustainability and growth. Buyback is a good addition to the repertoire when it makes sense from a accretive shareholder value standpoint.
And other, you know, there are obviously other calls. We just got done talking about inventory. That's a great allocation of capital for the reasons that I talked about earlier.
Great. Really enjoyed the conversation. Thank you so much.
Thank you very much.
Thank you.