GLOBALFOUNDRIES Inc. (GFS) Earnings Call Transcript & Summary

Great. I think we're on. Anyway, welcome, everybody. I'm Joe Moore, Morgan Stanley Semiconductor Research. Very happy to have with us here today the CEO of GLOBALFOUNDRIES, Tom Caulfield.

Thanks, Joe.

Thanks for being here. So I wonder if you can just maybe give us a little bit of an overview of the company priorities. You've come a long way and a lot since you became CEO a few years ago, you transformed the company, going through a difficult -- cyclical patch right now, but maybe you can just kind of talk to generally how you guys are doing relative to the priorities you laid out at the IPO.

Yes, it's a little bit out of the order we talked about, but I like the question. The priorities for us is to stick to the discipline of our model. We, first and foremost, provide solutions that need to be differentiated and make our customers' products better. So the fact that capacity could be added all over the world as long as it's not capacity that mimics our capacity, it's irrelevant to us and our customers. So our job is to make sure that every dollar we spend in R&D is in partnership with customers because we're doing something unique for them. I think a great example of that is that we made an announcement -- need to speak up -- sorry about that -- announcement not too long ago, in January with one of our automotive customers, Infineon, about a long-term agreement. That's not a technology platform. We started developing with them in 2014. And it takes that long to get the technology ready for our customers to not have it on their road map and start to use it. So stay very focused with the right end markets, right customers to develop that unique single-source business. That's the priority for the customer -- for the company. Now, how do we deliver that to the customers? It's through a disciplined manner of adding capacity. We're a manufacturing services company. What we're able to do as a result of the 2021 chip crisis working with our customers is we all got ahead of ourselves to add capacity and everyone saw in 2021 that 2023 was going to be a down year for the industry. We're sitting today shipping revenue last year on the equivalent of 2.2 million wafers. At the end of 2024, we will have capacity worldwide in our global footprint to 3 million wafers. So we have the ability to grow our top line in a very capital efficient way, which means for the next x period of quarters until we actually can use all that capacity, we're going to have a disproportionate amount of free cash flow versus we've had in the past, very consistent to our investment thesis around the IPO, and that's what you brought up. And I'd tell you the last thing of that, what's really important about this 3 million wafer type capacity is it gets us to a point of scale where not only can we invest at 20% of revenue to grow our business in a meaningful way, we can also maintain that free positive cash flow. So we're really at a very unique point for the company, the investment part to get to scale. That sustainable is behind us. Now it's about the industry and the world cooperating, getting the economics around semiconductor is growing again.

Yes. Thank you for that overview. Yes. So I just wanted to start with more of an overview question. But maybe if you could also talk out specifically to the 2024 environment and how you see things playing out? I know you've already kind of guided for a correction of sorts. Can you just talk through that?

Yes, I think to talk about 2024, you have to look back into 2023. If you think of the industry and how GF performed like-for-like. So our competitors that are -- they either just do 28-nanometer and above or 12-nanometer and above, or the TSMC, their portion of that business, that's where we play. You could argue we outperformed. We were down 9% year-on-year, where others were down somewhere between 15% and 20%. Now a lot of that has to do with the fact the nature of our business being single source business, having long-term agreements with our customer, creating capacity for their needs and us working with them to figure out how to minimize the impact. I'd rather -- if I'm going to run a business through a cycle, I'd rather have a cycle that runs a little bit longer but has less amplitude to it, just a better way of managing your business you get and why go down very fast and they have to come right back up. How do you be more resilient business? And I think that's what we've done through the cycle. So what does that mean? We enter 2024 and we sit here. Joe, for me, it's a little bit of a groundhog day for you and I. We did this a year ago, and we sat here and we were like everybody else, at least I was. The second half of 2023 is going to be this big cost [indiscernible]. Why? Well, because everybody is saying that, and this is what we've seen in the past. At least for 2024, I think we have good data points to feel good about the second half. We saw in Q4 inventory, if it wasn't peaked it was peaking. And in many cases, we started to see it come down. That's a good news story. The bad news story was still historic high. So I think we have 1 quarter, 2 quarters or so of getting that inventory across a broad range of end markets, more normalized. So we have a good sense why the second half could be stronger for our industry. And this one is more databased rather than historically, it always goes and comes back up. So for GF, we reset -- or, as you said, in our guidance for Q1 to a much lower level. And we think that's the low point for us and the rate of pace of growth for the rest of the year really depends on how fast that inventory comes down and how strong the second half will be.

Yes. I mean some of the data is pretty compelling. If you look at, particularly on the smartphone side, your customers took inventory down by around 20 days in Q4. And to your point, they probably have to do that again for like 1.5 more quarters. But like, you're not going to reduce more inventory next quarter than you did before.

No one could hear me?

[indiscernible].

Thanks, Jack.

Thanks, Jack. Sorry about that. Maybe it's good you didn't hear what I said before.

But you've seen a lot of inventory reduction from smartphone, in particular. And it looks like they have to do it again this quarter, but like the pace of reduction doesn't seem like it needs to get worse and they're going to run out of inventory in the next couple of quarters.

And we're seeing that, too. And I think that's the area where I wouldn't say hopeful, but most optimistic for second half. It's 2 years in a row with handsets being down in volume, there's got to be some pent-up demand and some new features coming out that will drive some replenishment. I don't know how much of it is hype on AI-enabled phones, but they are pretty cool features that we're seeing starting to come out.

Yes, especially over a couple of years, there's probably more than hype, I think. Great. And maybe some other recent news. You've got the CHIPS Act grant kind of officially denominated at $1.5 billion. Can you talk about what that process was like, what that means for you? And how that process of getting that money works if the CapEx is this low.

Yes. There's a lot there. So for me, this has been a real journey that's still a story to be told. It started with first being a constructive voice as an industry player to get the need to create more capacity in the U.S., more capability and that it would need partnership investments, incentives, grants, whatever you want to call them. What we just announced was we've come to terms on an MOU. It's not yet a definitive agreement, but the hard terms have worked out. Now it has to become a contractual type of commitment. Now for us, I just spent the first part of my question for those who couldn't hear it, you'll hear a little bit more of it. We have a lot of capacity we've already invested in trying to grow our business. This CHIPS Bill is going to allow us to do a couple of different things: one, in our U.S. fab 8, the first part of that, we call it 8.auto. And this is a really important part about having a globally diverse footprint. It's not enough to have a mailing address that you have a fab in the United States. You have to have a fab that can serve a bunch of customers in different end markets, therefore, it needs a broad base of technologies. We don't want fab 8 to be 12-nanometer FinFET. We want it to be that. We want it to be RFSOI. We want it to be silicon photonics. We want it to be 22FDX. We want it to be 40-nanometer embedded memory for auto markets. 8.auto is part of the funding that will take our fab 8 facility and diversify it similar to what we do in Singapore and in Germany. And why is that important? We now serve many customers who want to use in the U.S. It's not enough to just say, I have a fab. The answer is what I build there, you don't need. And so that's the first part of that funding. Second part of the funding for fab 8 is when we get to that critical mass of needing new capacity, that funding will go a long way in closing the economic equation for investing in doubling this fab campus. And then about $130 million of that $1.5 billion is to reinvigorate and retool our 200-millimeter capability for wideband gap in Burlington with a really strong play on GaN. And it really makes sense to do that there because 6-inch technology goes to 8-inch technology, you get their productivity improvement, and that's our play for both power and RF for the future in that site. And then the last thing I would tell you, while the CHIPS bill is tremendous, it's a 15% CapEx offset. The ITC adds another 25% that's already in existence and states participate, so when you're all said and done, you're getting somewhere north of 40% offsets to go create capacity and capability. And that's what the overall funding program is.

And there's a lot of focus on CHIPS Act for you as a U.S. domicile company, but you also have a lot of capacity in Dresden, in Singapore and then also eventually in France, can you talk to the subsidy environment in those regions as well?

I think everybody is being very competitive and really the offset has to be proportional to the cost difference to operate and to create capacity. So if it's an area where a part of the world where to add capacity because of regulatory environments or because of the cost of labor, those offsets have to be higher to make it globally competitive. So if you look at what we get in Singapore, it's not dollar for dollar, we compare. We say, how do we make sure that it creates the right global competitiveness. And so we still continue to work and have expansion plans in Singapore, working in close partnership with the EDB. We have a European CHIPS bill that we will participate in at the right time there. And the key is to grow our global footprint in a thoughtful way, making sure it's always about certainty, durability and profitability, as don't add capacity because someone is giving you money, make sure there's a real need for it and to do it in a way that's balanced across the globe and we have redundancy of capacity. So it's truly resilient, right? It's not like mailing address only capacity, it is capacity that can be funged across the planet.

Great. And I assume that subsidy environment happens because your customers are asking for it, like your end customers want to be partnered with you, want you to have that capacity and not have to be reliant on Taiwan or things like that.

I think there's more and more of a recognition of the importance of supply chain resiliency. It's beyond geological, geopolitical and single point of failure. There's just too much at stake in the semiconductor industry to have such a high concentration. But resiliency doesn't come for free. And so it's a balancing act for our customers to figure out how do we go create the resiliency, but still do it in a way that's economically efficient. And that's where government subsidies to help out. Look, I think the analogy to our industry is going from fossil fuels to clean energy. We all know we need to do it, but the realization is, it's not going to be free. It's going to require investments, but it's for a long-term good. And for our industry, it's going to take investments to create global resiliency, and it's not going to be for free, and we have to figure out how to do it at the best economics for everybody.

Great. Last year, right before this conference, I think you had inked a deal with General Motors. And I still -- it's a year ago, I still use that deal as an example of how important your role is. General Motors 3 years before that probably didn't know who GLOBALFOUNDRIES was and now is kind of actively working to get a capacity commitment on behalf of fab suppliers a complicated thing. Maybe you can just talk to that. And is there -- are you seeing that OEM pull-through from other companies in autos or other areas?

Yes, I think it comes in many different flavors. That was a very unique deal. Actually, it was a 3-way partnership with a design company, GF and General Motors, but the overall and overarching point was, how do they plan their supply line well into the future. This is a long-term agreement that shipments wouldn't start for 3 years and then it will be a 10-year deal. And the auto industry is unique in that. It can -- it has those types of timelines. And we've seen deals similar to that, but have the elements of how do they lock up long-term visibility and commitment to supply, but to do it in a way where there is some flexibility. We don't mind giving flexibility as long as it's symmetric, right, where you can't have it where we're all committed to something and make investments and have to be ready to ship where our customers are not there. And so what we're finding is -- meaning the customer is not committed to it. What we're finding is customers want to start early, create a framework of boundaries of things that minimum this, maximum, as they get closer to when they really want the capacity, honing in more and more on that gap between what we're committed to do and what they're committed to do is very small. And that's some of the model fine-tuning we're doing as the industry is getting more accustomed to creating supply chain security through long-term agreements.

And that really showed up in your numbers. I mean you nearly tripled your automotive business last year versus the average semiconductor company and autos that grew about 10%. So I know we're going through a bit of a correction in automotive inventory as well, but you clearly have a lot of momentum there.

Yes. And I think the story that you don't see is a lot of that business that started to ship last year began in development 10 years ago. We have 3 versions of embedded memory on 40-nanometer. That's auto grade. Two of them are customer-specific variations of that, that we did co-development to go to the marketplace. And one of those was that a deal we just talked about earlier with Infineon that we announced a long-term agreement earlier this year. And so the reason why that's an important backdrop is our customers, we have to be ready with technologies for our customers who think about their roadmap 3, 5 years out. So we have to be ready 3 years ahead, so customers are willing to take the risk to put it on their roadmap so that they can do the work they need to do to actually productize it. And what we're learning is what makes me excited today is we're seeing more and more of the sockets we win today through partnerships. We say, what, 3, 5 years from now, we know what that looks like. And what's going to feed us between then and now are all the things we did 5, 6 years ago to get positioned for that. It's a very sticky business, the kinds of markets we're playing in and as we keep remixing our business. And automotive is a key play. You called it 2022 -- 2021 was under $100 million. 2022 was $375 million, last year $1 billion. And even today, as we sit here, knowing there's some inventory and a little bit of pressure on unit sales in cars, we still believe we have growth in automotive this year on a $1 billion base from last year.

That's great. And I think underlying a lot of what you're talking about is a key theme from, again, predating the IPO is you're innovating a lot in these nodes. You're not at the cutting edge nodes that people talk about, but you talked about GaN, you talked about specialty auto processes. Is there still opportunity for you to push forward in that area?

Yes. There's -- I think the best way for us to know and to be successful there is with our customers. And I cannot tell you how many customers talk to us. You've got to save us from single-digit nanometer. It doesn't fit into our -- we don't need the cost per transistor. We don't need the performance. More importantly, the volumes for the breadth of different products we want to have don't support the design cost and the tooling cost. How do we work together to make sure that I, customer have relevant products in these end markets and not being forced to go to a very expensive technology. And for us, the opportunity is to continue that differentiation. What does it mean? We have MRAM today on 22-nanometer, really auto-grade technology. But guess what, for next-generation smart mobile devices, we need magnetic immunity. So we need resistograms. So we need to go develop that technology and then spread it across our portfolios. We are not started the R&D programs that will add features that customers want. Our job is to make sure that we down-select in deep partnership with our customers to have the right ones in place.

Great. On that note of technology going to single-digit nanometer, there seems to be some anxiety now about that. And I know you've talked about there were always some programs that we're going to do that, particularly on the compute side. But now that we're sort of in a loose or cyclical environment, people worried about more aggressive pricing on single-digit nanometer that the TAM starts to shrink. Can you address that concern?

Yes. I don't think the TAM is shrinking. And if you first start with we're not over-indexed on that capacity on our 12-nanometer FinFET. So we're not worried about filling millions of wafers a year type of capacity. I think what happens when transition is placed in the down market, the rate at which opportunities move, which you knew were going to move because we were on that scaling time, it can outstrip the rate at which you could bring technology in. And if this same transition was taking place in an upmarket, we wouldn't even be having this conversation. So what are we doing? We're going to continue to invigorate and invest on the 12-nanometer program platform to make sure that it's -- has a great future. Recent deal announced between UMC and Intel, I think, demonstrates there's demand for 12-nanometer in the future, right? So when things migrate away, other things migrate in, but it's got to migrate to the right capability. We need to accelerate that. But we also have the ability to take that 12-nanometer technology and broaden its capability. We could do more 28-nanometer. We could do 22FDX. These are all technologies we're bringing to that same site. And our customers now seeing the success of 22FDX are asking us to start working over the next couple of years to create a 12-nanometer version of FDX. Same lithography node, same tools, just a different way of creating transistors and capability. So I'm bullish on the capability of 12-nanometer. We just -- the balancing of things coming in and coming out, sometimes is not perfectly matched at any given time and it's more challenged in the down market.

Yes. Okay. That's helpful. Can you talk about the competitive dynamic? You mentioned the Intel -- UMC Intel Power relationship, so there's other people that are sort of looking to alleviate geopolitical concerns around that. TSMC is adding capacity, China is adding a lot of capacity to do something. So can you talk to generally what you see from a competitive dynamic?

Yes, I think there's the competitive dimension and how does GF navigate through that. I think the biggest part of it is you have to understand our strategy. We are, by definition, single source differentiated business. Even our 28-nanometer node is built different than our competition. We're gate-first technology, they're gate-last technology. Why is that important? When customers design on our platform, it's very difficult to design away from it. Now the other end of that is if they've chosen a competitor's platform, right? It's hard for us to get into that business, but that's why we need to start early in this process and win on our platform. Now that platform creates a moat around our business. The capacity that's going to go into China that I believe, by the way, is going to take a very long time to become usable capacity. Having a checkbook, right, in either a warehouse or a factory to put tools in does not mean you're a foundry. And you could see as Intel tries to become a foundry that it's a long haul to create all of the PDK, standard cells and libraries, foundational complex IP. It's a lot of work to create something that's a design environment that marries to a technology that can be actually manufactured. So this semiconductor manufacturing is not about getting a bunch of tools, getting them connected, run a high-purity clean room in Europe, you're a manufacturing partner. That's like the furthest from the truth they can be. So what does GF do? Keep differentiating. If capacity does come on in China, it's going to be aligned probably most likely not to GF, but to where the bigger market opportunities, TSMC, that's the UMC plays to be aligned with TSMC and GF continue to grow its business, making sure we're delivering real value to our customers on differentiated business. So it almost becomes isolated from the greater capacity that's put on. And then the last thing I'd tell you as much as China wants to have a local for local capacity, fabless companies in China who want to be global players, need global for local. So they want to be a local player with international footprint. They need capacity to tell their customers, don't worry about geopolitical, I'm servicing you out of Singapore, the U.S. or Germany and they're coming to GF. So we may not be able to give the local for local, but we can give the global for local in the China story.

Interesting. Okay. Great. Maybe you could talk about the long-term supply agreements and how that's evolved over -- as we've sort of headed into a downturn. How do you -- seems like you've used those to give yourselves some visibility, how do you sort of make sure you're also using that to cement customer relationships going forward?

So the reason the LTAs became such a big deal, you have to go back to 2021. It wasn't so much, just capacity reservation. It was about making the investments then if I go full circle to where we started today, this is why we have, at the end of 2024, 3 million wafers worth of capacity is because those LTAs were to go get the confidence to go make those investments to have that capability. And with the customers' commitment to go use that. Okay, we go through a cycle, what are we going to do? We have to work together. How do we go make sure that we preserve the economic intent of those contracts, but also create long-term win-wins for our customers. And I think one of the reasons you saw our business in this cycle have less amplitude to it is the nature of those relationships with our customers. We've used them now, in many cases, to offset near-term decrease in business with longer-term commitments for new design wins, for example, how do we make sure that this capacity gets put to good use. Now we're in an environment we are today where we're in a cyclical downturn. Even in that environment, we signed a long-term agreement with Infineon in January. But I think for 2024, the rest of this year, you'll see probably less of those deals because now it's more how do we use the capacity we have, we're not investing. And what customers will look to do is do more capacity reservation, then let me go and co-fund or find a way to create the right economics to add more capacity. So I think it's a better way of doing business because it creates certainty for GF and GF's customers in what we need going forward together.

Great. So the last question for me, and then I'll turn it to the audience for any questions. What are your priorities for 2024? Obviously, the cyclical recovery will or will not happen as you and I are thinking about it. But what are you trying to achieve? Where are you with the customer relationships that can be bolstered. How do you build your pipeline of business for the next decade this year?

Yes. You almost answered the question. The priorities for this year is continue to navigate as we've done in this downturn. I talked about our performance of top line being down less than everybody else. We actually grew margin with underutilization. All that means is when we start to get this business back and utilization comes up, we have the ability to grow profitability. We're going to -- in 2024, even though we're going to spend a healthy amount of CapEx, we're going to be 2 to 3x grow our free cash flow from the $320 million we did in 2023. So it's really important that we demonstrate that the sustainability of this model that when we're not investing in capacity, we're throwing off free cash flow and we're doing it, whether it's an up cycle or a down cycle. And then when you make the business for the upturn is all the business you drive and win in the downturn. And for us, that's not only sockets for 2025 and beyond, it's also creating those partnerships for the technologies that we're going to develop that don't ramp for 5 years from now and staying very focused on that with our customer gains and making sure the R&D dollars we spend reduce the R&D dollars our customers need to spend to do products with us.

All right. Very helpful overview, thank you. So if there's any questions from the audience. All right. If not, then we'll wrap it up there. Tom, thank you very much. Appreciate it.

See you in a year.

All right. Good to see you.