Corning Incorporated ($GLW)

Good morning, everyone. Welcome to the fireside chat with Corning, and I have the pleasure of hosting Hal Nelson, who's the SVP and Chief Operating Officer; and Edward Schlesinger, who is the EVP and Chief Financial Officer. So thank you both for coming to the conference, and thank you to the audience as well for being here. Maybe it's a good time just given that you just did your investor event. So we'll focus a lot on that in terms of the discussion here.

You just did the investor event on May 6, where you extended your springboard plans through 2030 with a $40 billion sales target. Can you just help investors understand the building blocks and sequencing as they think from your trajectory from the $20 billion revenue to $40 billion revenue?

Yes, thanks. So first of all, thanks for having us. It's great to be here. Thanks for everyone attending. So maybe before I answer that question, just to go back for those that maybe haven't followed -- about 2 years ago, 2.5 years ago, we launched a growth plan. We called it springboard -- we were about a $13 billion company. Our operating margin was around 16%, and we had put a significant amount of capacity in place across most of our businesses. And we had expected to fill that capacity grow sales and improve our margins. we had set a target of growing sales, about $8 billion from that level, so getting to -- from $13 billion to $21 billion by the end of 2028. And we've clearly been tracking well ahead of that plan. . A lot of that growth has come from our enterprise business, which supplies passive optics into the data center space. Additionally, we've improved our margin. We're running at about a 20% margin today, and we did that a lot earlier than we expected. So that's kind of where we're at today. And as Samik mentioned, we held an investor event. Hopefully, some of you listened in or attended. If not, it's available on our IR website. We expect our sales run rate, which today is probably around an $18 billion run rate, $20 billion by the end of this year to go to $30 billion by 2028 and $40 billion by the end of 2030. And we -- I would say there are sort of 3 things I'd call out. First and foremost, the enterprise business, which has driven a lot of the growth to where we are today. We expect that to continue as data centers continue to get built out, larger data centers, larger clusters, scale out driving a lot of that growth. We'll start to see some scale up in the enterprise space as well in this time window. And I think that will continue to drive the enterprise business at a rate of growth significantly above the rate of GPU growth somewhere in the 1.5x the rate of GPU growth is what we would expect certainly for the next couple of 3 years. The second thing is if you take the rest of our businesses and you kind of just put them all together, we expect to actually generate a pretty significant amount of growth, call it, mid-single-digit growth -- our carrier business in the optical space, we expect that growth to be driven by data center interconnect as well as fiber-to-the-home. We're building a solar business. I'm sure we'll talk a little bit about that. That will drive some growth. We also have our Gorilla business and our Advanced Optics business, which is part of what we call glass innovations now, and that will drive growth. And then lastly, we're starting to see the early phases of photonics or co-packaged optics near package optics. We expect that to start in this window of time and to accelerate through the 2030 period. So that's what really drives the growth. And we put out that plan. We also put out a high confidence plan where we expect that $40 billion to be somewhere between, let's say, $35 billion and $40 billion by the end of 2030.

Okay. Great. That's great -- you are now extending to 2030, the Springboard plan and taking a forward look to the end of the decade. Clearly, one of the questions that sort of investor minds is what's driving your confidence to forecast that sort of time per out, right? When you think about the AI investment cycle, to have the confidence that it sustains in 2030, what are the indications you're getting from your customers? How much forward visibility are you getting to give you that confidence in the 2030 time frame.

Yes. I mean, I think all the market projections that you all follow and we follow hyperscale CapEx continues to go up, certainly for the next several years, larger clusters, NVIDIA has talked about going optical in their network, certainly at least to start in a hybrid way. And I think that underpins a lot of our confidence. The most important thing is we've signed a number of long-term agreements Three, in particular, with hyperscalers. We talked about the one with Meta in detail back in January, and then we just signed a long-term agreement with NVIDIA. So a lot of our confidence comes from what we're seeing in the market space, what we're hearing from our customers and what our customers are willing to do with respect to signing up for taking capacity that we'll put in place.

Yes. I think I might also add that when we announced these plans and these expectations, they're based on a history of us having work with a preferred seat at the table of our key customers, where we've been spending time developing those products. We've been spending time preparing to scale out the supply chain. And when we have high confidence and agreements in place, that allow us to move forward and communicate externally, then that's what we do. .

Okay. Okay. So maybe on that front, you did have the announcement with NVIDIA recently. And maybe to get more details around the sort of around the agreement itself. Can you flesh out what is the total magnitude of the capital support that NDA is willing to provide, I think, some of the release that you had talked about a 10x connectivity capacity improvement increase, 50% increase in fiber capacity. So how should we think about the amount of capital support that the customer is willing to provide on that front?

Yes. So as you mentioned, we announced in the NVIDIA agreement that we're adding 50% to our U.S. fiber footprint, and we're adding 10x to our connectivity footprint -- and a lot of that is underpinned by demand from NVIDIA. NVIDIA is actually providing a multibillion dollar prepayment to support that capital deployment and they're making an equity investment. They purchased through pre-funded warrants, 3 million shares. They have the ability to purchase up to another 15 million shares. So I think that funding will support the capital. They've also committed to minimum commitments take a lot of the capacity we're putting in place. And I think most importantly, when I think about these agreements is that we have a technology partnership as part of the agreement, and that allows us to understand their technology road map to continue to innovate. And as Hal mentioned, we didn't sort of show up and sign the agreement. We've been working on this relationship and the product sets and technology that we would need to be able to supply the supply chain for quite some time.

Go. How exclusive are these agreements? I think one of the concerns as much as everybody likes to see the customers sort of provide capital support, one of the concerns has been NVIDIA is engaging with everyone on the optics side and locking up capacity, which might exclude some of the other companies. So how engaged are you with other semiconductor companies similar to NVIDIA? And are these agreements exclusive? .

Yes, we have the ability to supply everyone in, I'll call it, the OEM CPU space or server space. So there's no exclusivity in that respect. We do have specifics in each of our agreements that are exclusive for the particular customer, but we're -- we'll be able to continue to supply across the entire supply chain.

Okay. Okay. So we should expect customer diversification to play out as it generally has been with you historically.

Yes.

Great. So maybe moving forward with the announcement you also had with the hyperscaler engagements. I think a couple of weeks ago when you had the investor event, you did update us that you have now 3 different hyperscalers, including meta, which was the one you first announced for capacity ramp dedicated to scale out fiber. Can you outline the size of the 3 hyperscaler engagements put together? How are these agreements similar or dissimilar to each other, the ones -- the 3 that you've signed?

Yes, they're all similar size and duration to the meta agreement. We typically defer to our customers in terms of what we go public with. So we let them drive that discussion. Meta wanted to do that. So we had that event back in January, we actually had a factory opening event I think it was end of March. So you may see more public announcements come out from us. But generally, those are the -- we're looking to sign longer-term agreements that allow us to have this technology relationship in there.

Okay. Okay. So when we think about the capacity increase that you need to do for these hyperscaler customers relative to what you've now promise to do on account of the NBD agreement? Do they -- are these sort of overlapping agreements and it's the same capacity increase that's been underwritten by these customers? Or do these separate capacity agreements in the sense you do a separate 50% increase for NVIDIA and the hyperscaler capacity increases on top of that.

Actually, I think it's pretty simple and straightforward. I mean as we've shared, we've been deploying capacity and making those investments for some time, particularly as we talked about last year, and it's across all dimensions of fiber, cable and connectivity, and it's particularly focused at those customers where we have the long-term agreements in place to ensure that they're providing some commitment to that capacity for us that we deploy, and then we'll be able to deploy that capacity for the hyperscalers to build out the optical connectivity and be able to supply the components that NVIDIA needs for their compute and their architecture. And so it's as simple as that.

Okay. Okay. Moving to the other big item that you discussed at the investor event, the 10 billion photonics or as you call it, MAP, right, by 2030, you're some expecting also some of that map to materialize in 2 -- can you highlight firstly what's driving the incremental confidence in the overall $10 million map? And how to think about the likelihood of the realization that you're thinking in 2027 specifically, which is a lot more near term than medium term? .

Yes, sure. You're right. We did make an announcement about our new Photonics map. And of course, the reason we did was because of the confidence. And we haven't just come upon that. We've been working on that for some time with critical customers. And you saw 1 of the announcements from NVIDIA. And so what the industry is expecting to see with the build-out of CPO, NPO, co-package piano packaged optics is to move first into scale out, right, architectures and then it will find its way into scale up architectures. And that's really because what those customers are seeking is to find lower latency, higher faceplate density with all the connections that need to get made, lower power and particularly higher reliability. And so what you really have to decide and think about in that $10 billion opportunity is, first, who's moving in scale-out and who's moving in scale up to go inside the box on both of those fronts into what extent. And our confidence really stems from the customer agreements that we now have in place. the demand pool that is being placed on us, and you saw a little bit of that with the announcement from NVIDIA on their commitment of what they're doing and what they want us to do for them.

Yes. One other thing I would add, we put out a plan in time because that's just how it works, right? We gave you a, call it, a 5-year window, 26 through the end of and we made a lot of assumptions in there. And when you have a mature business and you're seeing a wave of growth, whether it's secular or otherwise, it's a little easier to predict the timing versus a complete change in the way something is going to work. So in this case, you're changing an entire supply chain in a window of time. We have high confidence that the supply chain is going to move. NVIDIA has announced that they're going to go optical. I think you're going to see others follow over time. But again, we picked the window to give you this plan. So there's definitely a lot of variability around that photonics opportunity. So first, we sort of derisk that plan a little bit by taking some sales out in the '28 and 2030 period. And a lot of that risk adjustment relates to the timing of how you see that inflection in Photonics. But that goes the other way, too. It certainly could be faster than what we have in our plan. I mean we are just trying to give a plan that we think has a reasonably high confidence so that you guys can underwrite that. And I think we'll start to see that in the 27 time frame, and we'll know more -- like every quarter that goes by, we'll know a little bit more about how the entire supply chain evolves because it's not just Corning, it's not just NVIDIA. There are a lot of other players that have to make all of that happen. And we'll be able to provide you with a little bit more certainty. But we feel good about the magnitude, and I certainly feel good about growth well beyond 2030, but we just try to give you a shorter-term window to think about.

So -- and let me just follow up on that. you have a revenue contribution that you're expecting in 2027, does the shipment start in 2027 or this year? .

I think it's probably too early to say that, but I would say, assume more 2027.

Okay. So in terms of what you've done previously with your springboard plans, for example, is as you've got closer and more visibility, you've reduced the risk adjustment. So in terms of this photonics map, your higher confidence really builds probably once you start in 2027 with those deployments, and we can potentially see some of that risk adjustment coming down to your high conference .

I think that's right. I think we know investors are really interested in this space. We're obviously not the only supplier who's talking about it. You'll learn more I think about the market at large and what NVIDIA and other OEMs plan to do, and we will certainly share our perspective as we move through the year.

Great. So we discussed a lot about the volume ramps and what you're doing with capacity, Investors are also focused on pricing. So maybe as you referenced the 1.5x, just help us go a bit more forward, that's more scale out, how do you think about the longer-term content opportunity as you move into scale up and then we can go into more pricing-based cash. .

Yes. Okay. Let's do that. Let's start where you sit outside the box with scale out and scale up opportunity. And that's really mostly in the dimension of what we consider our enterprise business. And what we tried to do was provide to your point, some expectation of our growth rates relative to GPU build-out, right? And so the way we think about it is if you're thinking of where we are now, in its simplest form, we could grow linear with GPUs. However, there are a number of technical trends that we think drive us to have a more accelerated growth relative to that and a couple of them I'll mention. Number 1 is really on the scale-out dimension that you alluded to. We see data centers moving to a higher and larger cluster. -- size, right? So that growth rate is pretty substantial. And where we really begin to see a break point where the industry sees a breakpoint is when you begin to cluster greater than 130,000 GPUs. And the reason that becomes important is because you move from what is a current sort of 2-layer scale-out switch configuration with the leap in the spine to a 3-layer configuration with a leaf and spine and a super spine. So if you just went simply with that in simplified mathematics, we would grow 50% more when you see cluster size that go greater than 130,000. And right? And we see that, and we talked a little bit about the percentage of that. That's sort of number one. Number 2 is where we really pay particular attention to is bandwidth, right, like we all do. And so what we have historically seen, right, is bandwidth for GPUs and for the ASIC has historically doubled every couple of years, right? And then -- so we've got to figure out how you make those connections and between the lane rate or lane speed and the number of lanes, right? And with the serves driving that, we've made some assumptions about where that goes and what it tells us is that we're likely neutral to positive with the amount of fibers and connections that occur through what people are now seeing with GPUs and switch bandwidth and SerDes and so forth. So that's order number two. Then number 3 is what you alluded to is scale up. Today, scale up is essentially 100% copper, right? We have no space there. But as we talked about earlier, people are moving. There's already been an announcement from NVIDIA publicly that they said they're moving, right, to scale up. in a hybrid form. And so what you see is the opportunity to take those nodes and make a larger node, and that provides a substantial amount of connectivity for us. That is really 1 where you kind of have to decide back to Ed's point around what do you believe about the timing of when they're going to deploy scale up? Who's going to do it? When are they going to do it? And how deep are they going optically? Are they going from 0% copper to 100% optical, where on that continuum are they going to be? And those are some of the assumptions that you need to make in your modeling. We've develop our own view and particularly have insight from what our customers are telling us they want and need from us. I think when we talked about all of those 3 dimensions, what we see as a particular line of sight more in the near term that tells us, you put all those together, and we think that we can grow at somewhere between 1.3 to 1.5x the rate of GPUs. And then you go on the that's true 28. And then on the other side of that, we've made some assumptions, but there's some potential significant upside considering the factors that I talked about if the deployment is greater. So those are the drivers. That's the reason that we're seeing the build-out and the customer agreements that you're seeing us talk about or the pool that is validating what I just said.

So you do have this opportunity around content growth on a per GP basis. Outside of that, I mean, at this point, we can see demand is pretty strong. every part of the supply chain is starting to get constrained. How should investors think about opportunities around like-for-like pricing relative to content growth. And historically, this was an industry where you didn't really get as much pricing on the same technology. So how -- is that going to change at some point in the future?

Well, yes, pricing. I think that we are seeing pricing opportunities here given the strong demand. I think we see the pricing opportunities, though, most prominent when we're bringing forth new innovative products, right? And the reason that is so is because -- our intent is to make sure that we develop products that are valuable to our customers, either they are enabling them to meet some type of performance criteria or they're enabling them to provide a lower total system cost for whatever reason, we want to bring value to that customer. And in that case, we think that's when it's appropriate for us to share the value. They could benefit we benefit because what's important is what Ed alluded to earlier, is that we want to establish a long-term relationship with those customers. which gives us a preferred seat at the table when they're designing their products, when they're designing their architectures and they're telling us about the problems that they need help solving. And that's when we create value, and that's what we are good at. I think that's sort of how we think about pricing. It's a relationship-based staying with creating value. Secondarily, on the pricing front, I think the other thing that -- where we see value is mostly along the price cost margin dimension when you think about us being an integrated supplier in making fiber cable and connectivity, all of ourselves. And in that case, that allows us to buy from ourselves, create additional value and opportunity in the margin space there.

Okay. Just a quick follow-up there to the first aspect where you're looking at fiber and connectivity but maybe selling that more disaggregated to your customer rather than a full solution an innovation-led pricing is where you see the opportunity? Do you see more fiber and more connectivity products?

I think we actually see it across all 3, right? And I think you see it fiber cable connectivity. And everybody is moving to a more dense solution for a variety of reasons, right? So you have in fiber where we're creating smaller diameter fibers, right, lighter weight, more de in cable, higher, higher number of fibers per cable, and that's bringing potential amount of value because you can then put multiples in the DUC that you weren't putting before. And then on the connectivity front, we're seeing the need for more connectivity on the face plate so we need denser, form factor connectors. So of course, all of those dimensions, I think, we're innovating to bring value.

Great. So maybe just trying to wrap up here on optical. You have talked about the DCI opportunity being $1 billion by the end of the decade with 3 industry-leading customers that you're engaged with. Firstly, has that ramp met your own expectations? And should we be expecting a rebuild of the long-haul networks eventually materializing at some point because we all know sort of these investments in the network go in a more cyclical fashion. Are you already planning for those with your customers?

Yes. So we're public with 1 of those customers, that's Lumin. They actually talked about our arrangement on their earnings call. I think it was back at the beginning of February. So we've extended and upsized that arrangement with them. I think in general, it's met our expectations. I certainly think the $1 billion is a very realistic run rate. Maybe it happens before the decade, maybe end of the decade, maybe it happens by the end of the decade. And certainly, it could be larger. I think we are definitely seeing parts of the long-haul network get rebuilt, whether it's a full on rebuild and how sustainable that is probably still as all of these large data centers get put in place.

Okay. Let's move to solar before we put on out of time because I definitely want to discuss that as well. Last year, we outlined a $2.5 billion solar revenue target by 2028. Now you've raised it at the investor event to around $3 billion. What's the primary change in that driving that outlook? How much of the raises that you read ramped better than expected versus I think when you gave that target, you were not really in the module business, which you're now in? So just help us break that down. .

I can do it. The -- for solar, I think you're correct, we are seeing some strategic momentum on the commercial front, on the policy front, on the demand front. And just to ground everyone is that there are sort of 3 components or segments of our solar opportunity. There's the polysilicon, there's the wafers and the modules. And with respect to polysilicon -- we just put in incremental capacity last year to support the build-out that's sold out, and that's creating value already. So we're really pleased with that. On the modules, you alluded to, yes, we actually acquired module operation last year. We are now up and ramp that capacity to a place where we're sort of nearing where we expect it to be. And it's performing now -- and so between polysilicon and modules, both of those businesses, I'd say, are creating incremental revenue and performing at or above our corporate average for margin and profitability. Then you go to wafers, wafers is a little bit more complex. And if you think about it, wafers, we started, well, 2 years ago, kind of 2 years ago this time, it was dirt. right? So in that period of time, we have built a factory 1.2 million square foot factory and have ramped that produce our very first wafers at the end of last year and are continuing that ramp as we speak. And we expect that ramp sort of to continue through the back half of the year. And when that ramp continues the product is sold, and we'll begin to remove the arms of those incremental costs that we're experiencing due to the ramp and the build-out. And so we'd expect to see that improve through the back half of the year. And so I would say we're quite optimistic about what we've seen in solar and are seeing it in sort of a better light than when we first launched into it with the original projections.

Got it. And you outlined the trajectory of near-term margins, what you're seeing in the individual parts of it. But your longer-term margin aspirations in this business and particularly how dictated is it going to be by legislation or policy and tariffs as well in terms of how you think about longer-term margins.

Yes. Legislation tariffs policy all play a role. And I would say, so far, what we're seeing is tailwinds for pricing with respect to those elements. And really, it comes down to right upfront, there's a preference for domestic supply. And the primary reason there is a preference for domestic supply today is in order for you to access the 45x incentives, those are maybe many of you are familiar, the 45x are the production tax credits that have been provided for producing solar in the U.S. to receive those credits, you also have to adhere to peak regulations or foreign entity of concern requirements. And so with that preference towards U.S. content, we're definitely seeing tailwinds for being a U.S. manufacturer there.

Okay. Okay. You have talked about the U.S. manufacturing being a key differentiation as well. wheels talked about it quite a bit. Yes. clearly shows an optical what you're announcing an agreement is helping in solar as well. Where -- what are the other areas you would highlight where it is turning out to be a differentiation and would be a tailwind to your financials?

Yes. Actually, one of my favorite topics is our advantage from a manufacturing and advanced manufacturing standpoint, both globally as well as in the U.S. I know your question was specifically to the U.S., your question was specifically outside of optical and solar, but I'm going to emphasize a couple of points on optical and solar before I get your question because I think it's important to understand and give perspective for our manufacturing prowess. If you think about optical first and foremost, and Ed mentioned this before, is that we have the largest lowest-cost manufacturer of fiber in the world in the state of North Carolina. Soon, we'll have the largest cable manufacturing in the world in the state of North Carolina. So a very prominent position there. In solar, we talked about what we were doing, but what I would also pivot to is that in solar, we -- I mentioned we're providing the polysilicon. What it just so happens we're also providing the polysilicon for the semiconductor industry and semiconductor wafers. We are the only U.S. flag and 1 of only a handful of companies in the entire world that can produce polysilicon for semiconductors at the quality levels necessary for them to perform. And so when you take both of those businesses, optical and solar that you asked me not to talk about, we are building out in North Carolina. We are building out in Michigan. We are building out in Texas. We are building out in Arizona, right? Okay. So now to your question right? The other areas -- we also have other proof points. We talked about before the announcement that Apple made just last year for us to produce 100% of their watch and their cover glass for their iPhones sold anywhere in the world, and that is going to be produced in the state of Kentucky at Harrisburg. And so we're deploying that advanced manufacturing platform and capability there. Great example. I think another area that flies under the radar a little bit for us is in the semiconductor space. We make a lot of precision optics for photolithography for photomask, et cetera. We're building out that capacity in New York State, for example. And so I could keep going. But I guess what I would say is that we have 34 manufacturing platforms across 15 different states in the U.S. And I think that's a great example of our ability to manufacture in the U.S., reshore manufacturing here and compete on a global stage.

Thank you. Thanks for the details there. So maybe before we run out of time, just going back to the spring board plans and taking a more longer-term view here. It feels that you're migrating to a different phase of SpringBoard, where the first phase of springboard, you don't really require that much heavy capital investment to support the growth, which you had highlighted when you initialed rolled out those plans. It feels like going forward, there's going to be a sizable investment capacity to now pursue that growth that you're outlining. What changes should investors expect in this transition and how should we think about free cash flow conversion during this sort of second phase of springboard.

I may start and then you can jump in. I think that the way to think about it is that from a capacity standpoint is what we're trying to do with SpringBoard in this next phase of accelerating growth, as you said, is to make sure that we continue to enhance our financial profile of the company while we double it, right? And in order for us to do that, that means that we need to execute these long-term customer agreements that we've talked about so that there is a sharing in the risk and sharing in the cost. And so what we intend to have those agreements do is to create a higher degree of assurance of both the revenue inflow and the funding of those investments to support it. And if we do those things with that capacity deployment, then we're going to deliver higher revenues with cash following it.

No, that's great. And the thing I would add from a financial perspective is we really like the profile we have, and we want to keep it and make it better despite wanting to invest and grow. And maybe just a couple of highlights. We grew about 15% -- the last few years, we've got a growth rate projection of about 19%. So we expect our growth to accelerate. We moved our operating margin from 16% to 20%. We expect to be at or above that 20% and even while we continue to invest. We moved -- the 1 that I'm most proud about kind of the level of passion and how has on manufacturing is return on invested capital. We were 10% or so. We're now mid-teens we'll get to the high teens for sure. We might be able to do better than that. And if I think about the best value creation a company like us can have is when we can deploy capital and get a 20% return on that and grow the company at almost 20%. And -- that's a great value creation mechanism. So we want to maintain the profile we have today and improve upon that even though we'll have to do some investments. So we'll derisk those investments by getting some cash upfront from customers or other arrangements, but we'll also look to ensure we price right, we move up the value chain. We sell more solutions, we improve our mix and things like that, which we'll continue to improve that profile. Ultimately, maybe a simple way to think about cash flow conversion because that's really what matters in the end is that we'll improve our conversion because most of the incremental net income will add and net income should grow faster than sales, that will convert to cash at almost 100%, right? So if we're converting that incremental income at close to 100%, then we'll move our overall conversion up, and that's kind of the model we want to run in this growth phase.

Got it. Interesting. Great. We've run out of time. I know there are a lot more things to talk about, but yes, thank you for coming to the conference. Thank you to the audience as well. .

Thank you.

Thank you.