Wolfspeed, Inc. (WOLF) Earnings Call Transcript & Summary

Good morning, everyone. My name is Matt Ramsay, I'm the semis analyst at Cowen. Really, really happy to have Neill and Tyler here from Wolfspeed. And here's my shameless plug. Many of you probably know that we've been spending a lot of time in Cowen's research team on compound semis and silicon carbide in particular and tried best to put together a pretty comprehensive couple of hundred-page report a couple of weeks ago. And so it's really a big area that we're investing in, have a compound semis panel with one of the leading industry analysts in a couple of hours in this room. So if you're interested in this topic, come back. But obviously, Wolfspeed is a leader by far in the silicon carbide space, and there's a lot going on. So thank you guys for making time to be with Cowen and with our folks today, really enjoy seeing you. And Neill, if you want to kick off and maybe have a couple of opening comments. I know we've just gone through earnings. We -- there's a lot of things going on in the industry. So if you might want to open and set the stage, and then we can go from there.

Sure. So I think most people are familiar. So Wolfspeed were a high-power semiconductor company focused on silicon carbide. And if you look back over the last 3-plus years or maybe even a little bit longer, as the electric vehicle adoption rate has really ramped up over time. Silicon carbides become kind of a critical component for the inverter inside of the electric vehicle. As we've seen that ramp happen over the last several years, it's really created a big volume opportunity for compound semiconductors and particularly silicon carbide as it relates to the drivetrain, particularly inside electric vehicles. And that volume ramp, what that's done is not only has it been great for compounds and silicon carbide from a market standpoint, it's also provided the scale to start seeing the cost come down where it's opened up a lot of different markets in industrial energy through the sales and distribution channel and a lot of different applications. So we're seeing really tremendous growth. What we've talked about recently is our design ins and our pipeline over the last 6 months has increase dramatically, probably more than we had ever anticipated, I'm sure, we'll get into that here a little bit during the session. But we've just seen a really large pickup, a step function pickup, a higher from a design and a demand standpoint, and frankly, anything we ever really anticipated. So for some demand standpoint, that's going, I think, really, really well. And now it's really all about capacity. How do we get reliable capacity, not just for us but into the industry to support what's really a high-growth opportunity and really a game changer in the power electronics industry.

Thank you for the introduction. I'll just say, we're pretty informal here. So we want this to be a discussion. I don't want to -- you guys have to listen to me enough. So if you have questions in the audience, just wave your arms and we'll get those addressed. So anyway, just to kick off, one of the big conclusions that we had and the work that we did as a team in the last few months is that our own view is that silicon carbide demand is going to exceed supply for the industry through the end of the decade. That might be controversial to some, but I think that we're pretty convicted in that view. I wonder: a, if you share that view; and b, if you go to TSM or Samsung or any of these folks that have ramped multiple generations of silicon, they know what a yield curve looks like, give or take, like the whole industry is sort of bootstrapping silicon carbide to scale. And just thoughts about the risks? I mean, you guys are investing a ton as are some of your peers/competitors, and we'll get into that. But is the industry investing enough? Sort of what are the barriers to really come up that manufacturing yield curve for a brand-new process and a brand-new substrate?

Well, first of all, I'd say when you look at us -- well, let me take a step back. If you look overall in the industry, I think what you're saying is potentially true. I think that when you see -- as I said earlier, that demand step-up has been so large, is there going to be a supply and demand disconnect over time. I think that's probably likely as you look over time. I think that's why it's important that as we look at -- as we look at silicon carbide ramping from a capacity standpoint, this is very early days. And I think having all of these -- whether it's our materials customers or rivals in the device business bringing on capacity, we think that's really important. In fact, we plan for things, we plan for everyone to be successful in bringing up their ramps because it's so important to go do that. So number one, I think your assumption certainly makes sense, Matt, when you look at it from that perspective. For us, internally, though, when you look at silicon carbide, the number of people who actually know how to do this in the world, if you look like the number of actual silicon carbide material scientists or engineers, it's actually very few. And the one thing we have in our favor in that sense is that we've been doing this for a long time. In our prior version of the company, we had an LED business that was based on silicon carbide substrates for a long time. So we have a pretty good idea of what it looks like to ramp this material over time. And we've done it over 30-plus years. Our founders still work at the company, helped develop this technology a long time ago. And as we've ramped our technology over time and we've expanded the diameters of our substrates, we've seen good occurrence of how that can work. But that being said, it is a very difficult technology to work with. It's not easy to do from a material standpoint and getting yields and getting crystal growth kind of sorted out from a scale standpoint is not easy. And I would say even from our perspective, we've been doing this for a long time, as we move generations, it certainly -- is difficult. So I think when you put all that together and you say from a capacity standpoint, it's certainly going to be something that's going to be bringing reliable supply to the industry, I guess, I'll say, will be probably the biggest challenge over the next 5 to 10 years.

Matt, yes, for the benefit of the group, just to put this into perspective, we talked about this about 2 years ago. If you look at the last 20 years, of silicon carbide production. We've made more than 90% of the world's supply of silicon carbide in the last 20 years. We've got about 60% market share today of the merchant market. But hats off to you and the team, Matt, for the CapEx chart that you put in this report that Matt talked about because what you should really be paying attention to is who's putting the money to work within the industry. And there's a couple of people we've put over $1 billion of CapEx to work in the last 1.5 years, 2 years. But I think, Matt, the thing to pay attention to is investment. And we're talking about now essentially doubling the capacity that we announced back in 2019. And that's really because of the pull-in of stuff that we thought that was coming at the end of the decade is coming a lot faster. And that's where, if you think about the device opportunity pipeline, I joined the company about 3 years ago, it was about around $9 billion. And most recently, we talked about the device opportunity pipeline now being at about north of $25 billion. So that kind of speaks to the opportunity.

Well, thank you for that. I made my shameless plug, but the plug from you makes it a lot better. So speaking of supply and investment, Mohawk Valley, congrats. There's a lot going on, and it's good to see ribbon cuttings and exciting opening ceremonies and things. But and I have to get up there and tour the places and excited to do that. But maybe you can just talk about the plans for that facility and how quickly you can bring it online? Like what are -- as you think about full-scale production, what's on the to-do list? And what are the -- what things do you think will be easy to get off that to-do list? Why do you think there's huge hurdles that you have to get over?

Well, first of all, I think things are going well. I think if you look at where we -- what we've been doing, -- we've had our 200-millimeter wafers, both from a crystal growth and wafering standpoint, running for well over a year. We've been running those through a pilot line. And so as part of our deal partnership we've had with the state of New York, we've been running those wafers on a pilot line. So we had a very good result on that pilot line. And essentially, what we did after that is we took that pilot line, and we now -- we've deconstructed it and we've installed it into the fab in Mohawk Valley. So -- and now we're not running it today. We're running it on pretty much the same exact tool set. So we have a very good idea of what the yields need to be, what that ramp rate looks like for those -- for that technology over time. So we feel good about where we're at with Mohawk Valley. We've had the ribbon cutting. We are now running wafers in the fab, which is good to see. I think from a revenue standpoint, we'll start to see revenue start to ramp you want to think about this kind of March, June quarter, kind of next year in terms of when we start to see benefits from seeing that ramp start to take off from a revenue standpoint. But we do feel good about the underlying economics of what we're seeing, both from the 200-millimeter wafer and from the fab. As it relates to investment and as we look forward, inside the fab, we had a partition up. And our original plan was to kind of slowly bring the fab up, match the revenue run rate we talked about at our investor update last year, which was about $2.1 billion by 2026. And now the plan is we're going to take that partition down and really build the fab out as fast as we can. So because if you look at Mohawk Valley, more or less, if you take a look at the design ins we've had, some high-probability opportunities, you save some space for some industrial and energy opportunities, it's more or less sold. We don't have more capacity in there. So we have to move very, very quickly. Now the challenge with that if you go back and look at the material supply, if we had a partition up in the middle of the fab, that's about -- talk about half the fab was full from both a device standpoint, but also from the material capacity that needed to fill it. If we fill the whole thing is then we need double the materials capacity to fill that as well. So very quickly, we're going to figure out how to get the partition down, fill out the rest of the fab and also bring on materials capability to fill the fab as well. So that's really, I think, the next step. We talked about having a second material site in addition to another fab. So that's really the kind of the next step, Matt, is that we're going to change the plan. I think the trajectory of what we're seeing, not just from a design ins standpoint but the visibility we have on revenue now, we've got a high degree of confidence in and is really now a race to bring on supply to meet what's an increasing demand curve over the next several years by building out Mohawk Valley, bringing on a second material site and then having another fab behind Mohawk Valley as you get out beyond that kind of '26, '27 time frame.

So is it equipment availability, raw materials availability, yield. I'm just trying to -- as we try to scale here on sort of the list of worries or hurdles, I don't know what the right word is to get across. What's the list?

Clearly, on the supply chain side, there's challenges, right? I think we all know about this. That's something we watch closely. We've had some success of managing through this in Mohawk Valley. We haven't had a lot of issues. We were able to get our tool orders early that's all going well. We're going to continue to do that. We talked about raising our CapEx rates up this year. And really, that's around getting longer lead time tools and faster than we initially anticipated. But if you start thinking about new fabs or new material side, there's also a construction time frame. Now from our planning perspective, I think if you were to build a fab, Mohawk Valley a little over 2 years ago was a field of mud. Today, we're running wafers in it. So a little over 2 years to get to where we are today from a field to mud. We did a new fab today that would probably take longer, you're probably looking closer to 3 years just because of the supply constraints. But we've been able to manage through that effectively. We're looking far enough ahead and out into the future. So from a construction and tool availability standpoint, I think we've been able to effectively manage it. But I think those are the type of things you got to watch out for inflation and supply chain constraints. And I think that's pretty much where we're sitting right now.

Yes. I mean nobody has built a silicon carbide fab facility of this size and scale. So I think the turn up -- and we're being thoughtful about how we're kind of turning up capacity there. But I think -- so execution is going to be critically important and the thing to watch. And I think the other thing to listen for is the narrative, not only from us, but from the rest of the rivals within our group here about what's happening with customer wins. Because I think what we're all kind of seeing is we're trying to build Mohawk Valley out, I mean it's essentially sold out. Isn't it fair to say that?

I think so. And then let me just add one thing as well. As you look at Mohawk Valley, the blueprint that we've kind of laid out here, getting the partnership with New York, getting the reimbursements, bringing out the 200-millimeter wafer seeing how they're yielding, seeing the cost basis for the fab now start to come together, the blueprint in our mind of the unit cost economics around Mohawk Valley at 200-millimeter in an automated brand new fab, the largest in the world, and that's -- from that standpoint is very, very solid. So we feel really good about the blueprint that we're laying out. So as we start thinking about adding capacity moving forward, it would very much be -- take that same blueprint we've just kind of executed on, and we are executing now with a 200-millimeter wafer, the highly automated fab and then we'd really just do that again. So I think we feel good about what's the path. And I think as you get to the end of this year, we talked about bringing Mohawk Valley on through December of this calendar year. As you move into the first half of next calendar year, we have really good, I think, visibility to that. And as you start to see that come together, I think we'll start to see kind of how that blueprint plays out. And then it's really match up that demand with that blueprint, and that will be our game plan from a device standpoint.

If you bring on another fab, what you're talking about, do you need it in Europe? Do you need -- it's somewhere geographically diverse. What are the variables of picking a site here? Because there's a lot of geopolitical stuff that's going on in the industry where capacity is being built out or by whom?

Well, look, I think that we're leaving our options open there. And I think the #1 thing is how much funding can we get around the fab. If you look at Mohawk Valley and you look at the economics I just talked about, part of it was around 200-millimeter wafers going into a highly automated facility. In conjunction with the partnership with the State of New York, where we got about $0.5 billion of reimbursements related to the fab. I think we were doing another fab. We want to lay out that same type of road map, same economics, same type of reimbursement type of levels. Now the next fab we would do would likely be larger than Mohawk Valley, but you can think of the economics scaling to about the same size as what we're doing today.

Got it. That makes sense. At the same time, you're making the big push into the device space and on your own 200-millimeter. Your customers for 150-mil are -- keep taking targets up, which is great. I mean there are rivals in some sense, but they're also big, big customers. I was over in Paris a couple of weeks ago when ST pulled the $1 billion target in by a year '23. Infineon certainly has some ambitions to grow that business over time. ON Semi is doing their sort of vertical integrated own thing with GTAT, and we'll talk about that in a second. But are you also making the appropriate investments to support your sort of foundry customers from silicon carbide as they bring these targets up?

Well, look, there's no question of bringing up not just capacity on the device side, but we're bringing up our capacity on the material side as well. And what I can say is we talk about having unfulfilled demand, a significant amount, it's over $100 million right now, significantly more than $100 million. That's not just on the device side, that's also on the material side. So I think consistently with what you're saying, Matt, we are seeing a pull-in and a push for more materials from our materials customers as well, which kind of makes sense in terms of what you're seeing. And I think if you look at those customers, I think it kind of makes sense. If you look across the landscape of the industry from where we sit, our plans are always based on assuming that the materials customers that we work with, they're going to be successful with the efforts they have to bring on internalization of silicon carbide, which, by the way, is a good thing. The industry is going to need a reliable supply. We're very much in the early innings of driving this ramp with a significant amount of volume required. So from our perspective, number 1 is we'll take them at their word. They're going to do what they say they're going to do, that's great. we work closely with them on plans to make sure we can supply them to the levels that they need. And we think that's all fine and good. We'll continue to do that over time. Now the other side of that is we're not necessarily a backstop either. I mean, building a couple of fabs in a 200-millimeter potential material site and putting that all together, it's a pretty significant business investment. But if it came to that, we'd have to think about how we would manage if anyone wasn't successful in their efforts or needed more from us beyond what their capability of doing. But I think the overarching theme though is the industry is so big that it's really going to take all of us to connect to drive the capacity up over time.

Yes. And I think there are rivals, but anything that happens silicon carbide related, we think is great. We're trying to make a market here. And what we've seen actually in the last year is a substitution because customers could not get silicon IGBTs, they've made a conscious choice to move to silicon carbide. And I think that's why Matt, we're seeing ever pull in now. And we'll give an update to everyone probably later this fall on our longer-term outlook in terms of what's happening or what we're seeing, too. So I think it's just a matter of some of these market dynamics.

Do you have a sense or view on the cost of the inverter [indiscernible] on average, how much is and how it [indiscernible]?

Well, I think, first of all, I can't talk to the inverter that much. We do build modules, but we don't do the whole inverter. So I think that's a little bit outside of my scope of understanding. But I think the key here is, in my mind, as it relates to the inverter and the module as it relates to the overall vehicle is that the battery is -- regardless of battery costs are coming down, you talk about battery supply, I guess, it's another probably not a topic that we could take the whole hour on. But battery supply as well. The key here is that silicon carbide in the inverter is incredibly efficient. And if you think about the battery being the highest cost part of the vehicle, this is why silicon carbide inside the inverter makes so much sense because it can drive between 5% and 20% efficiency onto the battery. So as you think about supply over time. So first of all, the benefits could be a higher range between 5% and 20%. So you sold on carbide versus silicon, you get that range benefit. And right now, I'd say we're as I talk to auto OEMs and Tier 1s, that top of mind for everyone, getting efficiency out of their inverter and leveraging silicon carbide to do that. The second one is everyone's making billions of dollars of investments into the battery itself. So silicon carbide actually in a module and inverter based on silicon carbide starts to protect the battery investment. So if you think about it, you can either use that investment to increase your range or if there's a battery shortage, you can make a smaller battery, but still sell a lot of vehicles. It's actually a conversation we had with an OEM last month. They kind of come with a conclusion that it gives them a lot of flexibility to protect all of those battery investments. If you don't have the silicon carbide piece and the inverter sitting next to it. So I don't think it's so much around the cost of the inverter versus the total, but more about the efficiency and the assurance that the inverter with silicon carbide drives and protection of the battery, which is the highest cost part of the vehicle.

Another question on the wafer capacity. You talked about on the call that you're possibly going to fab and wafer. I was just wondering you talked about on the campus right now, there's [indiscernible]? It seems that you're completely sold out substrates right now. So kind of wondering like when possibly you'll start being able to produce at in Carolina? And then kind of the timing [indiscernible] you have more than that?

So the original plan around the $2.1 billion, excuse me, of revenue that we had for 2026 was to kind of slowly build up Mohawk Valley and then have enough capacity from a materials perspective in and around our Durham side, okay? So that was kind of step 1. So we are in the process of ramping materials capacity in our Durham site today to support all of this. So if you go back to our Investor Day, I think, we had a chart kind of showing on-campus. We've cleared out another building that's under construction. That's all going well. We'll add crystal growth capability there and wafering operations. We talked about on the call as well or back in January when we did the convertible offering. We would leverage that additional cash flow to kind of accelerate some of our plans. A lot of it was around bringing on additional capacity for materials, and that was both 150-millimeter and 200-millimeter. So there's still some ramp-up on the 150. There are customers that are looking for 150-millimeter materials and upsides to that, that we're looking at now and that's how we would fulfill that. The second piece to that then would be we're going to need -- as you think about that ramp up, if you look at Mohawk Valley, for instance, and I'll use as an example to get back to materials. The total build-out cost of Mohawk Valley round numbers was about $2 billion okay. The -- if you take the state of New York incentives, it gets you back to about $1.5 billion net, okay? And we were going to build about half of that factory. So what that means is by '26, if we're going to build half of it, it means that we have to now -- and then we're going to go to full, it means we need double the materials capacity over the same time frame. Now if you look at Mohawk Valley, the $1.5 billion net, the revenue capability was like $1.5 billion to $2 billion, okay? So then you got to go back now and double your materials capacity to go support Mohawk Valley in a much kind of shorter time frame and then be ready to build a second material -- sorry, a second fab on the back end of that. And then if that's say, double the size or more of Mohawk Valley, we'd have to double the materials capacity again. This is where the second material site starts to come in. You need a lot more material capacity a lot faster than we had before. And then from a CapEx perspective, it's actually -- it still remains a pretty good ratio. So the way we've been thinking about it is if Mohawk Valley net was $1.5 billion, and it could do between $1.5 billion to $2 billion of revenue. That's a really nice CapEx to revenue ratio, so it's below 1. However, when you add the materials element to that, it gets closer to about 2. So we're thinking about a 2:1 ratio for CapEx spend to support the additional, not just the fab growth but the materials growth. And we'll give an update later in the year in terms of how this all comes together, but that's kind of the thought process we're going through in terms of managing the investment. So I think from a CapEx to revenue or ROIC or whichever way you want to look at it I think it still gives us a really nice model that we can build out. It all depends on a lot of the things we talked about earlier, Matt, government incentives, working with customers, this type of things. So this hasn't been locked in stone, but this is really the direction we're headed. And the reason we wanted to talk about it on the call is so we can be more open about the level and the size and magnitude of the demand increases we're seeing and how we're thinking about kind of attacking that.

[indiscernible] the 2:1 CapEx to revenue ratio. So there's effectively a change that we'll have to spend less than $3 billion to $4 billion?

If you had a new fab, so if Mohawk Valley was 2 and then $2 billion, you just do the math or $1.5 billion net, you just do the math off that. That's right.

And that's over what [indiscernible].

Well, if you wanted to open a new fab today, it would take about 3 years you're probably talking about -- it was started at the end of this year, you're probably talking about '27 time frame.

What would you get back in subsidies?

Think about a similar -- now the fab will be bigger, but from a scale perspective, the blueprint would be very similar to what we've done with Mohawk Valley. So we got about $0.5 billion on a $2 billion spend. Think about a similar type of ratio.

Yes. And I think the other thing is what's changed since 2 years ago when we did the Mohawk Valley deal, is that customers now are coming to us. We've come up with the assurance of supply agreement, which we announced with GM. So there could be an opportunity aside from government funding that there could be some potential customer consideration funding too.

That seems to be the model of the ones that customers prepay to secure that capacity. I'm just curious, we have the sort of establishing out [indiscernible] customer base, right? So what's the potential to get briefings [indiscernible] to secure the capacity?

So we are working on that already. I remember on the materials, deals that we've done, we've gotten upfront payments related to expanding our materials contracts. We also announced the deals under our Wolfspeed Assurance and Supply program. We've worked directly with OEMs like we announced General Motors last year. We get upfront payments for them. So there are different avenues that we can work with customers on. But I can say that we are working with customers, I'd say, in a lot of different ways in terms of securing additional funding as well. So when you think from a funding perspective for doing something like this, there's a lot of different elements. First of all, we've got $1.3 billion or so of cash on the balance sheet today. I think we'd have to go reinvest our operating cash flow, we'd start to see sometime next year, start to reinvest that back into the business, and then look at government incentives, customer payments, whatnot to kind of chop that number down and then get to kind of a net financing number to support what we're trying to do. But the financing number, I guess, what I would say is a lot smaller than the total bill because that's the kind of the approach we would take.

Just to kind of follow on to that, the 1 thing that -- it's a little bit of a different angle to it, but the thing that I've been interested in is it seems like the decision to make like 150 label to external customers and keep the 200-mil stuff for yourself and your own devices, is that like a completely firm decision? And that's the way it's going to be? Or are we -- if there's upside to business and Infineon and ST and that's cash flow you can use to fund the expansion as well.

But here's a way to think about that one. So I get this question a lot. And I think there are some cross currents on this that happen from time to time. But I think we've got to take a step back and say, this is very early innings. We're just building the first 200-millimeter kind of supply right now as we speak and it's just coming up. The cost per millimeter squared of a 200-millimeter wafer today is higher than a 150 because it's early on. Now fortunately or unfortunately, we have a Durham cost footprint for our fab right now that is subscale, okay? We've taken an LED factory essentially, and we're building silicon carbide MOSFETs in there right now, which is not fully ideal though the factory is improving. And I think the team is doing a terrific job there. When we bring that to Mohawk Valley, that will have a much different view. But once you take even a higher-cost substrate on a cost per millimeter basis and we put it into our lower cost footprint fab at scale, we get the benefits for that. Now someone else is running a fab that's full or has been long for a long time or has been around for a long time. They're not necessarily going to see the benefits of that with a higher cost substrate today that would have to be marked up from us. So we're still very early on. I don't think it's like a commercial sale at this point from a 200-millimeter standpoint. Let's see how all that plays out. And I think that decision is probably still a couple of years away.

Yes. We were -- we had a presentation from ST in this room yesterday. And someone in the audience asked what I thought was a pretty astute question, which is in sort of the silicon industry, there's been a pretty clear delineation between design house, ASIC house, fab, back-end test and the industry been structured that way. And it seems like the silicon carbide industry is on its way to going vertically integrated as it gains scale. I mean we'll see if that's the -- if that's the end game, if any one sort of still not on board with that. But what's different about the market? Is it just that it's immature? Or is it that all the real complexity is in the raw materials and the substrates and then the designs on top are relatively simple? I'm just trying to understand the differences as this market gets economies of scale. Why vertical integration is the right way to go versus the silicon industry?

Yes. I think it says it's early days. I think it's because it's early days in the industry. And I think the play between -- the materials, if you think about the order -- the level difficulty in the process starting from the growing crystals all the way through the fab and your back-end operation. Crystal growth is the hardest one, okay? And the getting crystals grown, getting them to yield, getting them at high quality is very, very difficult. There's high barriers to entry. It's very, very difficult to do. And then as you start moving down the supply chain, as you said, Matt, it's start getting easier, but I would say designs on silicon carbide MOSFETs aren't that easy either, not as simple as you might think as we're dealing now with a very high-power device different from how it works with silicon. So the interplay between the material and the device itself is important in getting that feedback loop and trying to get these things to work. So I think over time, what you're seeing now is getting that learning cycle between the design and the material is an important element of this. We've obviously had a lot of experience doing this. Like we said, we've done the first silicon, so when did we do the 2012, I think.

2012.

So we've done like we've been doing this for a long time. So I think having that interplay is important. And for us, having that scale and understanding how those 2 things work together, is an important piece of the equation. And it's probably why people want to be internalized. I think also from an economic standpoint, I think you want -- you're going to -- given the cost of a substrate, it would make sense to have some of your supply internal. But again, it's very early days. So I think when things are this early in their lifetime, having them verticalized it just seems to make more sense.

Yes, because Matt, I think what we've learned in the 30 years that we've been doing this, is that the cycles of learning is where we're able to drive the cost down. So I think the vertical integration gives you great line of sight in terms of what Neill was just describing here. So I think maybe 20 years from now, it might look very, very different, but we think the vertical integration is really just a function of people trying to get smarter on silicon carbide.

No, that makes a lot of sense. Well, I want to thank you both for your time. We're getting the red light of death here from a little clock. But you folks are really interested in this topic, as I said, well, a couple of my teammates and I are hosting a panel on this in a couple of hours in this room. And thank you very much for the time and spending some time with us.

Thanks, Matt. Appreciate it.

Thanks, Matt.