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

Okay. Why don't we just go ahead and get started. First of all, thank you for your time after lunch session, another session. And we're very fortunate to have Wolf management with us. We got Elif, CTO of Wolf; and Tyler. I'm sure a lot of you guys have interacted with Tyler over the last few quarters. With that, let me just get straight into it.

What do you think are the key hurdles for Wolf at this point for revenue growth? Or is it just -- I mean, are there things happening in the marketplace that matter? Or is it just basically Mohawk Valley fab ramping at this point in time?

It's mainly ramp related. We have a really strong demand from the market actually for the visibility that we have for the next 10 years. What we are working is to ramp the fab, and then that has connection for us as a vertically integrated company all the way into the substrate. So we have made a decision several years ago to build the Mohawk Valley fab on 200-millimeter. That is the first 200-millimeter fab for silicon carbide. And then what we are doing right now is ramp our factories to make more silicon carbide substrates of 200-millimeter, and that's going to support the fab. So we're marching towards the plan of get to that incremental 20%. From there, another 20% and all the way to 100% between the wafer factories in the past.

So Harsh, for those that are not super familiar with the company, today, we make 60% of the world's supply of silicon carbide. And we do that out of Durham, North Carolina. The next closest competitor makes high teens supply, and that's also a U.S.-based provider. The reason why I point that out to you all is because we've been doing this for about 35 years. We've been -- we were the first vertically integrated company in the industry. If you go back a couple of years ago and look at some of our rivals, they maybe mentioned silicon carbide like once or twice on their earnings call. If you look at them now, they're probably 10 to 12 times. And the point being is this, is that there is a shift going on in the mid- to high-powered application, and that is silicon carbide. And so really, as Elif points out, this is something that we were working on for several years. But like the legacy technology is 6-inch wafers, where the game is going to be played is at 8-inch, and we are ramping a fab in Upstate New York on 8-inch wafers. So we will -- we're in production. We generated about $1 million worth of device revenue last quarter. That's going to continue to ramp. We expect to be at about 20% utilization at that fab in Upstate New York by the June quarter of next year, and that puts us on a really nice trajectory.

Can I put you on a little bit of a limb and ask you how that ramp? So 20% is $100 million, I think, is what it [ constitutes ]...

A quarter. Yes, that's right. Yes.

Per quarter. So you'll go from about $1 million in June of 2023 to $100 million in June of 2024 out of that fab, not necessarily revenues. But can I ask you how that ramp will look? Is it pretty linear? Or is it more cautious in the first few quarters and then scales up pretty [ high ]?

Yes. We're ramping in Q1 and Q2. So think of it this way. We're qualifying products right now. And I'll keep me honest here. We've qualified our largest device for customers, okay, and we've qualified our most complicated device. So we've done a couple of devices already qualified. We still have more in process. And that's why you really see the first half of fiscal '24 a nominal increase in overall revenue. And as Harsh points out, you'll see this kind of inflection point beginning in the March quarter of next year and then -- where you'll see a significant step up of revenue, and then you'll see that again in June. And as Harsh points out, at 20% utilization, that puts us on a run rate of about $100 million per quarter. But it's going to probably take us until the December quarter of next year to get $100 million in a quarter out of Mohawk. And that's just because of the start time that a wafer goes into the fab, goes through a test [indiscernible] inspection and comes out of packaging. So that's kind of the way to think about the trajectory.

I'm glad you brought that up because I just want to -- help me clarify this for investors. You have $100 million worth of wafers coming out, not necessarily revenues, $100 million of wafers coming out by June 2024, but then you have to test package. And so what would be the lead time by the time you get to -- so it's about 1.5 quarters to 2 quarters is what the lead time is. Okay. And your margin should inflect very nicely because this is the big one that you've been waiting for and working for. How should we think about your gross margins ramping as you go from where you are today to $1 million at Mohawk Valley to closer to $100 million out of the door in terms of wafers?

Yes. So for the Q1 guide, the midpoint for non-GAAP gross margin is 14%. And the way that we're thinking about it, if you get to 20% utilization between now and the June quarter, you're probably going to see gross margins move to about low 20s. So you see about a 6- to 8-point improvement over the next couple of quarters. And that's really a function of just moving more volume through Mohawk Valley. So we kind of feel like you're on the trough or, we call it, kind of the bottom on gross margin for Q1. And then you're going to see this stepped-up improvement as we move along.

Okay. And how should I think about -- so let's say it's 2024, June. You're doing $100 million now in terms of wafers. But at some point in time, it's going to get ramped and it will be fully ramped. How -- what is the profile of the first 8-inch global fab fully automated? You put a lot of work into it. What can I -- what should I expect out of gross margins out of that fab when it's fully humming and fully going?

Yes. I mean the target gross margin that we talked about is in the low 50s. And if you think about that, at full capacity, Mohawk Valley will do about $2 billion annually in revenues. And there's about a -- there's a flow-through of about 60%. So it should throw off about $1.2 billion in cash. And that's probably -- I'd say that's our expectation, but Harsh knows this. This is the first 200-millimeter silicon carbide fab built. So that's kind of the target of what we're getting to, which is kind of what you would see in traditional semis. But there's possibly opportunity just because of the level of automation. Do you want to speak a little bit to that?

Right. So currently, we compare our processes to our existing fab in Durham, and that's an old fab that we've been building and working through our history. That's also a very manual fab. So you literally take the cassette from -- we've been -- we kept upgrading the tools and the working condition, but we still have the humans taking the cassette from one tool to carrying to another. So with that, on a daily basis, that we have thousands of steps that the human touches to a certain wafer every day. When you look at our yields in the -- so silicon carbide is a very demanding, difficult technology. It's in the maturity curve for us at Wolfspeed that we feel really comfortable understanding material behaviors for almost 40 years. But then the material is also very hard, but at the same time, it's brittle. So when you combine the human interaction and the manual handling, one of the highest yield issues that we see in the Durham fab is actually from breakage and then from that impact. So our initial testing as we qualify the New York fab is it's fully automated. We see much better yield compared to Durham fab just because of the handling aspects of it. This is really great news for us because when you think about the material going from 150- to 200-millimeter diameter with the brittle behavior of the material that you expect higher breakage. So we do actually see a much better line yield [ in the fab ]. So from there, yes, we have our -- the anticipation in terms of the gross margin, but we'll constantly keep working [ in an upside ] as well. So we will see that in time.

So you guys have like very ambitious plans. You've got this fab at Mohawk Valley, which is now about to get going. You've got a fab in Saarland -- am I pronouncing this correct -- Germany, which is going to be even bigger than this one, I think $2.3 billion or $2.6 billion fab. And then you've got a wafer expansion, which is about $1 billion, if I'm not mistaken.

$1.5 billion, yes.

$1.5 billion. So you've got a lot of CapEx going on. That takes money, right? So in terms of funding, how do you guys feel about where you are? And is your funding "I can do an Elon Musk" because your funding is secured?

Yes, I would say that -- and Harsh knows that last October, we talked about a capital plan of about $6.5 billion between now and the end of 2027. In the last 10 months, we've raised $5 billion, okay? We've done that in 3 ways. We did a convert deal that was headlined by BorgWarner, major Tier 1. They bought $500 million of the convert deal because they wanted to have -- think of it as a little skin in the game. And there's an -- that got them by buying into the convert. They also got a capacity corridor for several years, which would generate a lot of device revenue for us. The next thing that we did, we did a deal with Apollo for $1.25 billion. And that was an opportunity for them to break into semis. And it was also a way for us to kind of secure this $1 billion of capital that we wanted to raise between now and the end of this calendar year. But then the third thing that happened was fairly unique, and that was the Renesas deal. So basically, we cut a deal for a 10-year wafer supply agreement, which had a $2 billion upfront customer reservation deposit. So a way to think about how that works is we drew the first $1 billion this year and put that on the balance sheet. Next year, we will draw that second $1 billion. But that's like a -- it's a 10-year term loan at about 6%. And once again, that bought them the opportunity to get a capacity corridor or a wafer supply agreement underneath that. So that's probably worth several billion to boot. So $5 billion in capital, we're fit for purpose on cash right now. There's still probably a little bit more to be raised, but the way you got to think about that is we have U.S. CHIPS Act funding that's still pending. We have European Chips Act funding that's still pending. There's possibly some other customer deals -- or customer financing deals that could be done there. So I think the heavy lift on the cap raise has kind of been done, and now we're going to just be opportunistic to optimize the cap structure over time and maybe look to raise that last [ sub ] amount that we've talked about.

Okay. So you've got a lot of things. I mean you talked about the Saarland fab, this fab. You've got the wafer stuff going on. If I put it all together and try to arrive at when you guys might become profitable on a non-GAAP basis, what would that look like, Tyler and Elif?

Well, I'd say this. Think about operating cash flow sometime in fiscal 2025 and then think about free cash flow generation in fiscal 2027. And we've talked about a couple of hundred million in fiscal 2027 for free cash flow generation.

Okay. So getting there. A lot of people were confused when -- so for lack of a better word, MVF was delayed by a couple of months, call it. And a lot of people got confused thinking that you might be having technical issues, or there are some other things going on that weren't very obvious. Could you help us like demystify a little bit about...

You mean like challenges...

Yes, challenges that you ran into that...

You got the resident expert here. So Elif is the expert on that. I'll let you talk.

So we actually look back and do this analysis a lot internally as well. So I'm going to take a little bit step back. When we decided to start building the Mohawk Valley fab, it was right at the beginning of the pandemic. So we still made the decision. We didn't blink, and then we kept building the fab. Our intention in the plan actually was such that the Durham fab will migrate there into that fab in time to create more capacity or space for making the substrate. But we are facing such high demand from the market, and then we also chose to continue to support our customers from the existing Durham fab. So that actually pushed us to create the different solution. So we end up doing that we have this additional building in the existing campus that will be called Building 10. We [indiscernible] that building to turn into the crystal growth factory. But then it came, of course, that the process itself is extremely challenging, that you need to heat it up to 2,500 degrees. I keep repeating this because I want you to imagine what it means from the infrastructure perspective, from the reactors and an equipment perspective. It's not an easy job that you go to a shop and then you bring in tools. You need to be really like the special infrastructure, power supplies and then the cooling infrastructure and all that. So that came with a lot of permits for us. So we went through all those challenges. But then once we actually qualified and make the building ready, we're actually demonstrating a really nice and steep increase with the 200-millimeter volume. So the rest is going to be move that material through our internal supply chain and then start feeding the fab. On the fab side, we're still being cautious that the way we bring up the fab that we have the first line. But there are some steps that we need to have the second tool. We're working very closely with our close partner customers that we don't want to let them down, that we don't want to start the fab fully and then have a hiccup. So we're waiting to get the tools in place and then just put everything in their places. So from the quality and the yield perspective, things are going really well for us, and then nothing -- no issues or delays. But then we've been dealing with the external equipment and -- yes. And then again, this is primarily coming up. We work with our external partners and suppliers to educate and bring them up because it's such a special and difficult technology.

Okay. Last night when we were at dinner, we talked about China wafers. You've got a very strong position in silicon carbide wafer. And we're lucky that we have you because you're the expert in that area. Where do you see the China Inc. for 6-inch wafers? Do you see them? Are they good? Is there a yield problem? And secondly, have you seen an 8-inch wafer out of China at all at this point in time? Or it doesn't exist yet?

So I'll talk about the 150 part a little bit that we see in here and read a lot of volume out of China. And then this goes back for me, actually, as I research for the source materials, raw materials that China has been reporting for good like 10-plus years that we have excess silicon carbide capacity report that. But then silicon carbide traditionally, especially coming out of China, is for the abrasive industries that you make. It's the second hardest material they produce with the Acheson method that silicon carbide to be used in grinding, polishing of the abrasive industry. So seeing the market potential, those companies, they convert their high-temperature equipment to crystal growth. But when you talk about -- think about the semiconductor, the quality part of it is rather difficult to attain. But over the years, we see good improvement on the 150-millimeter diameter. But then a good part of it is we ask for feedback from our customers and more so in the Schottky diodes range, but not so much on the automotive quality and the consistency. That's on 150. In the communications and the trade shows, we see demonstration of 200-millimeter wafers from China. But if you think about that, for us fully into silicon carbide coming with 40 years of experience, we demonstrated 200-millimeter back in [ 2015 ]. So when you think about that, we're now in the volume ramp phase. So there's still like [indiscernible].

Okay. So is it fair for me to walk away from this thing that 6-inch wafer, the 150-millimeter out of China, is still tough? Is that a fair statement? In volume because...

Yes.

Okay. So still pretty tough.

I think what we will see, we'll see a lot of volume from China. The yield will be a question, and the performance of that material will be a question as well as the consistency.

We'll see wafers but poor quality, maybe tough yield situation. Okay.

How do they perform in the fab and then more so, it's the customer will be a key is the question to answer.

Give us a sense of -- let's just say I fast forward. I have a time machine. I went to 2028. And I said, who would be standing around in the silicon carbide world? Would it still be U.S. Inc.? Would it be still like guys like you and ON? Or would there be other people in Europe and China that you see or even Japan that actually might have wafers, might have fabs, might have like product?

I think you're right. So we will -- we're a pure play into the silicon carbide. We were the first. We're going to continue with our strategy. We have a really strong position in the history, actually 10 trillion-plus hours of field reliability on the product. It's just that -- it brings a huge confidence actually from the customers. So we're going to keep building on that. And then you see a lot of the -- all the players that you [ count ]: ON, ST, ROHM, Infineon. They are silicon companies, right, silicon power electronics companies. And then they are all, of course, seeing the opportunity. First, they entered into the field as our customers. So we've been actually supporting all of the market with our substrate to kind of like establish the power electronics. And then you see then, they're trying to build their vertical integration as well because seeing the value of it. It's going to take time, but I think they will be around because they're -- looking at their history, I predict that they will be around.

I think it's really encouraging. We've been talking with the U.S. CHIPS office about our grant applications. But one of the things that you have to kind of think about, think about how hard it's going to be to repatriate silicon semiconductor operations from outside of the U.S. back into the U.S. And here, you have a market leadership position in silicon carbide. We have over 70% of the world's supplies made here in the U.S. So as we sit here today, we think it's a great investment for the U.S. government to help us build our operations and build out capability. But if you just think about the automotive industry, they've kind of burned the boats on internal combustion engines. They're going to electric vehicles. So we've gotten -- we petitioned the U.S. government that silicon carbide is now on the U.S. geological critical mineral list, which means it's a matter of national security. So like there are things that we are trying to do to help drive greater awareness for silicon carbide, which I think will keep -- certainly keep us at the top of the pyramid, so to speak, for some time.

That's great. And we have a couple -- we have 2 minutes more. But I wanted to ask you, give us -- so you've done silicon carbide for like your whole life. Give us a sense of how hard it is to make wafers. I mean we're talking about temperatures incomprehensible, 2,500 degrees centigrade. So give us a sense of like why this is so hot at a fundamental level. Why can't somebody in Europe make this in Poland or some of the countries that have experience with silicon wafers? Why can't they get to this place of making silicon carbide wafers?

Thank you for that. And I can actually talk a lot about that, but I'll try to keep it short, because I did my PhD studies on gallium nitride, bulk crystal growth. So compared to that, it's such a hard technology, gallium nitride. We were modeling the silicon carbide, which is already very difficult. But specific to silicon carbide, in my mind, the majority of the difficulty is coming from the temperature requirement, the 2,500 degrees. So a lot of us in this room, we don't even have the understanding of what does that mean. So the room that you're sitting is maximum [ 25 ] degrees Celsius. So imagine that. So the silicon technology, if we compare, it's about 1,500 degrees. So think about that. It's still 1,000 degrees different. I'll tell you one thing that the lava temperature is 1,700 degrees. That's very important because imagine that 2,500 degrees, how do you even get to that temperature? How do you accumulate that heat body? And then, of course, you still need to control your temperatures and then pressures. So you need your electronics to work. But then back to the 2,500 degrees, that lava is 1,700, 2,500 is actually like halfway to the temperature of the surface of sun that you can't even look at it. It's just matter is like blowing. It's not solid anymore. It's kind of like the [ stockpile ] environment. So we have this founding technology. The founders in the company are still like senior scientists. What they're doing like extreme fundamentals. They're like calculating what's going to come out of the run. They do all the calculations and the run comes out, the crystal growth, that it just doesn't work. So at those temperatures, at such high entropy that the thermodynamics and kinetics rules, they don't apply anymore. So what we've been doing, the way -- the reason I keep pressing on the 40 years, you rely a lot on the empirical results. You literally -- this starts with, "You try this one." "Hmm, this didn't work." "How about try that one?" "Okay. This didn't work either." They keep doing that over and over again, but then learning from that and building -- that is the empirical knowledge. But then doing that over the years, what we've also done, build on the knowledge and the data. So when the thermodynamics and the kinetics don't work, we rely a lot on the statistics. So for statistics to work successfully though, you need to be very like data-rich. So we rely on our [ rules ]. So it is -- so like it's just not one thing. It's an accumulation of a lot of difficulties, challenges and us finding the solutions to that. And it's basically that. It's just a little bit too much chaos, on the verge of madness, but creating something really exciting like this is...

Awesome. Thanks, Elif. And thank you, Tyler. Appreciate the time guys.

Thanks, Harsh. We appreciate it.

Thank you.