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

Good morning, everyone. Thank you for joining. I'm Vivek Arya and I cover semiconductor [Audio Gap] If you have any questions that you would like me to ask on behalf -- on your behalf, please feel free to send them along. But with that, a very warm welcome to Gregg, Neill and Tyler. And maybe, Gregg, just to kind of set the stage, there are people in the audience who might not be as familiar with Cree, just because you have gone through a period of restructuring and getting into better and stronger areas. So maybe just for the benefit of this audience, if you could give us an introduction to Cree. And just a state of the union, right, as you see it, what are the top 2 or 3 things that you find most interesting about Cree and the industry right now?

Well, thanks a lot, Vivek. And I appreciate everybody's interest in Cree. I've been with the company now coming on 4 years. And joining the company, it was really a company focused on lighting and LEDs. And when I joined the company, I saw a real strong potential for the wide band gap semiconductor aspect of the business, the silicon carbide and the gallium nitride type. So what we've done is we've completely transformed the focus of the company from lighting and LED to now solely focused on silicon carbide, gallium nitride and compound semiconductors. We actually sold our Lighting business, we sold our LED business, and we're now a pure-play compound semiconductor company focused on gallium nitride and silicon carbide. Now the company's roots were actually in silicon carbide, that were founded 30 years ago, a little over 30 years ago based on this technology. And what we saw happening is kind of a convergence of the growth of that capability that we've had internally from a development standpoint, hit at the same time the transformation of the powertrain of the automobile going from internal combustion engine to electric. And the demand for silicon carbide and EV and the value proposition, if you will, for silicon carbide in an electric vehicle is very, very straightforward. For the same amount of battery, your car will have a longer range if you use silicon carbide over silicon. And that increase can be anywhere from 5% to 15%, depending on who's measuring it and different formats and so forth. But that's pretty substantial. And range, of course, in an EV is everything. And so companies that are coming out with longer range of EVs are selling them very nicely. And of course, we're seeing a very big transition in the industry from internal combustion engine to electric vehicles. Our kind of bogey for the middle of this decade is somewhere around 5% of the cars becoming electric by 2024, 2025. Most pundits are saying that's a pretty short put. But they're also predicting that by the end of this decade, at least 30% of the worldwide production of vehicles will be electric. So there's sort of this, Vivek, there's sort of this huge kind of secular trend, if you will, that we're playing in, primarily because of silicon carbide's efficiency improvement over silicon. Now EV is driving a lot of the demand. We're also seeing a ton of demand in the broader industrial markets. We're seeing solid face circuit breakers and electric surgical instruments. We've got an industrial mixer application. We've got a power supply application for -- think of it as sort of lumberjacks out in the middle of Canada somewhere in high-efficiency-type things, and so we've seen a pretty substantial amount of interest there. So it's a complete transformation of the strategy of the company from lighting and LED to silicon carbide. We're in the middle of a huge secular transformation and secular growth story for silicon carbide, it's super exciting. And we're really pleased with the direction we're heading.

Absolutely. Do you think we are at a point where customers are sure that silicon carbide is the answer, but -- is the answer? Or are we still at a point of debating the cost benefit of silicon carbide versus silicon or other wide band gap materials, such as GaN? So where does silicon carbide kind of fit in versus, say, silicon where the cost, right, for the last number of decades continues to go down versus, say, other kind of intermediate materials versus GaN? Are we past that debate? Or is the Cree working, right, on initiatives to help kind of customers get past that obstacle?

Well, I can give you a feedback that I get from customers. And we've been in this COVID lockdown now for 1.25 years. And despite that, next in 2 weeks, I'll travel to Europe for my third time in the last 8 months, and it's because customers are super interested in meeting with us about the transformation of the car from a silicon-based inverter to a silicon carbide. And what I would tell you, Vivek, is 3 years ago, there was a lot of cost benefit trade-off. Is it going to be silicon carbide? Is it going to be silicon, et cetera? And at the system level, there's been plenty of studies done that say, for every incremental dollar you spend on silicon carbide over silicon in a vehicle. For every dollar, you save $3.50 to $7 in system cost at the vehicle level. And that's at a silicon carbide cost that we're now below at this point. So I think the cost benefit trade-off has largely been answered. And I mentioned, this will be my third trip in the last 8 months to visiting with customers in Europe. I think that debate is largely over. And in fact, I've had one customer -- one OEM say, 100% of their platform will go silicon carbide. Another one saying, everything new that they're doing is going to be silicon carbide. So they just see a situation where you're going to get 5% to 15% more range with the same battery. And of course, batteries are the most expensive things in a car. If you can save battery, it's a big deal. And so I think it's -- I think that debate is largely over. And so now what we're doing is we're trying to convince the customers to use our silicon carbide versus somebody else's. And that's obviously a stronger position than walking in with silicon carbide versus silicon. So I would say, for the most part, customers have made that transition. And I think they're moving pretty strongly forward with that proposition.

Got it. I think the range argument you make is perhaps the one that really brings it to life. Is 5% to 15% range extension enough? Like what does it really mean, right? So give us a flavor for what the cost of silicon carbide replacement is versus silicon? And then what does that benefit translate to, right? So how much is the benefit to the end customer when they use silicon carbide instead of silicon?

There's multiple benefits. Number one is, for the same given range, you use less battery. And batteries are the most expensive thing, and you could take your cost per cell or cost per battery down pretty substantially. And when you eliminate 5% to 10% of that, it just writes out any cost incremental that you see in silicon carbide. And that's just the start of it. On top of that, your other components around the device, the inductors, capacitors, the other discrete devices around your silicon carbide chip in the inverter are smaller, which means the size of the inverter goes down. Space in a car is very, very premium because there's only so much you can have. The cost of cooling goes down. Silicon carbide can operate at a substantially higher temperature. It's much more efficient. So you've got cooling costs are down. So all of these things happen. There's another argument that's happening though with silicon carbide and that is silicon carbides advantages increase with higher voltage. So it's a progression as you get higher voltage, there's more of a benefit of silicon carbide over silicon. Higher voltage means faster charging. And so as customers want faster charging, they're going to go to higher voltages, and the higher voltages mean you got to go with silicon carbide. So that step added benefit that comes on top of that. So I think the car manufacturers have really concluded at this point that the cost benefit trade-off of silicon carbide is pretty strong. And what's really interesting about what's happening in the auto industry right now is the carmakers are making this determination. They're getting very, very involved in the decision of silicon carbide and whose silicon carbide, et cetera, because they see that advantage at the system level. And the inverter might cost more money, but at the system level, it costs a lot less. So a lot of the discussions we have are directly with the OEMs on our platform.

Got it. One other pushback that, Gregg, you often hear about is, well, silicon carbide sounds great right now. But as we saw in the solar market, right, as soon as there was enough demand, there was a lot of low price competition from Asia that came in. So maybe as a start, give us a sense for what the supply chain looks like, right, substrates, wafers, devices, systems, where are the barriers to entry the highest? And which parts of the supply chain does Cree play in or want to play in?

So it starts with silicon carbide crystals. So growing the silicon carbide crystals and then slicing them into wafers. And there's a pretty huge barrier to entry in just doing this. Cree, by the way, is about a 60% market share player in doing that for the merchant market. So we've got a pretty substantial market share, and we've got pretty good understanding of that issue. So first off, you have to grow the crystals. To grow the crystals, you need machines to do that, the machines, you can't really purchase on the market. You've got to create your own machine to create a silicon carbide crystal. Silicon -- so you have to know how to do that. The second thing is you need a recipe to do it right. Silicon carbide grows at 2,500 degrees C. And when you grow silicon carbide, it can form into one of several hundred different crystal structures. One of them works. So you're not only having to build these machines, but you're having to build and grow the silicon carbide crystals in a very specific way, which is not for the vein of heart. And by the way, when your machine is running at 2,500 degrees C, you don't kind of open it off halfway through and make sure everything is working. You either got it working or you don't. So then -- so now you're growing these crystals, you're trying to get rid of any of the noncorrect silicon carbide crystal structures. And now you're trying to grow it with very, very low defects. Hard to do as well. All of these things are difficult. Now you got to slice it. You got to slice it into wafers. Silicon carbide is the second hardest material on earth, diamond being the hardest. So even the act of slicing it is tough. And then you got to put epi on it, which is also difficult. So all of these things have enormous barriers to entry. We have a 60% market share. And what we're doing is we're using that market share to do long-term supply agreements and to grow our capacity and grow our demand. And as the demand grows, it gives us more cycles of learning. As we get a cycle of learning, it allows us to increase our yield and decrease our cost. As we decrease our costs, we're able to pass those on to customer. The demand goes up. As the demand goes up, we get more cycles of learning. As we get more cycles of learning, we decrease our costs. And so we've created a virtuous cycle, if you will, or kind of a flywheel effect, where we are driving these costs. So that's all on the materials side of things, and we've got a very, very large barrier to entry. And from a technical standpoint, and then we've added this barrier of they're chasing us down a cost curve. Now once you do that, now you've got to put it into a wafer fab and create a device. There it takes intellectual property in terms of designing the device itself and having the right kind of metrics and things like that, we tend to score very, very high on that in terms of device capability. What's been a weakness for us is actually our fabs themselves. Our wafer fabs are subscale, nonautomated, very antiquated fabs. I think one of them -- I think one of the fabs we have in North Carolina, we're the fifth or sixth company that's owned the fab. So it's just been passed on to different companies. And so it's kind of a hodgepodge of different things. And so what we're doing to fix that is we're building a new fab in New York. This is our Mohawk Valley fab. It will be the world's largest silicon carbide fab, and we're making a very, very interesting transition with that fab. We're actually going from 150-millimeter wafers to 200-millimeter wafers in that transition. So it will not only be the world's largest silicon carbide fab, it will be the world's first and only 200-millimeter wafer fab. That fab is nearing completion from a construction perspective. We actually have first tools, have already gone into that factory. Neill and Tyler were up there last week visiting and seeing it kind of in action. And we will actually be producing qualification lots in the first part of 2022, and we'll be qualified in starting production in the back half of '22. We anticipate that, that fab will actually drive nearly 3/4 of our revenue in 2024. So pretty fast ramp from no fab to 3/4 of our revenue in basically 2 years' time.

Got it. We will come -- we'll talk about the Mohawk Valley fab because I think that's kind of a key differentiator for Cree going forward. On this topic of kind of supply chain and the competitive landscape, a number of players in the semiconductor industry, Infineon, ST, ON, I assume they collaborate with you in certain places, right? There are customers in certain places, there are competitors in certain places. Then we also hear of companies in Japan or Korea also, right, trying to get into the substrate and the wafer business. So on the substrate side, you mentioned -- or on the wafer side, you mentioned you have 60% market share. How about on the device side? How do you think that market shapes up over time?

What I would tell you is our anticipation is that over the next few years, we'll grow both our wafer materials business and our device business, but we'll grow our device business substantially faster than the materials business. When we talk about our device, our pipeline of opportunity or $10 billion opportunity pipeline, that's all devices. And so we fully recognize that our customers and wafers are attempting to build their own capability. I would do the same thing if I were them. I think they'll probably find it's harder to do than they think. And so it will be tough, but we supply wafers to them, and it's fine. But I think from a device perspective, we're probably going to see a pretty substantial -- we'll see a faster ramp in revenue of our device because the fact that we've got an internal capability already, and we've got pretty good scale on the wafers, just gives us a natural cost advantage and, quite frankly, a supply advantage as well.

Got it. Gregg, I'm curious, you have been in the semiconductor industry, right, over a long period of time. Do you think the silicon carbide market can have more than 2 or 3 suppliers? Like is the size of the market big enough to attract a lot of competitors? Or do you think that just the capital requirements, right? Because look at how much is being spent on your fab, and that is after you have a lot of scale and experience, right, in this work. So how many players can actually be there in this industry? Maybe help us quantify what is the cost of being in this industry and building the kind of fab that you're building right now?

Part of the challenge with silicon carbide is there's certainly a cost element, but let's just eliminate that for a second. The real challenge is intellectual property and know-how. So you can't go down the street and buy the machines. There's not a whole industry of vendors supplying silicon carbide crystal growth machines. You don't see that. Most of the companies in this industry build their own proprietary machine. So you're all, okay, now you have to have the know-how to do that. Then you have to be able to grow the silicon carbide crystals in a very specific way, in a very controlled environment, have very high quality of crystal. This high quality of crystal has become more and more prominent as an important piece because of the advent of the auto industry. And it's not just the normal, well, the automotive guys want high quality. It isn't that. What it really is, is if your defect density isn't really good, the automotive chips are bigger chips than what historically has been in silicon carbide. The bigger the chip, the higher the probability, you have a single defect in that chip if your defect density isn't good. And so if your defect density isn't good, you can't be in the auto industry because you won't yield anything. And so it's one of these things where one of the biggest barriers here, it's a growth market, and certainly, there's going to be a lot of capital thrown at it. The biggest barrier is how do you spend that capital? Or what do you buy? Well, you have to get knowledge and know-how and all of this sort of stuff. So there's kind of this inherent problem inside the industry in terms of trying to enter the business because it's not like the memory business or it's not like the silicon business or it's not even like the semiconductor business in general where you could build a wafer fab. And all it takes is money, it just takes know-how, and that's been one of the key barriers to entry.

Got it. And the one other interesting aspect, we are all seeing the headlines around the automotive industry going through all these supply shortages. What do you think -- what's the change, right, in terms of the customer behavior as they emerge from this experience in terms of them wanting to work directly with the semiconductor providers? Because if I take your pick of a large automaker, and I promised my investors that 20%, 30%, 40% of my output will be EV, I better start lining up, right, those supplier relationships, quite in advance, right? Like how is the nature of your discussions changed, given the experience the industry is going through right now?

Well, that's exactly what's happening, and I'll hit that in a moment. But just remember that the supply crunch that's happening right now and the importance of supply chain, is happening in an environment where there's another huge transformation going on, and that is from internal combustion engine to electric. And the car companies really understand that this is a big deal to them. For many car companies, the engine is the identity of the brand, the Dodge Hemi. BMW's world headquarters building is shaped like a 4-cylinder engine. And BMW names their cars after the engine. 528 is a 2.8 liter, a 530 is a 3 liter and so forth. And so the car companies already were kind of thinking through engines going away, inverter is going to be the big thing. They need to be involved in that inverter because that's the heart of the solution, if you will, or heart of the vehicle. So there was already this huge interest in the car companies and the technology going into the inverter. Now you layer on top of that, hundreds of car plants have been shut down over the last 6 months. And you mentioned about my experience in automotive, in my 35 years in automotive, I never shut down a car plant. And I never saw a car plant shutdown. And that includes earthquake and tsunami in Japan, Philippines issues, typhoons, all of these various different things. The supply issue right now is kind of unprecedented. So you've got the car companies focused on this big transformation of EV, and then you got them hit with -- they can't sell a $72,000 car because of $0.72 silicon chip. So they're highly interested in the supply chain for EVs and silicon carbide. And that's where we kind of stand out because we have a very, very unique position in terms of the supply capability for silicon carbide. I mentioned that I've been to Europe a couple of times. Both Neill and I will be visiting with customers, I think we head out there, the 26 or 28 or something like that of June. We'll be in Europe for almost 3 weeks visiting with customers. And a lot of the discussion is going to be around supply assurance, capability, et cetera. And what I would tell you is, I'd love to be able to tell you 2 years ago when we decided to build a new factory, we predicted that there would be a supply crunch in silicon and that capacity would be a big care about. I'd love to tell you we were that smart. We're not that smart. It was total luck, I guess. But here we are visiting customers now and being able to tell them, in 6 months, we have capacity coming on board that, we're going to start qualifying that capacity in 6 months, and we'll ship it to them in the back half of '22. We'll likely be ramping our manufacturing process. It's world's biggest silicon carbide fab. We'll likely be producing that and shipping to customer before the supply shortage in semiconductors and silicon abates. So it's just -- I love visiting customers now because our story is pretty solid. We've got a new factory in place. It's not one we're going to build. It's one that's done building, and now it's just a matter of outfitting it and qualifying it.

Got it. So as you look at the next several years of EV adoption, I know you mentioned that by kind of the middle of the decade, your assumption is only mid-single-digit kind of adoption. And we have seen much bigger numbers for the end of the decade. What is the penetration of silicon carbide in that, right? I assume it's only in battery-powered cars. Basically, what is the right way to think about units and content? Like if we have to describe your opportunity in terms of units and content, what is a simple way to do that?

Well, the way I would describe it is a couple of folds. So first off, let me just distinguish between electrified and battery. So you got battery and you got hybrid, et cetera. Basically what the European regulators have really figured this out. There was a big strong incentive for plug-in hybrid vehicles in Europe. We don't play that much in plug-in hybrids. We play mostly in battery electric vehicles. What the European regulators have found is that people were taking and buying plug-in hybrids for the tax incentive associated with it. And then returning the car after 3 years with the plastic wrapping around the plug still intact. The car was never plugged in. Now I've been to Europe many, many times. There aren't plugs all over the place. So people just buy a car for the incentive and don't plug it in. And when you do that, it's actually creating worse of an environmental situation than if you just had battery electric vehicle. So what we're seeing, and you're already hearing a lot of noise in the European Union about no longer supporting plug-in hybrids and really going after battery electric vehicles. So as we go from somewhere around 1% adoption to 5% adoption to 30% adoption, I think you're going to find the percentage of fully electric cars is going to be increasing through that time. In terms of content per car, the bogey is somewhere around $300, and that can be plus or minus a little bit. And that depends on how many inverters per car and what the -- how many kilowatts and different things. But if you kind of say it's about $300 a car, it's probably close enough for modeling purposes. And of course, that changes a little bit with time. But that's probably a pretty good bogey on that. And then the last thing that I would tell you is on my last trip to Germany, visiting with a lot of folks, there's a lot more discussion about having longer range and having faster charge times. And that's even for customers, car customers who have a relatively short commute to work. Maybe you only drive 40 miles a day. But if you have a range of, say, 50 or 60 or 80 miles, every time you come home, you have to find a parking spot with a charging station. If you have that 40- to 50-mile commute per day and you have a 400-mile range car, the number of times you have to drive around the circle of the block trying to find a charging station, you charge it on the weekend or you charge it at work. So I think there's a real appreciation for that. Faster charge time is a big important thing. I met with one of the bigger offboard charging infrastructure companies in Europe last time I was there. Charge time is a big deal. And so increasing the voltage and increasing the charge time is an important aspect.

Got it. Next thing, Gregg, you have come out with a model, right, $1.5 billion in sales, 50% to 54% gross margins. Maybe talk to us about what the progress has been towards the model? And I know in the last earnings call, you're still kind of saddled by a lot of the cost structure of the legacy fab. So where are you in terms of improving, right, some of the headwinds that we saw in the last quarter? And then how would you kind of grade the progress towards your longer term model?

Yes. Maybe I'll kick it off, and Neill can give you a little bit more color on it. What I would tell you is, certainly, the Durham factories are subscale. They're antiquated, they're not automated factories, and they give us a cost disadvantage at this point. And so -- and it will basically be that way until we're out of Durham. They will always be a cost disadvantage. As we move to the Mohawk Valley Factory, the most important thing we're going to be doing is we're moving to a modern facility. Just to kind of give you a sense for things, every day in our Durham factories because it's not an automated factory, there is a human contact between, well, a human and a single wafer, 10,000 times a day. In the new factory with automation, that will go to 0. Anybody in semiconductors knows human contact with a wafer is a bad thing, nothing good ever happens with that. So you get yield issues, you get defect issues, you have scratches, you have all sorts of different things. All of that gets eliminated. Automation improves speed at which you're able to change things, improve yields, et cetera. So it's a pretty big transition. But the bottom line is until we get the majority of our production in the Mohawk Valley Factory, we'll have this disadvantage. And maybe, Neill, you can kind of walk through the transition there.

Yes. And I think that's the timing that we should expect, and I think that when you talked about kind of some of the overhang or some legacy cost in Durham is kind of what challenges now. A lot of folks ask, they look at us today and they go you're kind of low to mid-30s. I kind of expect that range and we talk about mid-30s as long as we're primarily shipping out of the North Carolina kind of legacy footprint. But when you look at our plan going out to 24, [Audio Gap] 50% gross margin. So if you look at from a commercial standpoint, that's going really well. We've had $2.5 billion of design-ins over the last 5 quarters, and it has come out of our device pipeline. We're seeing potentially higher rates of electric vehicle adoption as we work through that period. So I think commercially, all of that seems to be on track, and we're working towards those goals, I think, pretty nicely. From a cost standpoint, we talked a lot about the materials execution. So down to shifting off this legacy bad footprint on devices and moving that to the modern facility that Gregg talked about in New York. But I think the timing of that will be next year, in the first half of next year, we'll start qualifying the Mohawk Valley facility and then with customers. And as you get down to 2023, you'll start to see us ramp that with customers more. And so I think you'll start to see more of that trajectory change as we get into our fiscal '23 time frame. As we get into fiscal '24, I think you'll start to see -- we anticipate to see more than 75% of our power device revenue coming out of that Mohawk Valley fab. So there will be a qualification period coming up this year. Then as you get into '23, you'll start to see that transition and more and more revenue coming out. And the cost footprint is just -- it's like a slightly different basis of what we have today. So that's the one leg of the plan that we look to execute. So what are we doing about it? We're building a nice, big modern factory up in New York. And I think once we load that, we'll be in good shape to hit the targets that we've laid out.

Got it. And maybe one last question, Gregg, as let's say, you start to see the benefits of 200-millimeter come through, what does it translate in terms of your TAM expansion in going after the larger devices market? And you have a relationship with Arrow. So talk to us about what does this imply in terms of Cree being kind of a more dominant supplier on the industrial side? Because we only heard of Cree, either in the context of autos or 5G, but talk to us about what having this cost advantage with 200 millimeter does in kind of exploring the larger industrial market for you?

Well, what I would say, it's a cost advantage, it's obviously important across the entire spectrum, including automotive. And you can't be cost disadvantaged and get into the automotive space. And what I would say is based on my experience, if you can -- if the -- automotive is kind of like the New York, New York, if you were. If you can do automotive, you can do anything kind of thing. The real important aspect of what we've got going on in terms of going after the industrial space is the partnership we have with Arrow. And here, internally, we have a huge scale disadvantage in terms of feet on the street selling assets, working -- trying to go after these customers where it's a $1 million to $5 million opportunity across thousands of different customers. We just don't have the bandwidth to do that. I joke a little bit, but when I'm traveling in France with the sales guy going from one account to another, I'm in the car with the entire Cree French sales force. And there's empty seats. So we just don't have the scale. So here's where we partnered with Arrow, one of the, if not the largest fielded sales and applications teams in the entire industry. They've been a great partner to work with, where there are sole and exclusive global distributor for world speed products. We've now done 2 product launches with them where they were the exclusive distributor, launching new products. They've generated billions of dollars worth of opportunity for us. In one of the programs we worked with them, they generated opportunities in something like 45 different countries, half of which we had no employees. So they've been great to work with. And so what we've done is we've augmented our scale disadvantage on the sales side with a partnership with Arrow, and that's going well. I think the Mohawk Valley fab gives us a cost basis to go after it, but we didn't have the channel to market, and now we've solved that.

Excellent. This is great. Gregg, really appreciate your time, Neill, Tyler, for joining us. We could go on for another hour, but we are at the end of our time. Thank you so much for joining us. And for people who are on the call, if you have any follow-up, please feel free to contact me, but we can end our session here. Thank you so much again, Gregg and...

Thank you, everybody. Thanks for having us. Have a nice evening or a nice day. Thank you.

Take care.