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

Welcome back to this session. Really delighted to have the team from Wolfspeed. Joining us, Neill Reynolds, CFO; and Tyler Gronbach, Head of Investor Relations. And I'll start with a few questions of my own. But if you have anything you'd like to bring up, please feel free to raise your hand, and I'll be sure to get you in. But with that welcome Neill. Welcome, Tyler. Really glad to have you here.

And just as a start, maybe, Neill, there's a lot of kind of macro cross currents. So would help to get your perspective on what you're really seeing on the ground from a demand perspective right now?

First of all, thanks a lot for having us here. Thrilled to be here at the conference. Again I started seeing people in-person again, I think first time this year, which has been tremendous. I think if you look at overall kind of from where we sit from a silicon carbide perspective, we're actually somewhat unique, I would think, in terms of the demand profile. Although a lot of people are talking about potential recession or maybe going into more of an uncertain future here, at least over the next months and quarters as we look forward, we're really sitting in the middle of a secular demand story right now. Think about silicon carbide, it's a propensity to drive benefits into electric vehicles and other industrial electrical applications or electronics applications. We're seeing strong secular demand. So we continue to see a lot of strength, not just in the short term, we still have a lot of unfulfilled demand, but I would say even more so in terms of the design ends, the order rate for the future. So as electric vehicle adoption rates continue to increase, that continues to drive down the economics in terms of silicon carbide and drive value and scale. And we're also seeing that in the industrial and energy applications, too. So we're seeing -- we're continuing to see broad-based kind of secular strength regardless of seeing some of these changes in terms of what people are looking at in terms of the economy. So we feel quite strong about that, and I think that really leads us to believe we're still in this situation where silicon carbide continues to be a capacity story. The demand is strong, and it continues to be a capacity story. And Tyler, do you any other comments on that.

I think for the benefit of the group, we're very early stage in this market. And I arrived at the company about 3 years ago and the device opportunity pipeline stood at about $9 billion. Today, it's over $25 billion. So to Neill's what point we're seeing and I think a couple of years ago, we had a very thoughtful plan that kind of outlined growth and the pandemic kind of opened up the aperture a lot. And while auto is the super exciting part that we can talk about today, Industrial & Energy continues to also be an expanding market. And the thing that changed in the last 2 years with the stock out of silicon, what happened was people started to decide, well, if the cost of silicon carbide has come down and it's available to me, why don't I design it in. So thematically, what we've seen in the last year is a lot of substitution. We recently had some customers up to our Mohawk Valley facility that came through our relationship with Arrow and there was a gentleman with a vertical takeoff and landing application. And he basically said I was designing it in with silicon IGBTs, but I couldn't get them. Now I moved to silicon carbide. And the overall cost of his design went down 20% because of the efficiency that he's gaining with the silicon carbide technology. So I think thematically, the story continues to evolve.

Got it. We'll talk about the demand side dynamics. Just one question on the supply situation. I think you did identify that there were disruptions because of China lockdowns, right, 3%, 4% type of impact in June. Where are we? Are you starting to see any changes in that?

Well, we put some piques into our guidance, some of the impacts from what we saw from the China shutdowns. Right now, we have packaged devices that kind of go through the channel. So a lot of our revenue right now isn't so much auto-based, although, we're winning a lot of automotive deals that will gestate overtime. And we'll see that start to blend into the revenue. A lot of that is on packaged parts. We had several packagers in China, which were impacted from an attendance standpoint, in terms of running product, just getting people in and running product. That then translated into a situation where logistics started to become a challenge, holding some of the material at the airports and shipping locations and things like that. As we came into the guidance of the quarter, we kind of baked in. We called it mid-single-digit million dollars of impact, as you kind of reflect on that put back. And since then, I think if you look out there and you read the reports, I think that region is starting to clear up, and we started to see things and improved to a certain extent. So I think we're -- I think things haven't really played out much differently than we anticipated when we moved into the quarter.

And I would add to that, what we've done, too, from a remediation standpoint, Joe Roybal, who joined the company back in December. Back office expert really has come in now and looked at the footprint and tried to figure out what's the better way to kind of remediate this going forward. So the good news is, I think we're putting a solution in place that we kind of solve for that going forward.

Got it. Now let's go to the more interesting part, right, which is the really structural dynamics and advantages, right, of this new technology. Let's make it right easy for the audience. How much extra cost does silicon carbide add and how much extra benefit does it provide?

There's a lot of different ways to look at this. But if you look at various studies, if you look at the cost of a substrate, so generally, a silicon carbide substrate will cost us, as I mentioned, a bit more than a silicon substrate. But the net benefit, as you put it into, like say, a high-power electrical or electric application, are pretty significant. So switching speeds and RDS(on) and other metrics around in terms of the capability of the device are a lot stronger with silicon carbide than silicon. So when people look at that, they see benefits -- system cost benefits, but also -- and people think if you take the system cost benefit, including weight, size of battery or potentially use it for range and better efficiency, I think it's between 5 and 7x better for every dollar you invest in silicon carbide, you get that type of return. So -- and I think that's pretty much what people have figured out, both in EV space and some of the industrial applications is, when they make that transition, you see some of the benefit and the return of putting in silicon carbide into an application, generally pays off. And that's really what we're seeing that adoption rate change. So both if you look at electric vehicle adoption rates, you look at silicon carbide within those vehicles, also in these industrial energy applications are seeing, as Tyler pointed out earlier, you're seeing a lot of transition as well. So yes, I'd say higher cost substrate, but the returns are significant, and we're seeing that from -- not just from our standpoint in terms of our design ends, but from an overall market growth, we're seeing a substantial pickup in overall demand because I think a lot of those benefits people are understanding, and they're baking that into their applications.

Got it. Silicon carbide very early in its life cycle, right? So naturally, very -- I think it's probably the most capital-intensive part of the semiconductor industry, if I'm not mistaken. How are you ensuring that the industry doesn't overinvest, right? If we were just talking about the capacity side, how much capacity do you think is going to come online from Wolfspeed and then from an industry perspective over the next few years? And if you could also then help us make that bridge to how does that compare to the demand that might come at the same time?

Let me start off and maybe, Tyler, you want to kind of jump in here a little bit. So I think overall, if you look at the market, if you go back just a year ago, and we had an Investor Day last November. And we look at the market size and the growth rate potentials. And I think even since then, we've seen a significant step-up in capacity. So I think when people start thinking about how much capacity is required, who are the players in the industry, from our perspective, it's not so much that there's going to be oversupply. Our view is that the bigger risk is that it actually potentially undersupply in the industry and for several reasons not so much from a fab perspective, but the amount of silicon carbide substrates that are required to bring on into the industry to fill those fabs and to fill those applications. So as you look over time, I think you're going to see a supply and demand disconnect, continuing with silicon carbide over the next several years, even after the end of the decade.

Yes. It's interesting, last week, the CEO of Yole, an industry outfit that kind of tracks the silicon carbide market, made the comment that it takes 10 years of an R&D cycle to gain scale in silicon carbide. So I think the next question, which is a really good one is, how quickly can capacity come on. And it's not something that you can take money and manpower, and all of a sudden produce a silicon carbide bull. So I think, Vivek, it's going to take time, and I think capacity will certainly come into the market. But I think one of the reasons why on our most recent investor call, we started talking and flagging about more capacity needed is because we're in a situation where our Mohawk Valley facility is essentially sold out. and we're just bringing that online. So I think it's fair to say that, as we look out into the next 3 to 5 years, it's going to be a very capacity-constrained environment.

Is there a way to have a simple metric that, let's say, over the next 3 to 5 years, EV adoption is X, but the capacity coming online is lower than X, right? Is there some -- I know it's hard to simplify everything down to those kind of metrics. But ultimately, that's what we really care about, right?

Yes. I'm not sure if it's that simple. And the reason for that, it really depends on the applications. And I think that it's very early days in terms of how the supply chain is going to sort itself out. So we make both modules, we do bare die and we obviously sell wafers and substrates. So we go into all the elements. We sell them to all elements of the supply chain as it relates to silicon carbide, and we've been doing all of that for a long time. So it really depends on how people are viewing that in terms of the technology transitions and the supply chain due to some customers want modules, some want to do their own module, want bare die. We've made a lot -- we maintained a lot of flexibility in terms of how we deal with these different customers over time. I think clearly, bare die will be a bigger portion of our business over time. That's something we're certainly focused on. But we do make ourselves available in terms of opportunities to fill various pieces of the supply chain.

Yes. And I'd add, Vivek. We were -- recently, we were talking with a global OEM. And he said, the 30% penetration rate is going to be an interesting demarcation point because what he was telling us is, they've done the math that basically says at 30% EV penetration what happens is the cost of an internal combustion engine car starts to go up because of the fixed cost of pistons, drivetrains, axles, all the things that are needed for an internal combustion engine car starts to go up. And the price of an electric vehicle should come down. And what will then probably happen is penetration will probably rapidly increase. So to Neill's point, it's hard to kind of give you that metric, but we're paying very close attention about penetration rates because it could lead to an even faster ramp than we anticipate.

And the contention that capacity will lag demand, are you also including all the capacity that ON or [ ROHM ] or Infineon or ST or others are planning to bring up?

For sure. And I think from that perspective, because of the size of the market opportunity, what we're seeing is there's other material suppliers that are bringing on capacity. In addition, as you point out, there are device manufacturers are also working on their own internalization efforts. And I think everyone is working on that to some certain degree. And then the way we think about that is, look, the overall industry, everyone needs to bring supply to the table. So hope we work with those customers on the material side, they're great customers. They tell us exactly what they need from us. We work with them on that, and then we kind of let them go and do their piece as well, and we're hopeful that they're successful. And the reason for that is we need to bring on enough capacity for the entire industry. So all of us need to bring something to the table. It's still very, as you mentioned earlier, of the back -- it's still kind of very early days, I would say, in this industry in the sense that there's going to be a lot of capacity required. And I would say even the fabs that we're building today the Mohawk Valley fab, we're contemplating a second fab given the design in levels and the capacity requirements. There clearly wouldn't be enough to supply the majority of the industry. So I think all of the participants right now really need to bring that to the table, and we just work with them closely in terms of what do they think they can do from an internalization basis, work with other substrate suppliers, which is fine, but all of us really need to be bringing significant capacity to make all of this work over time.

Got it. Are you concerned, Neill, that in the automotive industry, especially in EVs that maybe a number of your direct or indirect customers are not getting access to enough materials, say, batteries or others? So it has nothing to do with what you're providing, but are there other constraints in the automotive ecosystem that could ultimately affect the demand there?

It's possible. I mean we've seen the supply chain challenges, and maybe Tyler can follow up with some other additional comments. But we've seen the supply chain challenges, limit some people, and there's been some fits and starts. Although I think if you look at the overall electric vehicle, I think most people would point to over time, the battery and lithium potentially being the biggest issue there. From our perspective, though, on the battery front, silicon carbide makes such a big difference in terms of the efficiency of the battery. So think about all of these global OEMs investing all of this money into battery efficiency. And what silicon carbide then does is, it protects the battery. It makes it between 5% and 15% more efficient, and you can make your battery 5% to 15% smaller or you can add 5% to 15% more range. So we had a global CEO kind of talk to us about, in the event there was some sort of shorter, let's say, on the battery front, they can keep their lines going by just making smaller batteries. So that actually ends up being kind of a good thing for silicon carbide. And I think most of the customers in this area have really figured out that in order to kind of protect the battery, which I think, to your question, would be maybe the biggest risk. You really need to secure the silicon carbide capacity to ensure that you can keep your lines running and really have positive outcomes regardless of with some of these variables that could potentially happen.

Yes. I mean what we now understand from a design perspective, there's 2 considerations that the global OEMs have and that is, charge time and range, as Neill pointed out. And so even if there might be a shortage that could be potentially impacting supply, I think kind of the trains left the station on, should it be silicon carbide or should be silicon in the inverter because of what I just talked about, which is people. Consumer preference really is range and charge time. So I think those 2 drivers is really -- was what has kind of led -- that led us to a situation to where silicon carbide will definitely be in the consideration set.

Do you see a situation that just given the cost, the incrementally higher cost of silicon carbide that it's only used in cars where you need more than, I don't know, 200 miles. I guess, is there some crossover point versus silicon? Because I imagine the cost of silicon IGBT is also going down every year, right?

Well, what's interesting on that point is that during the supply crisis, we've seen silicon pricing actually go up. So we've seen it start to -- we've seen silicon and silicon carbide actually get a lot tighter in terms of the cost differential over the last couple of years as we've seen silicon pricing go up. So I think that as you look at the range benefits or which applications or which vehicles is it, a lot of discussion on, is it a 400-volt EV? Or is it an 800-volt EV? But what we've seen now is for new production, most OEMs are moving to higher power electric vehicles. So that shift is happening, and that's good for silicon carbide. But I think what that's resulted in, you think about the -- we talked about the battery protection, but you also think about the range benefits. When we talked to a -- the head of e-mobility in a major Tier 1 recently, and Gregg talked about this recently. And he said, look, congratulations, great pipeline. I love everything you're doing. And he was amazed how the pipeline has transitioned so much in the last 2 years. That pretty much they don't see any silicon anymore in a lot of the deals that we're seeing. And I think that's true, as we get down to the finish line on a lot of the design-ins that we've been working on, it's not really anymore do we see silicon carbide versus silicon. It's really silicon carbide providers down at the finish line running these deals now. And to Tyler's point earlier, as you start to see the volume of these deals pick up, as you start to see the transition of electric vehicles, over time that will really turn into high volumes for silicon carbide, which then brings us back to the capacity question again. As we start to see these ramps, as we start to see secular trends in terms of the cost of an ICE vehicle, as you get out to like this kind of '27, '28 time frame, there's -- it's more likely that you start to see this even pick up higher from a growth rate, which then brings us again back to the capacity discussion.

Yes. And I think, Vivek, the way to think about the yin and the yang of it is that, it's going to boil down to how much charging infrastructure can make it to the market in the next 5 years. And I think that's what's driving, I think, the automotive OEMs to say, especially in a place like Europe for us here in the United States we'll drive our car home, plug it in the night. But in Europe, you're going to be parking on a street, you're going to need access to charging. So if your vehicle only goes 150 kilometers and you're driving to and from work every day, you're going to need some place to charge it or you're going to want to -- you're going to buy a car with longer range.

It's a pain even in the U.S. not just -- personal experience.

Exactly.

So the next thing I wanted to kind of dive into is, where is the sustainable competitive differentiation in this market? Because we had one of your peers yesterday, who said that, look, the differentiation is not substrate by itself or device by itself or you need all 3. And according to them, they were the most differentiated because they can do that Trinity, right, the best. How -- where do you think is your biggest competitive differentiator? And do you agree with that statement that you need all 3 to be successful?

I think you do need all 3 to be successful. In fact, we've been, I think, doing all 3 for roughly 10 years. I think we're one of the first 2 actually to have the fully integrated capability. But I do think that if you think clearly, having integration and working these things together at early stages of technology, having substrate capability, having device capability and being able to do all -- and having module capability, which we've had for a long time and being able to do these 3 things and seeing how these all work together is important. Now from what's harder to do, a lot of people asking, what was the value equation here? Well, clearly, crystal growth, do you think about, very beginning of the process, that's the hardest part, growing solid crystals with good quality and capability is the hardest part now. Silicon carbide being one of the hardest materials in the world. It then becomes difficult to work with downstream, too. So you've got CMP and grinding, epitaxy, fabbing and then packaging, all of these things are a little bit more difficult than silicon carbide. And again, we've been doing this for a long time. So when you think about the integration between those 2 and having those engineering teams and scientists working together, in understanding how these things all play off each other is very important. And not just because you want to have the technical capability, but OEMs right now are very focused on the performance of the inverter and the performance of the module. Think about it, they're going to change out their entire profile. I mean most of the companies we talk to now from an R&D perspective, they've kind of cut ties with internal combustion engines. Anything new coming out in the future is pretty much going to be electric vehicles. They'll have nothing else to sell. So they're looking for that differentiation, that's replaced, but before was the engine, which will now be the inverter. And the module and the inverter is probably the biggest piece of that and then silicon carbide as an efficiency measure within the module then becomes important, too. So there's a lot of efficiency to be ring out of this process, I think, over the next multiple years and having that integrated capability to understand where those benefits are coming from, from an efficiency standpoint, whether that be all the way back from the silicon carbide material itself all the way through the end of the supply chain is incredibly important.

Just remind the group, we started making silicon carbide to support our LED and lighting business. So if you look at the last 20 years of silicon carbide production, we've probably made close to 95% of the world's silicon carbide. And then when Gregg Lowe arrived at the company back in 2017, he made a conscious effort to sign long-term wafer supply agreements with our rivals, with those folks to put more silicon carbide into the market. Why does that all matter? Because you can only get better at silicon carbide, the more cycles you actually go through. So if you start to think about the 20-plus years that we've been making silicon carbide to support our LED business, to support our wafer supply customers and now we're moving from 150-millimeter wafers to 200-millimeter wafers. The only way you get better is by going through that. So I'd say the moat that we really have is, we have great expertise in devices because we've been doing it for a long time, but don't overlook the material side of the equation and just being able to make something that can actually yield.

Got it. I wanted to touch on Mohawk Valley. You recently had the formal ribbon cutting ceremony. Give us a sense for what progress you have seen there so far? When do you think that facility is going to be up and running and producing revenue that is tangible that you can state, I don't know, and you recall then say, okay, this is coming from Mohawk Valley?

Yes. I mean first of all, let me say, I think the team has made just tremendous progress. If you go back just a little over 2 years ago in March of 2020, right when the pandemic started, our Mohawk Valley fab that we're currently running wafers in was a field of mud. And now we've transitioned that into a factory that's been -- that's actually kind of running and qualifying wafers. So that's where we're at right now. So we've built the fab. We've hooked up the gases and chemicals and the water systems and everything else you do in a fab. And right now, we're running qualification lots, and we'll continue to manage that, I think through the second half of this calendar year. And as we -- and then we'll start to, in that same time frame to our qualifying customers, could see a little bit of revenue out of Mohawk Valley in the calendar year, but we anticipate seeing more of substantial revenue between kind of in the March and June quarter next year, which is consistent with the plans that we've talked about previously. So I think more or less, we're on track as to what we kind of thought we would have the capability of it. But I think the bigger thing is, now that we've got the cost of the fab kind of understood. We have people hired. Like I said, we have all the systems hooked up. I think the roadmap now, the blueprint for 200-millimeter wafers in combination with our highly automated 200-millimeter fab is coming together to give us a really good clean line of sight to future profitability looks like. Now as it relates to when are we going to get a good look at that, I mean we're starting to see it now. We're getting -- I'd say, so as I sit here today, I think we have better visibility to that than we've ever had. And I think as we get to the end of the year, finish our qualification lots and start to see some starts to pick up towards the end of this year and into early next year, we'll start having good visibility to that. And we'll be able to provide more concrete evidence around the kind of the unit cost economics of what that fab will provide. But I would say, if you take the combination of what we're seeing today and then extrapolate that out until the '24 or '26 plus time frame, is really tremendous, and we're really enthused with what we're seeing right now.

Any growing pains in that? Because I know you use a lot of specialized equipment, a lot of it you designed yourself. Also, but any challenges in accessing the tools that you need, anything else from the logistics?

No. I think we've been ahead of this in the sense of ordering tools and managing lead times. And in fact, most of the tools from Mohawk Valley we ordered before the pandemic, before a lot of the supply crisis start. So having that first set of tools is good. The second thing we've done technically speaking, is, we've been growing 200-millimeter wafers for quite some time. In fact, over the last year plus, we had a pilot line up in Albany, and we're running wafers on our -- and developing our process on that pilot line. Now what we've done since then is we've taken that pilot line down and we've moved into the fab. So we have a pretty good idea of how the 200-millimeter process works and how it yields. In fact, the yields on our 200-millimeter wafers right now are better than what we have our 150-millimeter wafers, which is a really good sign and something you don't normally see. So I feel technically speaking, the entire process, from growing the crystals to slicing to epitaxy to fab, I think we're technically in very good shape. Now is there a possibility of running this, growing pains and growing this fabs? Of course, we could see something pop up. I would see that as transitory, so to speak. You probably have some growing pains and issues as you start to bring things up that would just be natural. But from a technical standpoint, in terms of our confidence, what we're seeing from the 200-millimeter wafer, the process technology and the fab, we fund that through the back end, that looks really good.

What happens to the Durham fab? And when your new plant is up and running, right, the new and shiny plant is up and running, so what happens to the order facility?

Right now we have, as I stated earlier, despite some of the kind of market outlook that people are talking about and maybe Tyler you can talk to that part a little bit.

Sure.

We've got still over $100 million of unfulfilled device demand right now, and we continue to see that as we move out into fiscal 2023 and into calendar 2023. So we're still seeing significant demand for our products over that time frame. So I think the Durham fab also went into 150-millimeter. It's playing an important role right now in supporting our customers. As we bring up our new fab, we initially anticipated this being a bridge. But right now, I think there's kind of -- indefinitely, we've just got to continue to support our customers until which time we see some of the other fabs come up as we start thinking about things into the future. So right now, I think, indefinitely we'll keep making progress at Durham. We've had a new fab leader there, Missy Stigall, who we hired under a year ago. She's making tremendous progress. So -- and I think with the progress she's making, it could play a role in the supply chain. But I think one of the things there to think about too, though is that, just from a capacity standpoint, I think it's something like 15% or 20% utilization in Mohawk Valley is already bigger than our entire device footprint right now. So we're talking about a completely different scale as you start to move forward. We are continuing to see very strong demand.

Yes, I'd say this. I think the choice, Vivek, that we're going to have to make, and we're going to update everybody on the long-term outlook later this fall, probably, do kind of an Investor Day. But I think the choice for us on Durham is, do we keep running it and capture opportunity, more opportunity that's out there where it's probably at a lower gross margin than we can probably run through our Mohawk Valley. But I think if we can give everybody kind of the underlying economics of how Mohawk Valley is running, keeping Durham running a little bit longer may satisfy what I would call non-automotive customers. So it's just a conscious choice that we're going to have to make.

Got it. You still see requirements for adding additional CapEx at Durham or anything on the upgrade side? Or is the plan to just do the best with what you have there?

I think it's -- I think we've obviously made some changes there. There's not a big investment strategy with the Durham fab from that standpoint. I think we can bring out some more efficiencies. I think this is making good progress. I think there's more to be add in terms of what, her and the team are doing. And I think they've done a terrific job with it. We've really seen them kind of ride the ship, so to speak. So we feel really good about what that team is doing. And then after that, it's really about driving revenue through the new Mohawk Valley fab. So I think that investment you'll see is making are really on 200-millimeter automation and building out a highly automated fab with Mohawk Valley and really proving out this blueprint. And it's kind of come -- I won't say it's come out and happen sands, but we've kind of -- we've been trialing and building the supply chain over the last 2 or 3 years. But in my mind, I think the clarity now around the 200-millimeter platform wafer going into the highly automated fab and running that blueprint, and the early returns we're seeing on that, to us, that's the key. And I think if you see us expand capacity beyond that, it would really be taking that same blueprint and then just doing it again, maybe on a different scale. So that's really what we're working on now.

Yes. And for the group's benefit, just as a reminder, we've announced close to $9 billion of design-ins in the last 3 years, and you should think about 60% of that is for automotive and 40% is for Industrial & Energy and RF. But I -- the reason why I point that out is, this is why capacity is going to be so critically important because there are still large OEMs to make decisions on models. And so they're not going to award it to somebody, they can't look them in the eye and say, I've got the capacity to support your launch. I was with a global OEM, and that was -- they literally put a chart up and they said, okay, here's where samples get delivered. Here's where we're ready to ramp. How are you doing? And so I think it's not lost on us that at the end of the day, more capacity is going to be needed by us for the industry, but it's because key decisions are yet to be made in certain markets.

Got it. Now to our favorite topic of gross margins which is, walk us through the journey from where you are kind of in the mid high 30s, right, towards the 50% goal? How important is it to smoothly bring up Mohawk Valley in that? Or should we be prepared for some turbulence along the way?

Well, first of all, I think the roadmap is pretty straightforward. I talked about getting, moving from 150-millimeter wafer, to getting most of our revenue on a 200-millimeter wafer, where typically you get 40% cost improvement on the die level. Now of course, our yields are better in Mohawk Valley. So we would see cost improvement at the die level even beyond that, as we look forward. So we feel really good about the roadmap. So really, the roadmap is pretty clear for us. We've got to get volume through Mohawk Valley. Like I said, 15% to 20% of Mohawk Valley utilization is already bigger than our current device footprint. So we can -- we'll have the ability to kind of bring up the revenue. And when we start seeing that, we're going to build that now out as fast as possible. We're going to put CapEx to work to push the limits and really fill out Mohawk Valley as fast as possible to match the increased supply that we've had. So we're just going to continue to bring on that capacity. And when we do, given the unit cost economics I just mentioned, we'll start to transition and see the margins really start to flip over. So it's really a function of driving more revenue through our kind of new -- our new fab and when we do that, we just see a cost footprint that would bring the economics into play from a unit cost economics but also from our gross margin perspective. As it relates to bringing up the fab, we've been very careful with this. I think there's reasonable metrics out there about how quickly -- how many wafer starts per week or per month you want to bring in it to a fab and how quickly you bring on a new line. I think we've been cautious. I think we've managed that and we think we're managing that effectively, and we've got really good plans, I think, around bringing that online. And as we get success, we'll kind of add volume over time. And I would say that's kind of like I said, all on track. So no changes there. Now bringing on a new diameter wafer and building a new fab, I guess, as I said earlier, is there going to be a few bumps in the road? Probably. But I don't think there are things that won't be -- things that we won't be able to solve. I think technically, we're in really good shape.

Got it. I had one other question about the technology side. So there are a number of companies investing in the GaN market, right, who believe that they can also be right competitors in the auto. Is it as simple as saying, well, the main inverter right is always silicon carbide -- is going to be silicon carbide, but you could see a lot more competition from GaN and other parts, right, whether it's onward charger or other DC-DC converters and so forth?

That's possible. I think that -- I think if you look at the content in an electric vehicle that's power semiconductors, 85% or so is in the inverter. So it's in the module and the inverter. So that's really where we spend the vast majority of our time because if you look at the EV or the potential market there, the vast majority of it is within inverter. And I think silicon carbide is pretty much the part of choice now. As I said earlier, when we talk to OEMs, that's really the only thing that's at the finish line in terms of bidding. So it's really silicon carbide in these deals and looking to win the inverter. As you move into the smaller parts of the vehicle in terms of the market share within the vehicle, onboard charging can play a role potentially, I think, with GaN, DC to DC conversion. And from a technology standpoint, we do see it there from time to time. But we also win our fair share, I would say, within the onboard charging and DC to DC conversion. It really just depends on the OEM, the deal, the magnitude of charging, that type of thing.

Yes. The interesting thing, as Neill points out, it's about $300 to $600 worth of content of silicon carbide content. But in the case or in the case of charging, it's a very small percent. The majority is in the inverter. I think the other thing that's important to note that the charging infrastructure, there's $1,000 of silicon carbide content in a supercharger. So it's -- there's great opportunity in the inverter, and we're also seeing it in charging infrastructure, too.

Got it. How are you handling the dynamic whereby a lot of your customers, right, on the substrate side, very large important customers are also competitors on the device side because as Mohawk Valley starts to ramp, right, and you get this really nice cost structure, how are you going to do that trade-off between, should I make a device out of this and get higher ASPs? Or should I give it to a substrate customer?

Well, I think, first of all, we're in very early days. Now we have a 200-millimeter wafer. We've got 150-millimeter wafer. We've done several extend and expand deals with a lot of the other semiconductor companies who do kind of power devices. And I think we worked with them very well. And I think they're good customers. We take care of them in terms of we don't take like all the bad ones and give it to them and keep the good ones for us, so to speak. We take good care of our customers in that sense. And I think as I said at the very beginning, this is turning into a reliable supply gain. This is what silicon carbide is about. It's not so much about who has the demand but who has reliable supply. And I don't think us alone can go do that. So I think we've got to figure out how to help support them for what they need. And if they come to us and say they need a certain percent of their external supply coming from us, we'll work with them on that. And that's what we've done so far. So I think we'll continue to manage this. It's early days. Lot of people ask me about 200 millimeter, are you going sell that to people. I think we've got to get our own ramped up, and we're still going through our own learning cycles to get the very initial wafers here. So I think that will be several years before that there's a decision point there. So I think we're still all relatively early on in how to manage this. And I think all of us, everyone in this area -- both us and our device rivals are all going to have to bring not just device capacity, but to some extent, materials capacity as well into the marketplace just if that fill was required from a demand standpoint.

I see. Do you see China becoming a competitor in this market? There are a lot of parallels drawn between this market and what happened in the solar market and the enormous price deflation. What are the similarities and differences with that?

I think if you look at just anywhere, I think a lot of people are trying to get into silicon carbide materials development and for good reason, I think, because it's going to require an immense amount of capacity as time goes on. If you look at China or other places and pretty much anywhere, the challenge is still going to be that, you can't just throw money at this. You can't just try and outspend this or throw people at it. The number of people in the world that actually have really good solid knowledge about how to grow silicon carbide crystals and manage it is actually very small. And the second thing is, silicon carbide requires a lot of learning cycles. So if you think about, as Tyler pointed out, we've got a lot of long-term deals with various companies on the material side. And what we've done with those long-term deals is we've taken that internally, we've been able to build scale and then drive additional learning cycles around that capacity. That's not something you can just do without the know-how. So we already kind of know how to build our own reactors and grow the crystals, learn from that, make changes, feed that back into the kind of the circular kind of learning process, but it's difficult to kind of get into. You got to know how to build -- you got to be able to know how our crystal looks to build a furnace. And from that perspective, it's not easy to kind of get those learning cycles going only because of the material is so hard to work with. It runs at over 2,500 degrees C. So you can't kind of lift up the lid and look inside as you're kind of going on. So as Tyler pointed out earlier, as Yole, CEO pointed out, one of the breakouts recently, it takes about 10 years to kind of figure this out. So I think that while I think people, whether it be China or otherwise, are going to look to try and get into this, I think it's going to be a lot harder than they think to do this, which is fine. In the meantime, we'll continue to work with people on the material side to ensure that we can support some of the industry growth and really support and ensure that this is a fully commercialized, capable technology for the long term.

Terrific. Thank you so much, Neill. Thank you, Tyler.

Thank you. Enjoyed this, here.

Thank you for the conversation.

Thank you for having us.

Thanks so much. Thank you.

Okay. Thank you, all.