ON Semiconductor Corporation (ON) Earnings Call Transcript & Summary

Next up is one of our top picks in the whole semiconductor space on semi. We've got the dream team here, Hassane El-Khoury, the CEO; and Thad Trent, CFO extraordinaire. Why is it one of our top picks, an absolutely amazing turnaround story architected by these guys to my left, gross margins 10 points above the previous peak operating margins, more than 15 points above the previous peak. And then you also have a very nice growth aspect with the silicon carbide and just the general automotive market. So guys, thanks for coming. And why don't we start there? Before we dig into the 70 million questions on silicon carbide that I'm sure myself and the audience will ask, and by the way, this is intended to be interactive. So if anybody has any questions, feel free to ask them.

But let's start with automotive. Automotive has been strong for you guys all year. How are things going? And we've had a number of your competitors say everything from the previous company said automotive is great. It's going to be up new issues, and you've had other companies say automotive is correcting a little bit. So where does ON stand on the automotive spectrum, and we'll jump in from there.

Yes. From an Auto bore perspective, obviously, we still see that as a positive end market for us, for a few things. The biggest growth driver for us is electrification. And from an EV perspective, you can see a little bit on the global from an EV, the numbers of units are going up, and that's going to drive the growth for us. But even if I take outside of the silicon carbide growth that we're seeing, automotive has grown. Now obviously, it's not at the same rate as we've seen in the last few years, but there was still growth. And that's the content that we keep referring to where apples-to-apples, we have more content in newer cars than we had in previous cars. And it's not a lot of people that look at it from the glass half empty of, well, we've been building premium vehicles, so content is higher. Well, name one entry vehicle that doesn't have more power than it did last year, more screens. Every car today has power windows, power seats, power steering, the key term is power, which is what we do. So all of that is net incremental content for us. So any new car being built is going to have more content and every new EV being built is going to have almost 14x more content just on the drivetrain. Because if you think about the content for -- on a drivetrain in an internal combustion engine, onsemi has about $50 worth of content. That same car turns into an EV that's $750. So you have $700 incremental dollar per vehicle that converts from ICE to EV, that's where the underlying strength in our business is, and we see that continuing.

Hassane, maybe you could take us through what your sort of current average and peak content is in like an ICE versus a hybrid versus an EV car? And give us a sense of -- you talked about your content has gone up even in an EV car over the last few years, like how much has your content gone up over the last few years? And then maybe give us a sense of your content per each 1 of those 3 models.

Yes. Look, I mean, the plug-in hybrid there it's very subjective. So I'll focus on the 2 extremes because that's really where the focus is. And that content has -- look, in 2022, what we talked about in our last -- not this Analyst Day, but the prior one when we deployed our strategy, we had a vehicle that went into production with $841 worth of content. That's not just from a power perspective because tangential to the electrification, there's something we can't really forget about because onsemi is a very big player in automotive also is the ADAS side of it. Every vehicle now has cameras in it for some level of autonomous driving, whether you believe in Level 4, Level 5 or when we will achieve full autonomy. Let's talk about just Level 2+, which is from surround cam, the surround view, which at least requires 4 cameras to adaptive cruise control to lane departure and all that is driving content in automotive orthogonal to electrification because you have that same trend happening in internal combustion. So when I talk about content for onsemi in automotive, I'm not only referring to just the power semis, which is, of course, a big growth, that's a $700. We got almost $1,000 on all the other stuff. So net-net, that's why we always talk about we're not pegged to SAR or units sold, it's more on the content that is going into these units.

And then how do you -- looking into 2024, how do you guys feel about the automotive market there?

Feel pretty good. Obviously, you have the continued execution, continued growth in silicon carbide. But even outside of silicon carbide, we're going to see some of that more content. We -- some people may not know, we're #1 in LED drivers for automotive. Anecdotally, LED is the new chrome, right? Every car identity now has more LEDs than in the '60s and '70s was chrome. That's more content, ultrasonic sensors, image sensors, power. All that content is driven by megatrends that OEMs are using to differentiate their product from every other product. And that's where they put the value, and that's where we pull into that demand. So we're very comfortable and very excited about the future of automotive. Including in '24 and beyond and led by, of course, the continued ramp in silicon carbide.

With this increasing content and increasing growth in the automotive market, has the margin profile for your automotive business changed over the years? Has it gotten better? Or do you sort of price to some sort of margin and any implications going forward?

No. So we don't price from a cost-plus to hit a certain margin. What we -- maybe that historically was how the company did. But you said we've been in a very...

Different mode.

Very different mode, right now, is value. So as long as we create technology and we create products that add value to the customer, we're going to extract that value. And that's across all the products. From a margin profile, of course, if you look at the last 2 years, our margin profile, like you said, went up almost 20 points.

We like that, by the way.

That's right. We do too because it allows us to invest in more competitive technologies and reinvest in our business, that at the same time, our auto and industrial has become about 80% of our revenue. So we've driven a fundamental mix shift in the business, on the end market and the products and the value that we create for our customers in these markets and the value that we extract with our product. So that's been all the multipronged transformation of our strategy that we're not done with it yet. We put out a target of financial target over the next 5 years at 53% gross margin. We feel comfortable with it given everything that we have executed to, but yet we have to still execute to. And that's a lot of it is new products, and the scale that we will be able to achieve. So that's all been positive. So we're not stopping at where we've delivered. We actually upped it, and we've always talked about our financial target as a milestone, not a destination. So that's how we run our business.

Just to give you a funny anecdote, I saw a friend of mine yesterday who retired 5 years ago from Wall Street, we're talking about onsemi, he goes, "Yes, so is ON still trading between like $8 and $15. Where is it in that $8 to $15 range? That's still higher buddy."

Yes. Very different companies.

Out of zero. Before we move on out of auto, could you just touch on the sensor business a little bit. To me, this is a bit of a like an undersung hero. You've made some key acquisitions there. It's a pretty nicely growing business. Just give a sense of like size, growth rates on the sensor business?

Yes. From the image sensing business in general, a lot of us focus on the turnaround and the new financial performance of the company. But within the financial performance of the company is a very successful turnaround of our image sensing business, where margins are around 50, they reach 50 from a kind of the same level as where the corporation was new products. So we're leveraging our new products into the new markets. So what does that mean? The last 5 years, the number of cameras per car has doubled. We expect that number of cameras to double again in the next 5 years. So there's unit growth just from a penetration and more cameras per car going forward, but also more cars getting that Level 2+ functionality in it. So that's one. The other trend that's helping this business is the megapixel or the resolution for these cameras. When we talk about now the future is 8 megapixels versus 1, 2 and 3 megapixels in the last few years. 8 megapixel camera is about 2.5x the ASP. So again, there's an ASP uplift, more unit, that's what's going to accelerate the growth. So net-net, I go back to the prior comment about automotive. When we talk about content driving growth just 1 camera going from 1 resolution to the next give you more than 2x the lift. That's growth even with the same number of vehicles just by changing the resolution. So these are things that we are indexed to more than the number of vehicles that sell.

Great. Let's touch on the Industrial business. It's another fairly large business for you guys. Just talk about the trends there and what you see over the next, I guess, year prospects for 2024?

Yes. So the same as automotive. We see, I would say, overall, stability. We've seen some softness exiting last year. I think that's behind us. We see stability in the industrial market. That's a general industrial market. But within that, what's important for us is what end markets within the industrial we anchor on, that's renewable energy, energy storage, energy conversion, and energy distribution. So think about it as solar, wind for generation, energy storage systems, and chargers for the deployment. All 3 of these have seen the exact same growth because they're driven by what I call the sustainable ecosystem by the e-mobility. So when EVs ramp, you need the infrastructure and needed charging. The charging is going to ramp when they see the EV. So that combination is uplifting both markets. That business for us in the renewable side has grown in the last 2 years, about 70% year-on-year. That's tremendous growth. And it's not small business. It's a meaningful part of our company, and that's going to be also growth. So that's going to overshadow any flatness or call it, short-term market fluctuation in industrial. But over the long term, it's going to drive industrial above GDP plus that we look at.

Yes. One question I get is the downturn is hitting everybody else, why isn't hitting ON? And if you look at your non-silicon carbide, I guess, your core semi business, you have seen some fluctuation and perturbation in there. Maybe touch on that and the preemptive measures you guys have taken as far as keeping inventory low and utilization rates.

And just in terms of the markets and what we see, I mean, Hassane did a great job of covering the EVs, alternative energy, which are growing nicely. What we saw really Q2 of last year was those other soft markets, we saw consumer and compute a small piece of our business, but we saw that getting soft. We saw the consumer side of industrial getting soft as well. I think we saw that decline for a couple of quarters, and it's basically gone sideways since then. Now what we've done, starting back in Q2 of last year, as we started taking utilization down pretty hard. We've taken our utilization down from 83% down to 70% this last quarter. At the same time, our margins are holding, right? So it shows the structural changes inside the company. We've managed our disty inventory extremely tight. We've got it at 7 weeks, we've been running in that range for really 2 years. We'll continue to run there. Historically, the company ran 13, 15 weeks, right? I mean so we proactively manage that. And the distributors would take much more inventory if we'd ship it to them. We're actually not shipping it to them. So the LTSA coverage that we have is actually protecting us and we're building the LTSAs. And that's allowed us to take that utilization down and kind of start with the other markets. And as I think we go forward, we're assuming other than the secular drivers, we're assuming this market kind of go sideways for a period of time. If we're wrong, there's a nice tailwind behind us as we take utilization back up and start cranking again.

Yes. Two questions on that Thad. Number one, can you talk about internal inventory add-on where it is right now? And what's the ideal range is?

Yes. So we've been building inventory for 2 reasons: one silicon carbide with the ramp of silicon carbide. I know you're going to hit on that. The other piece is we divested 4 fabs and we've been going through fab transitions. If you exclude that and you look at the base inventory, excluding that inventory, our inventory was actually down 6 days last quarter. So in total, I think we're about 163 days. Clearly, we'll be below that at steady state once we get through these transitions and the SIC ramp. But that's another indication of just how we've been managing. We've been trying to take inventory down and keep it lean as well.

And when do you think you could be at the ideal inventory level barring anything flipping out up or down in the market? Is it 2, 3 quarters out or .

No, I think we're further than that just because of the silicon carbide brand, right? We've still got a ramp for next year for '25 as well. So it will take a while to burn through that. And the fab transitions take 3 years anyway, right? So those are strategic builds for long-term support for customers.

Great. One thing just on the LTSA, Thad mentioned, we're not building to backlog. We haven't built to backlog for the last 2 years or so. We've always said we're under-shipping demand. And fast forward to now, it puts us in a much better position because we haven't really been shipping to demand, whether you believed it or not. We've always been under demand. So as demand kind of readjusted the gap got smaller versus it got under our supply. And that's why we're in a much more or better position than some of our peers because we didn't blindly ship to backlog while trying to figure out what is real, what is not. We pegged it to LTSAs. The decision we made also over the last few quarters when you're talking -- Thad and I talking about the structurally, how are we going to get ready for the uncertainty in, call it, the next coming quarters, whether it's 2, 3 or whatever the number is, we had a choice of, okay, do we bring lead times in and get utilization up. We thought it would be way more cautious and much better set up for whenever the recovery happens to actually take utilization down and keep lead times where they are because there's no business value of bringing lead times in, given that most of our business is under LTSA, we're getting that visibility. Lead time is an artificial number at this point because customers are booking to what they need. So we're more proactive about not focusing on building inventory and do that for the lead time rather than structurally getting us set up or whenever the recovery happens, like that said, all that turns into tailwinds.

Actually, 2 questions on that. Number one, on the utilization rate. So whenever the utilization rates go back up, what would be your gross margin, especially by then, I would assume that silicon carbide is back to like corporate average at least are a little bit better. What would the overall gross margins look like when you guys are back to, say, 85%, 90%?

Yes. So a couple of things. So silicon carbide will exit this year at corporate average for silicon carbide. So that dilutive impact peaks here in Q3. When you think about utilization, we've got a couple of things going on. So it's not just utilization. It's what we call Fab Right. So we went through the last 2 years doing Fab Liter, divesting those 4 fabs, shrinking the footprint, bringing on the East Fishkill fab for capacity. And now it's a matter of getting the mix right across the network, finding the optimal structure, driving capacity by actually doing that at the same time. But a rough rule of thumb, and it depends obviously what ramps, right? So it's hard to give an exact number. But if it was like-for-like, similar mix, every point of utilization has about 15 basis points of margin improvement.

1-5.

Yes, yes.

Okay. Great. And then on the lead time comment you made earlier, Hassane, can you give us a sense of where lead times, where and when your lead times peaked and where they are right now and how you see that going forward over the next year or so or 6 months or whatever time horizon you want to provide?

Obviously, every lead time peaked mid- to high...

We're still about there.

I am sorry, what was that again?

We are about high 40s.

High 40s, that's average, right?

Yes. Yes, average. We're still about there, maybe a few weeks off. There are some areas where we did bring it in. But to a first order, we're still at that level.

Yes. And you said you're going to try and keep them there as long as possible.

Until we see consistent signs of recovery.

Yes, okay. And then what would be your ideal lead time, not the customer because they want it tomorrow, but what would be the ON ideal lead time?

It depends on the business. It depends on the product type. So we have targets by every product type and every product technology, not to get into the weeds because if a product, if we have 1 customer, 1 product that's going to be built to order, longer lead time. If we have a broad range of customers its going to be shorter because we can kind of repurpose it. So that's the level we go to when we're architecting our lead time where we want to land. So strategically, we have those targets. We're just not, there's no reason to get there yet.

Okay. And is the shortage situation fixed? So is everything within like some sort of band around that 40 or something weeks? Or would you say that you still have some like shortages out there in terms...

We have some that -- technology base, there are some that are still constrained, where even now with demand where it is, we're still not able to support. And of course, that we have capacity coming online as we convert from the businesses that we exited. That takes time, but there are some technologies. If a customer comes in today and says, I need 10% more, there is not 10% more. So we're able to support the demand we have now. It's just upside like silicon carbide, for example, we've always said we're building 2 LTSAs. We're not overbuilding capacity and hoping to fill it because that never works. You always fill it. You may not just not like how you fill it. So we're signing LTSAs and then we go build capacity, and that is the cadence we're going through.

Great. One more thing on the lead times. So we've had some of your competitors here, and they've talked about, hey, we brought lead times down. Our orders fell up, we brought lead times down. Is there any fear or concern Hassane that if you're going to a customer right now and saying, "Hey, we can get you that product in whatever 42 weeks," then you have company X that says they can get in that product in 20 weeks, would that be a competitive disadvantage, have you ever run into that at all?

No. So 2 things. One is the products are not fungible. The fungible product where I can get it from you or from A or B, where they don't care whose logos on it, that's the stuff we've been exiting. So we're -- that's fine. If they can get it somewhere else or they can get it cheaper knock yourself up.

You don't want that product.

I don't want that product anyway because there's no differentiation back to my point on, we focus on value. So that's the primary driver. Number 2 is, it's not about lead time. Our customers have 5-year LTSAs. We know what they want. They know what they need from us over the next 5 years. Whether my lead time is 13 weeks or 40 weeks, I know what they want for 5 years. So that's what we built to, it's not about when it hits the backlog, we start scrambling to build, it is that long-term visibility, and that's the value of the LTSAs that we have, yes.

And in terms of the LTSAs, how do you feel about your backlog coverage for 2024 in terms of LTSAs. And is there an ideal number? Do you want 80%, 90%, 110%, what's the...

Yes. Right now, we're about on a 12-month period, we have about 75%.

Yes, 75%, 80%. Yes. So plus NCNR orders on top of that.

Plus NCNR, yes. So we're pretty well covered. And that's where we want it because you always want the flexibility, right? You don't want to tie everything because in the case a customer does come in and wants 3%, 4% because they're seeing strength, you don't want to be the bottleneck, and saying sorry, I tied it up elsewhere. So we have -- it gives us that flexibility. So we're comfortable with the coverage where we are.

Okay. Yes, if you guys are the bottleneck in autos, you don't get a call from Mary Bary, you get a call from Joe Biden. That's not good. Right, he can speak. So on pricing, has the LTSAs provided you a nice sort of pricing blanket for next year? And if you could compare on average, what pricing has done for ON this year in terms of how much it's increased? And would you expect the same sort of increase next year? Will that ease a little bit? Or could it be higher?

So there's no increase. All the benefit that -- think about it, '22 even and '23 there's no pricing. "pricing benefit" it's all in the baseline already from prior to that period. So everything here is all new product, operational excellence, self-help, gross margin, thousands of line items, is all our execution internally, on getting more out of our footprint, back what Thad mentioned, the fab from a manufacturing side, Fab Right -- Fab Liter to Fab Right. Any additional wafer we get out of each 1 of these fabs, all cost structure drops for all wafers because you've got 1 more wafer out, that's the focus that we've had. So pricing is not a factor. Mix will be because, as we've always talked, all new products are at or above our target margin. So as those products go to revenue and become a more material portion of that revenue. Of course, ASPs are going to be better and margin is going to be better, but it's not apples-to-apples price increase that's helping us. So that goes back to while we always focus on why our results are sustainable because it's been 2 years of that self-execution and the sustainability of the results even in a softer market.

Okay? One of your illustrious competitors was here maybe speaking with the German accent, talked about some expedite fees that had really helped them and have now gone away. Do you guys -- any expedite fees for...

I'm sorry, I'm laughing because somebody...

I am close to crack you guys up a little bit today.

You suddenly brought it up today, I'm sorry, I never counted on expedite fees to hit the margin target. Because we've always said ... .

They have never mentioned that before.

So it's a little bit funny. But we've said -- we did have expedite fees by the way. But we've always said and we've been very transparent about we're passing the cost onto our -- to customers. So if we're not getting an expedite fee, there's nothing to pass on to customer. So it wasn't a margin -- it was margin neutral for us. We did it just because, look, those are real. We got an expedite. The customer had a choice. Do you really want it. tomorrow or can you wait for 5 days. And we ship at USPS, whatever. That's the conversation we have with the customer. So it was customer choice. We never forced it because we were trying to hit a target model. So it's like nothing changed today...

Not germane to the ON situation, I'll take that.

It's got no impact. .

Sure. So one question on the long-term contracts. How iron clad are those? I mean I would imagine that in the part of your business like industrial that had somewhat of a correction, there might have been some renegotiations going on there. Are these tough nuggies guys. You said you're going to take the product, this is like the product you're going to take? Is it, okay, well, you don't have to take as much product, but we'll charge you some sort of cancellation fee or something like that? Or how does it work?

Yes. So look, if you rewind even the last year, 1.5 years, and people always asked the question about, well, really, you're going to go after your customer if they want to get out of it and so on. And I've always said one thing very consistently. If I get one thing from the LTSA, I'm going to get a phone call, as soon as the customer sees softness. That's it. I want a phone call because that's the conversation I want to have. Historically, backlog disappeared 30 days before I ship it, and you're left holding the bag. So the LTSAs, somebody is going to get the bat phone and call and say, we got a problem, 6 months, 9 months from now, let's have a conversation. So we'll have the conversation, but it has to be a win-win. What does that look like? Look, if demand is down, it doesn't really help us maybe in the short term, but it doesn't help the company in saying, "Well, you got to take, I don't care, put the inventory on your shelf" it doesn't help anybody to have inventory at the customer if demand is not there. So we'll have the conversation about, okay, well, we know we have 50% share on this product, and we increase the share, offset the gap and so on and so forth. So we have those conversations. From a company perspective, we would rather ship product any day than have fees because those are onetime benefits. I really don't care for them, right? I like the sustainability of it, which is, okay, let's talk about revenue. Let's talk about share gains. Hey, maybe we extend the LTSA and you bolt on more, let's put more products that you maybe are not buying from us, but we can qualify because we have a new product. So all of these are win-win. Now, if the customer is not interested in a win-win, but a win/lose, that's where it's ironclad. It's a legally binding agreement that both parties in a win-win will renegotiate an amendment. But net of that, we're not going to be left holding the bag. And so far, I will tell you, conversations with the customers have been very, very productive.

And the one thing I would add is pricing is locked, right? So it's not a pricing discussion. It can be a quantity discussion just if demand is changing but it's not a pricing discussion. And even over the last couple of years, we couldn't even keep up with LTSAs. So we couldn't commit to the full volume that, that customer wanted. So in some cases, we've been undershipping demand as well. But pricing is locked in. These things on average, are 4 or 5 years, and we've got them going out 10 years. that's good visibility on pricing.

Yes. And to your point, it's not all -- the conversations are not all negative because there are customers where we didn't have enough, so we were shipping 70% of what they really wanted. So now they're coming back and saying, okay, can you at least go to 80%? So some of them are net positive. They want more because we haven't signed up for more because we could really -- when we have supply assurance in the LTSA, I'm not going to sign up for something I can't deliver to. Well, now they're coming in or other customers that we didn't have capacity to even give an LTSA too. They've been looking at it from the outside they're in now and saying, "Okay, I want an LTSA for the next 5 years," whether the product is constrained or not they see it as a competitive advantage and a benefit for them Yes. Last quarter, we added $4 billion in LTSAs. You would say, okay, that's all net like incremental. So that tells you that customers now see the value as a tool versus as a protection.

Great. So let's talk about half year times if they can get any half year for you guys, silicon carbide. I have to ask the obligatory, how are things standing on the $1 billion revs for the year and the gross margin goals? I think you answered part of that earlier. It's funny. Parag and I commiserate every quarter. You guys have an illustrious competitor in silicon carbide, they blow up every quarter. I don't want to mention their name, it runs with both lead. But our phone lines light up. We spend 3 days saying, it's not going to hit on, and then you guys have a good number and it gradually goes away, but it's like clockwork. So just want to ask you still on track for the revenue targets?

Everything is -- when it comes down to operational, I can tell you that I control operation and execution every day. I can't do anything about the market, we'll have to navigate it. But when it comes to our team's execution, our team is executing greatly. And I appreciate every day they come in and deliver the results that we have, including on silicon carbide. Last 2 quarters, we talked about how silicon carbide is actually ahead of where our own targets were. Look, haters can hate. That's fine. Results don't lie.

What sort of growth rate or margin ramp could we expect for the next year or 2 beyond the short term for silicon carbide?

So what we talked about last quarter, we said silicon carbine margins doubled and we delivered high teens operating income. Now of course, the reason we delivered both because you can say, behaviors can say, "Well, you doubled from 1% to 2%." Well, okay. But when you give the operating income, it's not 1% to 2%, it's actually meaningful. We're still on track to exit the year with the dilutive impact at the corporate average for silicon carbide. So progressing very well. That's been our target since we started this journey, and we're on target. The other target we will also on track to meet is a majority of our substrates will also be internal. That's also a testament to our execution with our GTAT acquisition, that there's a lot of negativity around. But we've proven that we're executing very well to that, and that's on track to deliver a majority of our substrates. And that will -- because we're adding more capacity, you're going to see that margin kind of stay with the corporate average. But then as we get the scale and we get the absorption from the capacity side for silicon carbide specifically, you're going to see that business at or ahead our target.

What sort of revenue growth should we be looking at? Will you guys grow sort of market average, better than the market after this year?

Yes. So our focus for silicon carbide, if you we expect the market to grow about 30% to 35%, we expect to be about 2x the market.

Well, I'll take that.

So that's the growth we see over the next 5 years. And obviously, the trajectory is not a straight line, because you have ramps. It's a new product ramp, just like anything else. It's always as you add more and more customers and as you ramp LTSAs. We have the LTSA secured for that performance. So it's a bottoms-up performance. Right now, it's on execution mode.

And can you talk about the competitive environment? We had your 2 biggest competitors here earlier in the week present. And then it seems like every other month China has got some silicon carbide start-up. It's like whack-a-mole or something like that popping up there nowhere near commercial liability, but I'm sure you get this question all the time.

Yes. Yes. Look, so I don't look at -- as a company, we don't look at China any different than we look at European competitors or North American competitors. We look at the competitive landscape with exactly that, the competitive landscape, no matter what region they're from. Every region has got a different dynamic. We need to be able to win against all of them, and the only common denominator to win against small competitive threats is technology. . As long as we have the best technology, and we execute like we have been on that technology, we're going to win. I guess the proof is the LTSAs that we've had, we have $11 billion of LTSAs on silicon carbide, and that's what we're going to be delivering to. So you don't hear me talk about funnel and pipeline, and all of that is literally contractual agreements that we have with our customers to do just that.

Great. Two quick questions on gross margin for that because I know we're running out of time. So you guys are increasing the CapEx. I would imagine that's going to increase depreciation. Thad when do we think depreciation peaks? And is that going to be much of a headwind to gross margin over the next 2, 3 years?

No. The way we look at this is we're building the capacity to support the LTSAs, right? So it's not like we're not going to have revenue associated with that. So if you look at our depreciation today, it's 5%, 6% of revenue. As we go forward, we project that to stay basically in that same range even as we're adding capacity. . So we're going to -- our capacity or CapEx is going to be high teens this year and next year. And then it starts to taper down to 10%, 11% over the next 4 years. And that's -- at the same time, we're bringing on additional capacity even at that time because we go to 8-inch in silicon carbide, which is a capital-efficient move for us because most of the equipment is retrofittable. So short answer about the same percentage of where we are today.

And then last question on East Fishkill, the latest diamond in the rough, you guys are buying and polishing. Give us a sense on how things are progressing there and what the margin ramp should look like?

Yes, look, we've got a 250 basis point headwind that's in the numbers right now. The cost structure in that fab is much higher than we anticipated. It's blocking and tackling. We know exactly what we have to do. We have a whole fab network. We know how to benchmark material cost, cycle times, everything. So we've just got to go work on that. The challenge is the fab is fully loaded today. So it's hard to go take cost out when a fab is fully loaded, you have to be really careful that you don't break the fab. So we think by the time we exit '24, we'll have that dilutive impact back under control and back at the corporate average.

Great. With that, we're out of time. Thanks, everyone. Thanks, guys.

Thank you.

Thank you.