Great. Thank you, everybody. I'm Joe Moore back again, very happy to have with us the management team of ons emi, Hassane El-Khoury, CEO, and Thad Trent, CFO. So maybe if we go off script, breaking news today that you've acquired the United Silicon Carbide business from Qorvo. Can you talk a little bit about that? You know, what benefits do you guys see from that?
Sure. So if you think about the strategic rationale, of course, we're a power semiconductor provider, including silicon carbide and wide bandgap. And we've always said that the competitive advantage that we have is the ability to give customers a solution that's tailored for their specific needs and/or optimized. So how does that, the acquisition of the silicon carbide JFET business, fit into this overall strategy? And more appropriately, how does it fit the customer need? With the emergence of AI, the AI trajectory from a power perspective is increasing every generation. And now the AI server is going from 40 kW to, call it, hundreds of kilowatts. As that transition happens, you start to need more of wide bandgap technology devices rather than just the silicon, from an efficiency perspective, but also from a cost and area. Because a rack doesn't get any different.
The size of a rack, you know, the X, Y, and height are almost specced. So therefore, real estate is important. So you need a more compact power solution that is able to provide the efficiency when you deploy at scale. That's what the silicon carbide JFET does for us. It's a smaller die. It's a higher efficiency, higher frequency product than this regular, even silicon carbide. And with our ability to, over time, bring that in, it gives us that also competitive advantage from an operational side, in addition to the strategic side. So more power drives more efficiency requirement. More efficiency requirement makes silicon carbide JFET very appropriate for AI. Other markets that we play in, where that becomes very competitive also, and more suited, we have a trend in industrial for more solid-state breakers, for example. Silicon carbide JFET is suitable for that.
It's actually the most ideal solution for that. Fast forward into automotive, with the proliferation of electric vehicles, you have the battery disconnect switch that's going from electromechanical to now solid-state. Silicon carbide JFET, again, is the ideal solution for that. So you can look at it from a strategic rationale. The technology is superior, great team, great technology, now going into markets that are exactly where we go and play with the rest of our portfolio. So that's back to the cross-selling go-to-market that we have been pursuing.
Great. And Qorvo wasn't able to extract this value because they don't have the breadth of power that you guys have. Is that?
It's the breadth of power, but also it's the customer alignment and the channel alignment. So the Qorvo, the sweet spot of Qorvo's market, and they're successful in their market, is more on, you know, the mobile, the RF, and so on. And although they have some more diversification in their portfolio, that business requires a portfolio. Because sometimes it's silicon carbide FET, sometimes it's a JFET, sometimes silicon can do it. So you have to be able to give the customer that optionality, number one, but also be able to use our existing sales force in a very efficient go-to-market. You know, for us, our sales force will look at it as now it's another product in our sales bag for the same customers we actually call on today, through the same channel to go-to-market. So it's a very efficient go-to-market as well for us.
Great. Well, congratulations. Sounds like a great deal for you guys. The other new development was the development of Treo, the programmable analog. Can you give us an overview of that and how that fits into your long-term strategy?
Yeah. So Treo is, again, if you think about how we go-to-market, our focus is adding value to the customer system. And the way we go about it, we solve value through technology. So we've introduced the Treo, which is a mixed-signal analog product platform, which is based on BCD 65. So BCD 65nm BCD 12-inch internal fab at East Fishkill. What does that do? It's two components of competitive advantage. One is the technology component. BCD 65, as I mentioned, is 65nm, 300 mm fab, number one. More importantly, it is a 1 V- 90 V monolithic capability of the whole platform.
We can go from low voltage all the way to 90 V at 65 nm, which means from a competitive advantage, we are able to provide a single-chip solution for a 48 V rail, as an example, in AI, or 48 V with the zonal architecture in automotive, in a single chip, combining power and compute in a single platform. So that's what the platform does, and the ability to go up to 175 degrees Celsius. So very well-suited, auto, industrial, and AI for a very broad range of products. What we also said in the announcement is we were starting, actually, we introduced products. We're sampling products in all of these markets. Revenue starts actually in 2025 with a target of 2030 of $1 billion, with a margin of 60%-70%.
Again, when we talk about our portfolio progression over time to get into the higher-margin product, higher value, and drive the corporate margin towards that 53% target, that's part of that trajectory. On top of the technology, there's a process and capability that we introduced with the Treo, which is an SoC-like go-to-market. What it means, at its basics, we have a collection of IP blocks. And based on the product we want to make, we integrate and we consolidate those IP blocks to make a product. That capability, which is very typical in SoC, nobody's done it in the analog domain yet. We've been able to achieve it, which takes us from concept to sample to a customer in six to nine months. That go-to-market in the markets we're playing in, like AI, you have to be able to get that pace.
When AI generation is, call it, on average, an 18-month kind of beat, six to nine months is half of that. We're able to do multiple products to proliferate as soon as the architectural decision on the SoC is done. That's a different value outside of just the technology is that capability.
Can you talk about that with, you know, you're talking about margins that are the high-performance analog types of gross margins, you know, but you're competing there with companies that have broad catalogs in your, you know, how does that competition work? Do you have to have the same building blocks that they have to achieve this?
Yeah. So our intent is not to go in the general-purpose analog mixed-signal. It's very targeted to where we, as a company, provide back to the cross-selling as a go-to-market. Let me give you an example. We provide, and we're leaders in silicon carbide MOSFET, whether it's JFET or the MOSFET, or even IGBT. But let's take the silicon carbide. We're leaders in that market. If we're able to make a driver with compute and a good driver in it that is designed in conjunction with our silicon carbide FET, that together perform better at a system level, think about it as a chipset for kind of if your background is in the digital domain. A chipset always works better together, but they can work separately.
So having a driver that makes our MOSFET work much better, from a customer perspective, they will buy our driver when they buy our MOSFET. So it's that cross-selling that we're going after from a mixed-signal analog portfolio, not necessarily the general-purpose off-the-shelf analog that, you know, TI and ADI kind of do. So it's very targeted go-to-market where we believe we add value in conjunction with the rest of our portfolio. The same goes with the PMIC. The same goes in AI, for example. We have the best FETs in the industry. Integrating the point of load with some discrete or just discrete with controllers and PMICs and so on, we're able to provide the best performance for our customers because we have the best of both technologies. Others do not. So it's a very targeted, cross-selling-driven go-to-market, not just a blanket analog portfolio.
Yeah. Okay. Good stuff. Thank you for that. So maybe moving to the overall industry environment, you know, the number one debate we get for onsemi is gross margin. You know, you've done such a good job. You've got gross margins in the mid-40s% with utilization in the mid-60s%. Can you talk to the sustainability of that? You know, I mean, obviously, it's sort of a challenging demand environment. Can you sustain the profitability that you have?
Yeah. Look, our margin right now is driven by utilization, right? As you said, our utilization is 65%, kind of a trough utilization historically for us, and typically, in previous downturns, when we hit this level of utilization, our gross margins were high 20s, low 30s. So the fact that we're holding the mid-40 shows that we're structurally a different company. You know, from this point forward, whether utilization goes up or down, that's going to drive the gross margin. Given the outlook that we're seeing right now with the uncertainty, we think for the next several quarters, we're going to be, we'll be below seasonality for the next few quarters, so we may tick utilization back a little bit, you know, and margins will move slightly with that, but that's temporary, you know, and as utilization comes back on.
For every point of utilization, up or down, it's 15-20 basis points of gross margin up and down. So you can kind of do the math there. We peaked out utilization around 83%. So going from 65%-83%, you can do the math. It's a nice tailwind there. You know, like I said, in the short term, if we've got to take it down and not build a lot of inventory, that's fine. That's temporary. And when the market recovers, we're going to turn it back on. We've managed our inventory in the channel and on our balance sheet very effectively. If you look at our working inventory, it's 113 days. Our sweet spot is 100-120. So whenever this market does rebound, we're in a really good spot to turn utilization back on quickly.
Then, you know, there's about a quarter latency for that to hit the P&L. You start to see that through gross margins very quickly. It's not like we've got to burn through inventory at the disty and our balance sheet before we see utilization improve. You may see some short-term headwinds to utilization and gross margin, but, you know, I think that's all temporary. You know, structurally, we're a different company. We're positioned well for the market rebound.
Look, the balance you're trying to strike, of course, 2025 for us is going to be all about cash flow. You know, we don't control the market environment. We don't control the demand overall. So what we do control is how we manage our business. And managing the business for, back to like Thad mentioned, you know, focusing on not building more inventory than we need, keeping it in that spot will drive more cash flow to the bottom line. And with the CapEx going down a click now gets that cash flow, and you've seen us very active in share buyback, more cash flow generated, good capital return to shareholders. That's kind of what we're managing too in a 2025 environment.
Okay. Great. So to maybe double-click on a couple of those, you talked about the 113 days of operating inventory. You also have the strategic reserve, which comes up a lot. So can you just describe the rationale for that?
Yeah. So we've really bucketed our inventory into the two buckets. One is the base inventory, our working inventory that's 113 days. We have another 100 days, which is what we call the strategic inventory. That's for our fab transitions and the silicon carbide ramp. So just as a reminder, in 2022, we divested four fabs. We've been doing the last-time builds in those divested fabs, and we'll burn through that inventory over time. So as the market rebound happens, that's, again, another way we're going to create, you know, strong free cash flow, turning that inventory into cash. But when we look at truly what's that working inventory, that's in our sweet spot. That other piece is by design. It's just taking a little bit longer to bleed through because the market's softer right now.
Okay. Great. And then in terms of the lower capital intensity, I think you talked about going to a mid-single-digit level of spending. Can you talk about how long you'd operate at that level? What would cause that number to go up or down from here?
Yeah. Look, we don't see any change to that number. The big investments are behind us. We made big investments in silicon carbide, big investments in our 300 mm capabilities at East Fishkill. And the reason that we were able to take down that intensity level quicker than we anticipated is because silicon carbide's yielding very well. We don't need the additional capacity. So it's all, you know, internal that we're actually ahead of us. So when you think about our investments over time, they've been brownfield investments, which is much more efficient than greenfield. So from this point forward, we think we're going to be running in this mid-single digit. And you can think about it as primarily maintenance. There's a little bit of capacity here and there over a multi-year period, but most of it's maintenance because the big investments are behind us.
So like as Hassane said, even in 2025 with uncertainty, we're going to create a lot more free cash flow just year on year.
Yeah. Yeah. Okay. That's great to hear. So you mentioned the demand environment. I guess it was around this conference last year that you guys started signaling weakness in autos. You were early to see that. I think the same was true in industrial. You know, a year before that, you were the first ones to see that. So, you know, where are we now in that process of reducing customer inventories in automotive? And, you know, obviously, there's good underlying organic growth in those markets. When do you sort of see us returning to that from a semiconductor perspective?
So we're still in the inventory digestion period because that inventory and how long it takes is a factor of end demand. End demand is not really picking up.
Okay.
And, like we mentioned, even on our last call or on the conference last week, we talked about the next couple of quarters being below seasonality because look what happened in the last four to six weeks. You've seen a further development in the European market from automotive. Just look at the headlines. So there are challenges here locally that are different and softer than they were when we talked about it in our quarter. Now, why are we seeing it? Why am I talking about it? Not just the headline, but from an LTSA perspective, the same comment you made, why we see it early is because we get the phone call because the LTSAs allow us that phone call. So we're starting to see that coming in with concern, specifically European Tier 1s that are impacted by the European OEMs.
That's why we're looking at kind of the next couple of quarters below seasonality until that works its way through.
To what extent do you think that European weakness is a function of the success of your customers in China? Because it does seem like that's at least an element is that, you know, European luxury cars exported to China seems to be disappointing. I know you have a good presence with the customers there as well.
Yeah. Look, that may be from a financial stance, you know, that there is a slide. When you have, and depending on the OEMs, if you have 20% share or a market share in China and that kind of shrinks, it will impact your financials, top line, profitability, depending on the OEM and so on. But there's also an intrinsic issue in general, which is the demand environment. You know, you have a conflict in Europe. You have energy cost. You have fundamental topic. And look, let's be clear. There's also the geopolitical environment. And not just in the U.S., but, you know, in Europe, you have an election coming up in Germany. So these are things that have to work through wherever it lands, good, bad, or ugly. The common thread and what's important for the industry is stability and predictability.
Yeah.
And that's why I mean, it's going to work its way out. And we anchor on something. Wherever it is, the market will then adjust for it. That's when you start seeing clarity.
Yeah.
From clarity comes consumer confidence. Consumer confidence will drive demand.
I mean, the shortages that we saw. I've done this 30 years and I've not really seen that before two years ago where every automotive insurance analyst is calling me to figure out when we're going to start building cars again. So like, it makes sense to me that there would be a buffer of inventory that got built after that. You know, I guess what's your visibility into how much may still be there and when do you think that may start to clear?
So that's a broad one. It's hard to say, like, oh, in automotive, there's so many. And because it's customer by customer. And to the extent we can, we drill in on a customer-by-customer basis. You know, inventory in the channel, it's easy for us to do. We have visibility. We know what's in there. The distributors will take more if we ship to them. But we're holding back because we're managing weeks of inventory very tightly. But when it gets to the customer, you have some customers in automotive that are down to a couple of weeks, some that are a little more strategic. They talk about two to three months. But as demand uncertainty happens, those two to three months, they go, well, maybe we can go with four weeks and five weeks until the market kind of gets better visibility.
So that's the moving target that we're seeing, but it's not the same across all customers. And the same is true, by the way, at the OEM. If you look at inventory, car inventory across all the OEMs, it is a very broad range. So you can't look at it as, well, you know, it's healthy inventory, you know, plus or minus a few days on an overall. You have to look at it on an OEM-by-OEM basis. Then the discrepancy and the distribution is much higher. Within the same OEM, different brands have different inventory levels.
Okay.
We're looking at it at that level. I track inventory at our customers as they do for the public companies that announce. You can kind of modulate. If somebody has a ton of inventory and they're giving me a rosy picture for 2025, I'm going to be very hesitant to take it at face value.
Okay.
We're doing our due diligence to make sure it doesn't get ahead.
Great. Okay. On silicon carbide, you've had some nice proof points with the win with Volkswagen earlier in the year. You've made investments in brownfield, kind of local for local locations. But it's obviously been kind of a choppier market than you would have anticipated. Can you talk about the outlook of silicon carbide in adoption going forward?
Yeah. We're seeing silicon carbide, you know, silicon carbide is a growth vector for us. We also see silicon carbide growth into 2025 because there are two factors that are driving silicon carbide growth. You talk about electric vehicles. And what I mean by that is electric vehicles as a percent of total vehicles made. So there's going to be more EVs in 2025 than there were in 2024, even in a, let's say, flat SAAR. So that's the penetration. More importantly, there is a penetration of silicon carbide into EVs. So if I take out the North America disruptor, North America EV leader, if I take them out of the equation, the penetration of silicon carbide in EV is only about 6%. So just silicon carbide getting designed in and ramping within the same EV baseline from a unit perspective is going to grow the silicon carbide specifically.
As OEMs shift to 800 V, silicon carbide becomes the norm for electric vehicles on the 800 V with a 1200 V silicon carbide device. That's what's driving the growth. It's penetration of EVs, of total vehicles, and penetration of silicon carbide into EVs. Both of those are what we see from a growth perspective. It's going to be lumpy. Overall, the trend is growth because of these two factors.
In terms of on's capabilities here, I mean, you have vertical integration that nobody else has with your own internal substrates. But we've also seen, you know, China substrates starting to become more viable. Can you talk about those dynamics?
Yeah. So our strategy has always been we're not going to be 100% internal source. We have the capability, which is a very important distinction. Having the capability meaning you can stretch up or down. But we have the capability. We said we're going to be about 70%-80% internal, 20%-30% external. Some of it is from China. And we were successfully qualifying good substrates from China. So that's a path. There's a geopolitical aspect of it where having all your substrates come from the outside, or China specifically, but just in general, if you don't control your supply chain in this environment with the geopolitical uncertainty, that starts becoming a risk for our customers. So our customers look to us, back to what you mentioned. They look to us as we have good supply assurance, but more importantly, supply resilience. We're able to get inside.
We're able to get outside and we're able to stretch inside if needed or reduce inside if needed. Where it's going to land is going to be back to we need to look at a stable geopolitical environment. If you can access unconstrained China substrates, that's one model. If it's kind of in the middle, you have to be able to differentiate between the two. That's why we're talking about, you know, potential expanding investments in Europe to be able to also provide our European customers more of a supply resilience with a full end-to-end capability of silicon carbide in Europe in addition to our worldwide. So we feel we are in the best place from a supply resilience and with our device capability, you know, device and package capability and the superior performance of our products, we are very competitive.
And that's why we're winning a lot of the designs that are coming up, like you referred to with the VW.
Yeah. Okay. Let me see if we have questions from the audience. If not, I can continue. Lee?
Great. Thanks. Hassane, just maybe a quick one. It looks as though Chinese platforms are moving a lot quicker. Like every 24 months, we get a new platform. So help me understand, is this having an effect on your product cycles? And when you look across the peers, who do you think suffers in that sped-up cadence? Thanks.
I'm not going to say who's going to suffer, but the market will tell us. But to answer your question, yes. The pace of innovation in China is very challenging to a typical design cycle from historical semiconductors. You know, people talk about a two-year kind of generation to generation. Sometimes it's two to three years, depending on it. So it changes the way we approach product development and the way we approach technology development. You know, I gave the example of Treo where I very distinctly talked about you have the technology and then you have how quickly you create products in a six- to nine-month. That pace is you have to have consideration when you develop the technology. Otherwise, you have technology. If it takes you two years to develop products, you're kind of going to miss. You'll do every other technology at the customer.
So you have to rewin your share almost, so we do it very effectively by when we design a technology, we design the go-to-market platform capabilities with it. And that makes us very efficient as far as being able to service the market at a faster beat than the design cycle, which gives us two bite of the apple because if you get two bite of the apple, you're going to hit every design cycle and every what they call in automotive, a facelift, meaning they're not redoing the car, but they're changing a lot in it. So you can get every single one, and that's how your market share starts going up over time because you don't have to wait a two-year or three-year to get the next opportunity to get designed in. Very, very important distinction.
Our ability to move fast, we're able to move at, you know, what our customers call it, China speed. And that's why in China, we're capturing a lot of market share. And we're still winning in the market share because we're providing an onsemi performance and efficiency at China speed. And that's the best combination that the customer can expect in China. What we are seeing, though, is in Europe, it's getting us closer to the OEM because now we're able to have iterative discussions with the OEM and do samples way ahead. So the OEM can do, you know, six to nine months of their own work and validate before we even start making the actual projects. So it puts us at a very competitive advantage. Perfect example, VW again. VW announced an MOU ahead of the win. During that time, we're using simulations.
We're using the go-to-market, the fast iteration, and then we got the award. Very important in the new world of how you go to market, which is directly where the value creation is, which is at the OEM because the efficiency we get at a device level means higher mileage at the OEM, not lower cost at the Tier 1. Very important distinction, and that's helping us because we had to get to the China pace, which helped everywhere else.
Other questions from the audience? We have a minute and a half left. Maybe we could close with just, you know, any of this weakness drive any change in the pricing environment? And I know you have good LTSA coverage. Maybe you need from your competitors. Just how do you see the role of pricing going forward?
Yeah. We're not looking at pricing as a lever because, look, if there's a demand environment, a weak demand environment, you would argue there's no, there shouldn't be any price elasticity. You drop your price, you're not creating more demand.
Yeah.
You just brought the market down. You destroyed your own value. So if you believe that you bring value to the customer design, then the pricing is not a conversation. It's the value of the product didn't reduce just because the market environment is soft. It actually is more important because it provides a competitive advantage. How we look at pricing over long-term, you know, everybody asks me the question of, are you seeing this annual price reduction? And my point is, I don't believe in that this whole, oh, January 1st pricing clicks down for whatever reason. But take the example of BCD 65. You go in, the customer has a solution. We're able to do that solution with a very competitive value-based product. The ASP may be lower, but the margin is 60%-70% because it's a smaller product, more compact, more advanced nodes.
It provides a value to the customer. We're going to capture some of that value, but net net, we both win, and our margin goes to 60%-70% on that product. Is that a price reduction? I would argue no. That's a market competitiveness because the technology enables you to do that. You'll see that, but not a same product, same volume, or lower volume, and you're having negotiations for a year-on-year decline just for existing.
Great. Well, we're out of time there. So we'll wrap it up. But congratulations on navigating all this so effectively.
Thank you.
Thanks, Joe.