Good afternoon, and welcome everybody to the final day of Needham's 26th Annual Growth Conference. My name is Quinn Bolton. I am the semiconductor analyst for Needham and Company. It's my pleasure to host this fireside chat with Allegro MicroSystems. Allegro is a leading supplier of magnetic sensors and power ICs, targeting the automotive and industrial markets. The company holds the number one share position in magnetic sensors worldwide, as well as a leading—as well as the leading position in current sensors. We believe Allegro is well-positioned for growing content per vehicle, driven by the trends towards safety and ADAS and electrification. Joining me from the company are President and CEO Vineet Nargolwala and CFO Derek D'Antilio. Vineet, Derek, thank you for joining us.
Thanks for having us, Quinn.
Good morning, and thank you.
I wanted to start with just sort of some general questions about business conditions. You know, I know we're late in the quarter, quiet period, so, you know, if anything gets too close to earnings, you know, just say so. But, you know, let me start with kind of the Automotive Market. It's really, to date, the only end market where we haven't yet seen a major inventory correction. But, you know, a couple of weeks ago, we got a pretty nasty pre-announcement from Mobileye that I think has raised concerns OEMs and Tier 1 suppliers may have been building inventory for the past couple of years. Can you give us your best sense where you think, you know, OEM and channel inventory are in the automotive segment?
Sure, Quinn. You know, I've been in automotive for a very long time, longer than I would care to mention. But, automotive OEMs have traditionally not been known for building inventory, right? So they've really invented and perfected the just-in-time model, and I don't think they've really gone away from that, even in the dire days of the pandemic and the supply chain crisis that followed. However, their contract manufacturers and tiers certainly have held inventory. Traditionally, that's been pretty lean. I think during the pandemic and the supply chain crisis, they were asked by the OEMs to hold more inventory, and the OEMs, very uncharacteristically for them, had given them some subsidies to go do so, right? So they were giving price premiums, they were giving support, in other ways around expedited payments and some additional inventory payments.
As that sort of has come to an end, and higher interest rates and continued higher interest rates are obviously taking a larger bite in terms of carrying costs, we are gonna see contract manufacturers and tiers pare back their inventory positions. I think that's natural. That's healthy from an automotive perspective. But it does mean, and we had some conversations with OEMs at CES last week, where the OEMs still want their contract manufacturers and tiers to hold, you know, call it somewhere in the range of eight-10 weeks of inventory, but the CMs and the tiers are trying to pare it back more towards the four weeks or five weeks of inventory.
So I think over the next couple of quarters, we will see some pare back of the inventory, but what's more important here is, what does the end market demand look like? Because I know a lot of people are trying to draw comparisons to industrial and consumer. The backdrop here is that the automotive end market demand continues to be pretty stable. Last year, the global automotive production was somewhere in the range of 88 million units. This year, it might be closer to 89 million. Within that, EVs grew about 29% last year. They'll probably grow about 27%-28% this year. So the end market outlook is pretty stable. So inventory adjustments are a natural part of the industry. They will happen. They will continue to happen. But what I focus on is, what is the backdrop of the market?
Compare that or contrast that to industrial and consumer, where not only was there an inventory buildup, but the end market demand really fell off, which exacerbated the inventory position. So I think it's important for us to recognize that some of the inventory management that will happen in automotive, that is happening in automotive, is natural. It'll continue to take place over the next couple of quarters, but it's a different dynamic compared to industrial and consumer.
Got it. It sounds like two key, key parts of that message, at least it feels like I would take away from that is, one, you know, we may be talking a few weeks of adjustment at the contract manufacturers. It doesn't sound like, you know, it's months and months of excess inventory, and the fact that the end market demand is still pretty healthy. You know, that seemed to be, you know, as part you know, in the consumer markets, the fall off in end demand really exacerbated that inventory correction, and just doesn't seem like we're seeing that on automotive production, especially, with rising EV and ADAS penetration rates.
Exactly right.
Got it. Okay. Thank you for that, because I know we've gotten a lot of questions around, you know, thoughts on the auto and inventory situation. Another question, I think on the last call, you had mentioned that disti inventories were sort of in the range of 10-12 weeks, maybe at the higher end of that. Is that still sort of where you see the distribution inventory levels?
Yeah, so let me start off on that, and Derek can chime in as well. So, from a contextual standpoint, our distribution business largely serves industrial and consumer. There's a little bit of fulfillment that happens in automotive, largely in Asia, but predominantly it's our industrial and consumer business that goes through distribution. And so, we just talked about the end market demand and the resulting inventory overhang and the digestion that needs to happen. So we do see our inventory levels still remain at the higher end of the range. The point of sale is starting to look encouraging, but it's still early days, and we've got to watch that pretty carefully.
Yeah, and that really varies by region, Quinn, as Puneet started to talk about, right? When you look at that, that average I gave, right, that's the global average. There are some pockets of the world, like North America and Europe, that still need some inventory fulfillment. We were stocking the channel in Japan earlier this year as we made the transition from Sanken. But parts of Asia we've talked about for a while, particularly Taiwan and China, for industrial and solar and data center, have been at elevated levels, and we expect that to last a few quarters.
Okay. Yeah, that was, that was my next question, is in some of those industrial and non-automotive markets, still sounds like you, you think we may have another quarter or two to go in that inventory correction process.
That's right, and that's what we talked about on our Q2 call at the beginning of November, that that's probably a couple of quarters to digest through that inventory in the clean energy space and in the data center space specifically, which is really in Asia.
Got it. Knowing it's late in the quarter, I'll ask, but you know, kick this to earnings if you feel that's more appropriate. But just, can you make any comments, you know, as we're coming through this inventory correction, just how your order environment has been? My guess is it's probably held up better on the auto side, been a little bit weaker on the industrial and other segments, just given the inventory correction. But any comments you can make around just order trends or the order environment?
Yeah, without getting into too much detail, I think what you laid out, Quinn, is directionally correct. You know, we have always maintained that our visibility to the business really comes from our design win pipeline. That continues to be really robust, whether it's with our core portfolio or our TMR technology, which we are very, very excited about and that we've now augmented with the Crocus acquisition. That continues to be super strong. You know, and we'll get more into our order patterns when we get to our meetings.
Perfect. Okay. Just a couple more for me. You know, and I think this has probably been the case for a couple of months now, but, you know, sales in the U.S. of electric vehicles, you know, seem to be a little bit softer, and that seems to be perhaps skewing some investor expectations. I think data elsewhere around the globe, especially in China, has been a little bit stronger. But just what are you seeing from the different geographies in terms of EV demand?
Yeah, it's a great question, Quinn, and you're not alone or we're not alone in wondering and worrying about this dynamic. You know, from our perch and where we sit, we have visibility to all the OEMs globally, and one thing that is in common is that every one of these OEMs is investing very heavily in their EV portfolio. In fact, I can't think of a single OEM that is investing in new ICE platforms, so it's all on EV all the time now. And so, you know, there is a dynamic or a different dynamic in each geography. We know, and we see the proof. The Chinese OEMs are clearly ahead of the pack. I think they're the ones that have really figured out the economics of building compelling and cost-effective EVs.
I think they've put their arms around the battery supply chain, as does, as do some of the OEMs, you know, the, the, obviously, major North America EV OEM really has got the economics figured out. I think some of the stumbles we've seen, maybe some of the trepidation we are hearing in North America, is largely around some of the legacy OEMs as they are figuring out the economics of EVs. And so there's always gonna be a few fits and starts, maybe some stumbles out of the starting blocks. But we have visibility to everybody's three to five year plans. They're all very robust in terms of new EV programs, EV designs, and they're all pushing the envelope to get to the right economics as quickly as possible.
Which is why we also see the design cycles, at least initially, compressed for some of the new EV platforms as OEMs try to bring out the latest, greatest technology in quick cycles to the market. So short term, I think, you know, as I've said before, the EV transformation is not gonna be a straight line, and there's gonna be a few fits and starts, but the long-term trend here is pretty, pretty compelling.
Got it. Thank you. You had mentioned TMR, that was sort of my next set of questions. The technology is a newer, sort of more precise magnetic sensor technology.
Yeah.
Can you just... You know, for folks who may not be familiar with TMR or the XMR technology, what are the benefits of TMR sensors versus Hall effect?
Yeah. So, you know, I'll remind everybody that we are the leader in magnetic sensing by far. We have a really comprehensive and a vibrant portfolio of Hall effect sensing technologies, as well as MR sensing technologies. We also believe that Hall will be perfectly fine for a broad swath of applications for a long time to come, and we'll keep investing in Hall. What we are more excited about now is the TMR portfolio, and as applications become more demanding, the need for higher accuracy, better resolution, better response time, is really starting to become more pervasive in applications around advanced steering and braking, around battery management, and really understanding very accurately the current draw across the different parts of an electric vehicle. That's where TMR comes in. TMR stands for Tunnel Magnetoresistance.
It's a very advanced form of magnetic sensing, and it is eight times more accurate than Hall. The response times are in nanoseconds versus microseconds for Hall. And from a power draw standpoint, TMR is really ultra-low power. So any application that runs on a battery will want to use TMR because it will draw very little power and can make the battery last longer. So one great example is the application we inherited through our Crocus acquisition for continuous glucose monitoring. It's a small application today, but has really great potential, and these are consumable devices, but they need to operate for at least two weeks, and the TMR application is perfect there. So we will see more and more demanding applications, whether it's in automotive or e-mobility, whether it's in industrial, clean energy or automation platforms, gravitate towards TMR.
We had roughly pegged the SAM to be about $1 billion by the end of the decade. We think it's gonna be much bigger than that, as we've dug into the technology and we've had customer engagements. Anecdotally, you know, last week at CES, we had a really great slate of customer meetings, and over 75% of the customer meetings were focused on how we can accelerate TMR to bring into their applications and really get them to use and take advantage of this technology. Couldn't be more excited about where the industry's headed, where we are headed with TMR, and how we can serve our customers in a much more value-added way with not just our native TMR, but also what we had at Crocus.
Understanding it's probably a, you know, more expensive manufacturing process than Hall Effect, I could imagine that, you know, if it's much lower power consumption, you know, it could apply to a good part of the EV auto market over time. You know, anything that's, that's power sensitive, anything that's high performance sounds like a natural, you know, use case for TMR. But do you see TMR taking over a good part of that higher end of the market, particularly in EVs over time?
We believe so. So the more demanding applications should gravitate towards TMR, so whether it's battery management, solutions, whether it's, more advanced steering, braking applications, maybe more advanced inverter and onboard charger applications. Anywhere higher resolution, more accuracy, faster response time is required, we see that being a natural fit for TMR versus Hall. And over time, we think the economics of TMR should look very similar to that of Hall, but we have some work to do to get there. But-
Okay.
I think the margin profile will look very similar.
You had mentioned the higher precision. It certainly sounds like current sensors and, you know, measuring current flow off EV batteries certainly could be another application.
That's exactly right.
Okay. I wanted to ask again about the Crocus acquisition. Allegro had been internally developing TMR technology. You went out, you acquired Crocus to sort of complement your internal design efforts. What did Crocus bring to Allegro? You know, were they sort of applying TMR in different end markets? Did they have a sort of different flavor of the technology that might be more applicable to certain subsegments of the market? You know, how does it - how did it complement Allegro's TMR development efforts?
Great question, Quinn. You know, we had been investing in TMR for almost a decade, and we are in the market today in automotive TMR. So, the top two EV manufacturers globally today use our TMR position sensing. But we had really developed TMR for automotive. Crocus, on the other hand, had come out of the MRAM world and was really perfecting the TMR stack, on a standalone basis and had found application in industrial and consumer, including medical. And so we felt that, you know, while our overall TMR solution was really strong, their TMR stack on a standalone basis was very superior and really best in class, and so a natural fit. The physicists that we've inherited with Crocus are now an integral part of our engineering team, working very closely with our IC designers.
And so as a combination, we believe we've got the best TMR technology in the world, and the applications are complementary as well. Crocus TMR wasn't auto-qualified. We're in the process of doing that now, and very soon we'll be able to sample the XtremeSense TMR, which is now the joint brand for Allegro, across our TMR portfolio, with automotive customers. We're well in the process of doing that, and we will then also expand the position that Crocus had in industrial and medical, with our own portfolio and our own sales resources. So it's really exciting to see the two teams come together. By the way, I'm really pleased with the speed of integration. In the two months that we've owned Crocus, the business is now completely integrated. All Crocus parts are now available to order through the Allegro channels.
Our sales people are empowered to go sell Crocus, and so it's really gratifying to see the two teams come together, and it just highlights how middle of the fairway this acquisition was for us. We know the technology, we know the business, we know the customers, we know the supply chain. So from that perspective, it's just been a, you know, a complete slam dunk for us.
Excellent. I was gonna ask how integration is going, but it sounds like that's, that's done. The other question I had was, you sort of touched on it, your efforts to make the Crocus TMR sensors automotive grade. Sounds like that process is underway. How long does that take? Is that something that's measured in months or quarters, or is this a, you know, kind of longer process, where you've got to do burn-in and really, you know, kind of stress the quality to bring it to the automotive grade level?
... Yeah, it's another great question, Quinn. So in diligence, we had already put the Crocus TMR stack through its paces, and we are very close to having what I would call a generic auto qual for the Crocus TMR stack on our CMOS. Now, as we start to engage automotive customers, there is a unique qualification cycle that'll happen with each customer and each part, and it's multilayered, right? So we will qualify the part in our labs, in our four walls. We'll then qualify the part as part of a subsystem with a tier or a contract manufacturer, and the final qualification will be done by the OEM at a vehicle level through winter and summer tests.
So it's automotive, safety is always top of mind, and so, you know, it'll take a few cycles, a few couple of years before we start to see the Crocus TMR parts in cars on the road. But the auto qual process is already underway, and very soon here we'll be knocking, you know, notching design wins with automotive customers, with the Crocus TMR stack and our CMOS, and moving very quickly to production.
Great. Wanted to move to a few product and design win-related questions. The first kind of just a bigger picture, Tesla's Cybertruck, I think, may be the first vehicle manufacturer to use a 48-volt electrical system throughout the entire vehicle, and I think Tesla has begun to recently share that architecture with other auto OEMs. I know that I've seen some press releases talking about sharing it with Ford. Generally, how do you see Allegro transitioned as we move from a traditional 12-volt to a 48-volt electrical system in vehicles?
Yeah, it's a great question, and I think this is an exciting development in the automotive industry. I've got to give our engineering teams a ton of credit. They were very prescient in seeing a future transition towards higher voltages. And so about 5 years ago, we converted most of our portfolio to be scalable all the way from 6 volt to really 110 volt. And so the good news is, most of our portfolio is already ready for 48-volt. And if somebody were to read the specification that Tesla put out on 48-volt, essentially it's a manifesto of how to build a 48-volt architecture, you would notice that it has a bunch of Allegro parts listed. And so the inbounds from other OEMs have only increased in intensity.
Everybody had been experimenting with 48-volt architecture and sort of put it to a side, saying, "Well, maybe now is not the right time." But believe me, now has become the right time, as that document was put out, and now every customer wants to talk to us about, "Are your parts 48-volt ready? And we notice you're on the list of parts that was put out as part of the specification." So I think for us, this is a very natural extension, and now every one of our customers, other than Tesla, is interested in sourcing our parts, and they're relieved to know that our parts already work on 48-volt, as opposed to requiring new developments.
Yeah, no, that's great to hear. I didn't realize that Tesla's architecture, you know, kind of went as far as listing, you know, sort of approved or, you know, specific part numbers as part of that architecture, so that certainly seems like a strong endorsement.
Yeah.
Good. Moving to the next question, can you, you know, give us any sense on what you're seeing in the data center side of the business? I know data center cooling fans has been part of your business. It's been a little bit slow for a couple of quarters, but you know, first maybe just an update on the data center opportunity.
So Quinn, you know, data center is part of our industrial business. You know, we call it cloud infrastructure. Industrial overall is about 20% or less of our complete total company revenues. Having said that, the design win activity in data center continues to be really strong, and we are seeing more engagement on AI-related solution sets, which is really liquid cooling as opposed to air cooling. A lot of our tiers that we work with, the solution providers, have started to really double down on their investment in liquid cooling. That naturally results in more content for us as these solutions are more complex, require, in addition to motor and motor pumps, they require more IC content to ensure that, you know, the right liquid at the right temperature is getting to the right places.
So long term, we feel really bullish about what it means in terms of data center activity and design and momentum. Obviously, as we've been pretty transparent, right now they're going through an inventory digestion period, which we still think is gonna take a couple of quarters to work through.
Okay. No, I was gonna ask about liquid cooling, thinking that it might be a headwind to the, to the cooling fan business, but it sounds like, you know, you, you, you're seeing opportunities in terms of motor drives or the pumps to, to move the liquid around the liquid cooling systems, and, and sounds like you may have even higher dollar content in some of those emerging liquid cooling applications.
That's correct. It's still early days in terms of the architectures evolving around liquid cooling, so it's hard to sort of pin down exactly what that content might look like. Suffice to say, it's certainly bigger than that of just fans.
Yeah, and I think it's Supermicro has sort of stated they think that perhaps as many as 20% of data center racks would need to be liquid-cooled, so that sounds like that could be a nice opportunity for Allegro as well over time, understanding that we're going through some inventory digestion in the near term in this segment.
That's correct.
Got it. You know, kind of moving to other applications within your industrial segment, can you talk about some of the larger opportunities you see in the clean energy segment?
Sure. So when we think about clean energy, it's pretty vast, centered around solar, EV charging, and then battery-based energy storage solutions. And so the backdrop here is that as the demands on the grid continue to increase, and there are continued restrictions, if I can put it that way, around supply, clean energy, specifically solar, is probably gonna be the fastest-growing source of new energy globally over the next, call it 7-10 years. But that's not gonna be nearly enough. And so storage solutions, which draw energy at off-peak hours and store them to supply at, you know, peak hours, whether it's for charging purposes or for other consumption, I think are gonna grow exponentially.
When we look at solar growth, when we look at growth in distributed energy storage solutions or micro-grids, as some people call them, or EV charging, and you know, obviously there's been a lot of news about more investments in EV charging stations, whether it's to build new ones or fix old ones, we feel really strongly that this is a great opportunity set. We solve the same problems that we do in this sector, that we do on vehicle for electric vehicles, right? These applications require inverters. There are chargers or charging infrastructure. There are batteries that require management. So the solution set that we have to offer to intersect this opportunity is very similar to that we offer to electric vehicles. Obviously, the parameters are a little bit different.
Less stringent, if you will, so, you know, they're less safety critical, so, you know, we just have to tweak the parameters a little bit to meet the requirements of this market. So feel really good about our position here.
It sounds like that would be sort of likely using a lot of your current sensors and power ICs in those applications, I imagine.
Exactly. It's the same portfolio. It's current sensors, it's motor drivers, it's gate drivers, it's power management ICs. So it's really the vast majority of our portfolio that we apply in automotive that will get applied here, and that's our strategy, right? We're an automotive-first company, so everything we do is automotive grade, and we do it, we design, build, and launch platforms for automotive, and then we tweak it to apply to these industrial markets where we get great technology leverage, and that's really our business model and our strategy.
You know, we talked earlier about the consumer markets kind of face both inventory correction and slower end market demand. I know the solar market today is pretty soft. Is that sort of similar to consumer markets, where you've seen both inventory and a reduction in end demand, or do you think it's largely inventory, maybe a little bit more similar to what you were talking about for the auto market, where end demand is healthier, we just have to purge some of this excess inventory?
Yeah, it's more the latter, Quinn. I think the end market demand continues to be good. As I pointed out, really, solar is the only way to bring on new sources in a pretty short period of time and in a cost-effective way for, you know, most of the regions globally. So we think end demand continues to be strong over the horizon. It's really the near-term inventory digestion that we have to work through.
Quinn, in the consumer market, the last, you know, 10% of our business, which is the consumer market, right, that does have some applications for PCs and gaming-
Yeah.
That's a combination of both. That's, that was a weak end market that many peers were experiencing, coupled with the channel inventory digestion because of the weak market.
Right.
Yeah.
Yeah, no, and hence why I think why it was a much steeper correction, you know, that not only Allegro but other companies have seen.
That's exactly right, same dynamic.
You guys on the quarterly calls, you know, do a great job of giving us an update on your automotive design win momentum. You know, I assume, but I'll ask, you know, I assume it's continued pretty healthy through the December quarter. But anything to call out on the design win? Are you still feeling good about the momentum you're seeing across the automotive customer base?
Yeah, we are still feeling good. We'll, we'll have more color to share during our next call.
Okay. We will, we will stay tuned. Another question we get for a lot of companies, you know, China, I think, is sort of 20-25% of your sales. Obviously, there's a push more and more within China for local production. Are you starting to see the Chinese, especially the EV, you know, companies, looking to source, you know, sensors and power ICs from local Chinese suppliers? Are there any, you know, really local suppliers in China of these devices that you know have the quality to you know to compete in the automotive market?
Yeah, it's a great question, and you're right, China is about 25% of our sales, Quinn. And China is really important for us, right? As we think about our future growth, it's a big part of our growth plans, and so winning in China with the Chinese OEMs is really important. The other dynamic I would point out is that, you know, used to be most of our business in China, the majority was design wins that were logged outside of China, but our ship to was China, so we were winning with global OEMs. That's pretty flipped now for us in the last few years, where most of our business now is with the local OEMs, and it's designed in China, so the design wins are in China. So the local sourcing has become even more important.
You know, we have moved in the past year and a half very aggressively from what I would call a monolithic supply chain to one where we have more options for our customers. Specifically for our Chinese customers, we are now moving towards a China-for-China supply chain, which would involve definitely local OSATs, local assembly, and final test. We will be shipping locally as opposed to shipping from outside China, and then eventually we'll source some wafers for some of our select parts in China as well. We will appear more local and be more local, as a source. Our China OEMs and customers tell us very clearly, they love our technology, they wanna keep buying from us because they wanna be seen at a global level as a high-quality product. They don't wanna compromise on quality.
But they would love for us to become local, and we wanna serve our customers in China, and that's why we are going down this path.
When we were in China, Quinn, you know, we're talking about the 25% of sales that's shipped to China, like most companies would report.
Mm-hmm.
There are three types of customers. There were the global OEMs that are North American and European, they're the ones who brought us into China. That's still a decent-sized piece of the business that largely gets re-exported outside of China, so subject to the same standards, emission standards, safety standards in the EU and the United States. There are also the Chinese global manufacturers, right? The large manufacturers that are public in the United States and Europe, that are making inroads in Europe, same, subject to the same standards, so they're using the high quality of our products. And then there are the local Chinese manufacturers that are manufacturing just for the domestic market, where there is clearly more competition.
Got it. No, thank you. Thank you for that color and your strategy to be seen... You know, to become more local in China certainly makes sense as I think those supply chains bifurcate, you know, over the next few years. Wanted to sort of ask a last question just kind of around the business before getting to some financial questions. I believe you've now introduced a couple of gate drivers from the heyday acquisition. You know, can you give us the update on the heyday or the gate driver opportunities in some of the, you know, EV and energy infrastructure markets that you're going after with those gate drivers?
Sure, would love to. So you're right, we've launched a couple of gate drivers. Our high-voltage isolated gate drivers that was built on the heyday acquisition for GaN based devices right now, right? So we've launched two parts. We're seeing some really good momentum with customers, and early this spring, we will be sampling drivers for SiC devices. And so our automotive customers can't wait for us to bring those SiC, not SiC devices, but drivers for SiC devices to market. Having said that, we are seeing some automotive customers experiment with high-efficiency GaN-based devices. So for lower voltages, GaN, it's not just our belief, but it's the industry belief, that GaN will be more efficient. And so there are some OEMs that are actually looking at GaN as well.
But within a few months, we will have a complete portfolio to drive GaN and SiC devices. And the advantage of our gate drivers, compared to what's available on the market, is that we really combine three chips into one. It's an isolated, so we have the isolation, we have the transformer, and the gate driver, all functionality baked into one device. So we can save about a third of the package size, and that becomes meaningful when you think about there are as many as 32 SiC devices in an onboard charger, right? So there's huge space savings, and of course, there are system cost savings, which is not lost on the OEMs.
Is the application of GaN more in the onboard charger, or are you starting to see GaN potentially even in some of the traction inverters, maybe more for 400-volt batteries than 800-volt? But where are you starting to potentially see GaN play in the auto?
Yeah, I appreciate you drawing the distinction, Quinn. It is more in the onboard charger now. I think inverters will stay with SiC for the time being. But GaN continues to evolve and expand in its application set, and so we'll... You know, let's watch this space, right? I think both technologies have plenty of room to coexist, and we will have drivers for both of sets of devices.
Perfect. We've had one question come in from the audience, and then I'll move to financial questions. The question is, are you exposed to Chinese EV companies and batteries?
The answer is yes. So we are working with all of the Chinese OEMs, but especially, as Derek pointed out, the global OEMs from China. The ones that are leading the charge around EVs and really expanding their presence globally, especially in South Asia, Southeast Asia, as well as in Europe. On the battery side, so we work with tiers that use batteries sourced in China, so it's less about where the battery comes from, it's more about where the tier is and what they're trying to do. So we are working with all the tiers, whether they're in China, whether they're in Europe or in North America, to bring their solutions to market.
Perfect. Moving to financial questions, maybe for you, Derek. Foundry capacity utilization rates have been on the decline through most of calendar 2023. Do you feel like you've been able to negotiate better wafer pricing as we head into 2024 from your foundry suppliers? What do you expect the mix to be between your three foundry partners, TSMC, UMC, and Polar this year?
... Sure. So I'll, I'll start with the mix. The mix is, the last 12 months has been approximately 60% UMC, 30% Polar, and 10% TSMC. Kinda TSMC got as high as 15% about a year and a half ago. And so we make decisions based on foundry capacity, based on technology, quality, cost, and quite frankly, now geopolitical. And so when I think about cost, 200 millimeter wafers, everything we do is on 200 millimeter wafers. That capacity hasn't really freed up as material as it has at the 300 millimeter. So we haven't seen a price decline in 200 millimeter wafers. What we have seen, both at the wafer level and at the commodity level, is an abatement of the inflation levels that we saw a year or two ago.
So some of our suppliers, particularly the smallest one on the wafer side, had price increases that were in the teens and above that. Those have certainly come down, and we're in the process of negotiating the calendar 2024 contracts right now with those wafer suppliers.
Okay. But certainly seeing an abatement in the rate of inflation on some of those input costs, it sounds like?
Absolutely. And, you know, we've also seen a moderation between pricing between the Asian suppliers and our supplier, Polar, who's done a really good job of keeping their pricing and their cost structure pretty consistent over the last couple of years.
The mention of Polar brings me to my next question. About a year ago, the company secured a $150 million investment, led by One Equity Partners to expand their Minnesota fab. Can you give us any update on the status of this expansion? Are there any benefits that come to you from an expanded Polar with that investment?
Yeah, I'll start with the benefits that come with it. So with the expanded Polar, they obviously become a much more scaled fab. They're a pretty small fab right now. With scale comes cost benefits for its customers, which Allegro is 50% of their output. So we would get cost benefits there. We'll get technology benefits from upgrading the technology within Polar to keep track with our technology. In terms of the transaction itself, you know, Sanken owns 70% of Polar, and they're working with One Equity Partners on a combination of CHIPS Act funding, which they've announced publicly, in that transaction, which is constantly being... You know, we'll make an update on that when that transaction comes to close.
Okay. And do you still own the other 30%, or?
We do. We own 30% of Polar, yes.
Got it. Okay. Perfect. And I imagine, you know, with, with the move, well, you know, I think probably around the world, to having more local supply chains, having a bigger Polar gives you, you know, flexibility to the extent you need or want to bring more manufacturing back to the U.S.
Yeah, it's a real asset, Quinn. It, it really is a big asset, and it features prominently in our discussions with our North American customers.
This one, hopefully not too close to earnings, but I'll ask. Your December quarter guidance, you know, included the lingering effects from the UAW strike. Did those, you know, effects largely subside by the end of December, or do you think that those effects, in terms of production, schedule adjustments, and whatnot, could linger into the March quarter?
We'll defer that to earnings, Quinn.
Okay.
Thank you.
Yeah. The next question for Derek, just on the gross margin. You've got about 200 basis point, you know, hit to gross margins in the December quarter, about 100 basis points from the mix shift, more towards direct OEMs, away from disties, and another 100 basis points from the integration of Crocus. When do you expect the sales mix to normalize? Is that something that happens over a couple of quarters? And same question, how long do you think Crocus might be a drag on corporate gross margin?
Yeah. So the sales mix, which would be the normalization to more distribution, I think is a couple of quarters. We've talked about it takes a few quarters for the inventory and the industrial and the other to work itself out. So that's a few quarter process. We talked about that at our Q2 call. So that'll be a tailwind to gross margin when that comes back to normalized levels. And on Crocus, what we talked about is, we'll expect to start to see synergies, cost synergies in Crocus in the first half of calendar 2024. Those will primarily be operating expense synergies. The cost of goods sold synergies by moving some of their probe and test to our facility in the Philippines is a little bit longer lead time towards the back half of calendar 2024, and of course, wafers take a little bit longer to get qualified.
Got it. I realized, we are out of time. I sort of lost track. So I'm gonna end it here. Vineet, Derek, thank you very much for joining us at the Needham Growth Conference. We really appreciate your participation.
Thanks for having us, Quinn. Good afternoon.
Thanks.
Thanks, everybody.