TD Cowen 54th Annual Technology, Media & Telecom Conference

Okay, good morning. Welcome to day two of our 54th annual TMT conference. I'm Josh Buckhalter, semiconductor analyst here at TD Cowan. I'm pleased to be joined on stage by my colleague Lanny True and Dan Galvin from Mobileye. Gentlemen, thank you for joining us.

Thank you for having us.

Maybe just to start the conversation, we just got out of earnings season. There were some inventory dynamics, but also it seems like the auto and specifically your backdrop is improving. Can you sort of walk through what you guys are seeing near term and into the second

half as well? Yeah, maybe I can. I'll take this one. This is Dan. So I think the business performed really well in Q1. We grew revenue 27%. Like you said, that there was some safety stock normalization because inventories had ended the 2025 really low. But even if you exclude that, we grew 15 or 16% in an environment where the global auto industry was down about 3% so you know like we have the last number of quarters we continue to outperform the industry in terms of our top-line growth we decided to kind of you know take advantage of our good balance sheet and good cash flow we announced a buyback that's intended to kind of offset the dilution from stock based comp and the the mentee acquisition so I think that was a positive we talked about some execution proof points around supervision which is kind of upcoming launch we did a 2,000 kilometer drive on an unplanned route in the US for the first time with kind of a production level system so yeah I think I think we had a lot of positives to talk about in the first quarter we see kind of continuation of pretty strong order flow you know our outlook for Q2 was higher than street expectations. And then, you know, for the back half, like there's still a lot of uncertainty out there. So we sort of just like left the forecast or left the outlook at what I would call probably the lower end of the range that we've been seeing the last like five or six quarters.

Okay, maybe one more on this topic and then I'll pass to Lanny to ask some questions about China. You know, as we think about that back half outlook, you know, how much of that is indeed embedded conservatism? Because I think, you know, on the call, you guys described, upward pressure to IQ demand, and really everything seemed to be tracking well. So I guess how much of that is just, given what was going on geopolitically, you felt like it was prudent to keep things in check for the back half versus things you're actually seeing in sort of

the timing dynamic with the inventory restocking? Yeah, I mean, I think sentiment was, at the time, was basically if companies can hold their guidance for the year, that's a good thing, right? And I think we've been we've been kind of in this like 9.0 to 9.5 million units per quarter range in q1 and q2 it looks like we're seeing some upside from you know chinese oem export activities some upside from 8s adoption from our core customers and you know that you know we're hoping that that continues but you know if there is you know production schedule reductions which we haven't seen yet at all. We'll have some cushion to incorporate that for the back half.

Thank you. Yeah, on that topic, you know, I think a lot of investors are keyed in on weakness within the Chinese domestic market, but China exports seem to be holding up pretty well, and that is a benefit for Mobileye. How do you think about the durability of Chinese exports, especially as companies have talked about, you know, demand increasing and looking pretty good over the course of each month this year?

So I can take this. So I think we have been pretty consistent with a couple of Chinese customers in the past few years. What has changed is that we have launched a few new products that replaced non-mobile solutions, and that they picked up their export volumes, especially to emerging markets. I think that's the change, that they managed to increase their sales in foreign markets that are not Europe, specifically. And just for reference, two of our bigger customers, Cherry and Gilly, that have significant export volumes, they have increased their volumes in the export markets in more than 100% in the first quarter of this year. So that created a strong tailwind for us, and we think that it can continue. There is plenty of room for them to grow in the markets that they are now penetrating, which is mostly in Asia Pacific, in Eastern Europe, South America. They are now working on India. It's still early to see if they can manage to penetrate India. But I think what's important to say is that it's not cannibalizing our volumes in Europe or in the U.S. or in Japan at this stage, like the bigger markets for us in sales. It's more of increasing the pie for us because it's new cars in markets that today we're not selling units to. So I think overall, the fact that our solutions are proven, cost efficient, they can pass regulations in the future in Europe if they want, when they will start selling more in Europe, which they plan to, it gives us a good level of comfort in our relationship with them.

Yeah, you also mentioned that India was a new theater of growth for you, and you have the recent deal with Mahindra. Can you talk to us a little bit about the deal structure and products that you're selling with Mahindra?

Yeah, zooming out on the Indian market before we talk about Mahindra, it's a very interesting market, although for 10 years people are saying India is the next China in 10 years. It's been 10 years of saying this, so hopefully it will happen sometime soon. But in the auto industry specifically, it's the third largest market in terms of number of cars sold per year. Historically, it's been a market for low-cost cars, roughly $10,000, $12,000 per car. And in recent years, there is a strong demand there for higher-priced vehicles, more advanced features. And Mahindra, that's their claim for fame. They sell SUVs in $30,000, and they've been growing in double-digit percent year-over-year for multiple years now. I think they're expected to surpass 800,000 units this year and reach a million units in 2028. So what's interesting for Mobileye in India is that it's less than 10% ADAS adoption rates today. In 2027, there is a new regulation coming up called Bharatan Cup, and they will mandate emergency braking and potential additional features for new cars sold in India. And that organic growth, going from 10% ADAS adoption rate to maybe 60, 70, 80%, it can be additional millions of units per year that have ADAS, that today does not have ADAS. So for us, we have a very good relationship with Mahindra, but also with Tata, and some non-Indian OEMs that sell in the Indian market, like Hyundai, IKEA, like Nissan, Stellantis, and others. So I think we're well positioned to benefit from this. And one last sentence about the Mahindra collaboration. So what Mahindra wants to do is not just to offer ADAS in India, but to take it to the next level and offer advanced ADAS in India. And our deal with them includes supervision and surround ADAS, so two products. They want to have surround ADAS in roughly 50%, 60% of their cars starting in 2028, and supervision will be around 10%, 15% of their cars. So it will position them as not just a luxury company, but innovative that brings new technologies to the Indian market. And for us, it's a very nice growth opportunity.

Yeah, I wanted to go back to China for a moment. That's been a tricky market for you and your peers. On one hand, they clearly push the envelope from a technology perspective forward, but there's also the pull for localization. They were sort of where you led with supervision as well. Can you talk about how you view competing in that market differently than the West? And what are your, I guess, the long-term view of how much growth you can get out of the region?

So I'll divide it into domestic China markets and export markets. And you can also divide it between Chinese OEMs and non-Chinese OEMs. So there are multiple dimensions here. With the Chinese OEMs, I think, as we said, We have strong positioning for the export markets. We also have some moderate volumes in the domestic market as well, because it makes sense for them to have the same solution for the same car, whether it's sold in China or outside of China. For the non-Chinese OEMs, we think are more dependent on their ability to sell cars in China, which is hard to predict. There were some years in which there was a negative trend. Last year, it changed. VW, for example, managed to start growing again in volumes in China. So it's hard to predict, because China specifically is very influenced by geopolitics. And there could be, due to trade considerations, maybe they'll open up more for some companies. That's a little bit of what's happening now. Maybe they want to open the door more for European to improve their relationships with the European market, because they want their cars to be sold in Europe as well. So it's hard to predict where things will head. I think for us, our focus for advanced technologies is in Europe, US, Japan, Korea, not in China. But we do want to be participating in the Chinese market because it's a very good proof point for our technology.

I want to shift gears and ask about competition a bit. I mean, Mobileye has been the incumbent in ADAS for pretty much as long as ADAS has been something we talked about, but it's obviously attracted a lot more competition as the market has gotten bigger. Could you speak to Mobileye's differentiation at the IQ hardware level and also the software stack and basically what's mobilized right to win on base ADAS, but also the advanced ADAS programs as well.

Yeah, I think we should maybe categorize base ADAS and surround ADAS as a part of base ADAS. This is like the high volume market, then there is the high end market for privately owned vehicles, and robotaxi markets. Kind of three distinct markets, each has its own competitive landscape. For the high volume market, I think that we pretty much managed to maintain our incumbent position and dominant position. We won more than 98% of the volumes that our customers have published for these high-volume cars in the last three years. So we not just managed to maintain our position with our existing customers, we also acquired two new customers, Volvo and Subaru, both never had Mobileye products, or at least not in the last 10 years, and they now decided to start working with us again. So I think in this high-volume segment, in the entry part of things, where cost efficiency is extremely important, but also performance is very important because of regulation and the cost of safety issues, we have around more than 250 million cars on the road. We have never had a recall. Touch wood. This is not wood, but touch something. So I think it's a very, very strong proof point. And the cost of replacing us for the OEMs and the risk, just, you know, it's a hard math for them to do. We do have some competition. Bosch is trying to, you know, be more aggressive there, some small startups, but I don't think that's where the severe competitive pressure that you're referring to comes from. On the robotaxi side of things, I think, obviously, we have the front runner, which is Waymo, and Tesla that is kind of contending, and then there is a herd of companies that are claiming to be a part of the Robotaxi landscape in the future, but none of them have proof points yet. Maybe they'll succeed, maybe they won't, maybe in five years, two years, 10 years, we'll have to see. I think we have, in the Robotaxi side of things, our competitive advantage comes from the fact that we own the hardware and the software, that we are working with VW, who's a very scalable OEM, and we have been working on this for multiple years. We accumulated all the data and all the experience, we're getting close to launching with a scalable partner with a very cost efficient system as opposed to some of the other competitors that use hundreds of thousands or tens of thousands of dollars per system so when we'll come to the market we come from a cost advantage standpoint and with a scale volume scale advantage and then it leaves the this high-end sliver in which there is a lot of competitive noise and I think that you know obviously there are many companies there, but what's important to realize is that the OEMs, especially the big OEMs, are having different interpretations to this segment, and they're trying to understand what makes most sense for them. This is not yet an area they need to have. In some cases they want it, in some cases they're not even sure if they want it. As opposed to the base ADAS, which is something they need, this is something that they're contemplating, if they need it or not. And this gives way to multiple different interpretations. Some companies want to work with that company, that company. They want to own the software. They don't want to own the software. They want to design the architecture. So it's very confused right now. But the volumes there are still very small relative to the high volume segment. So I think it will take a couple of years for that to also kind of, for the dust to settle in this segment. And I think when we launch our products with VW, with Porsche and Audi, it will help us to kind of separate ourselves from the pack in the high-end segment.

And I want us to make a couple more comments. So I think on what we'd call the low-end, high-volume segment, I think the point there is that for what the OEMs need to have in their cars, we continue to win at an extremely high rate. I think this China export, the ability to sustain, you know, fairly meaningful business with Chinese OEMs, the surround ADAS opportunity for kind of next-gen ADAS at higher prices, it gives us the ability to grow that ADAS business sustainably. I think that's a different kind of perspective of what investors had maybe a year ago on the business. And then on the Robotexy side, I would say like we do, there are front runners out there, but no one's running away, right? Waymo has added 100 cars per month over the last six or seven months. That is not really fast scaling. Tesla's still at 20 cars operating. So I think that there's time to see this kind of herd come through and see kind of who can succeed, who can execute, and who can't.

I think something investors and us too have struggled with is benchmarking, right? It's like I don't know what the right way to judge the performance of a mobile solution is. You regularly, I think, try to inform the investment community. If you're looking at IQ chips, for instance, don't just look at tops. That's not what your customers care about. Could you maybe provide a little insight on how you win business and what your customers want from you? How much of it is the software layer and the experience you've built up over the last 15 years versus raw chip performance?

Yeah, so I think, again, let's separate between what they need and what they want. The high volume part of things, because put yourself in the shoes of a decision maker at the OEMs, making a decision that affects millions of cars is a very, very risky one. If you have a delay in the project, for example, the consequences can be, you know, tens of millions, hundreds of millions. If you have safety issues, if you have quality issues and you have recalls, it's a huge impact. as opposed to a small project that has high-end technology that is starting with one car that will sell maybe 10,000 units over its lifetime. And if it takes longer, if it's too expensive, if it has some quality issues, then you can manage it. It doesn't risk the entire company. And so I think for what they need, it's primarily the performance, the functional performance versus the cost. It's not just performance and it's not just cost. They cannot really optimize just for cost because then the risk of performance and recalls can bite them. And performance is not sufficient, because if it's too expensive and you apply it on millions of cars, then you're eroding your profits. So I think it's really important to be the most optimal solution in this graph between performance and cost. And that's been mobilized, I think, competitive advantage since its inception. That's why we're doing our own chips. It's to control the cost and to drive them lower as much as possible. So I think in this high volume, what they need regulation-driven, where safety and performance is extremely critical, but also the price is extremely critical, we have a well-recognized competitive advantage, which is evident by the numbers that Dan said about the win rate. In the high end, again, it's more about, I think, a little bit more confused there. It's not clear what are the benchmarks. What is the benchmark for a level 2++ system? What's the right performance bar to put for an FSD-like system as long as it's supervised? Is it like 50 hours of mean time between failures, 500, five hours, as long as the driver is responsible? It's not objectively defined yet in the industry, not just by mobile eye, which is, I think, why there is room for more actors to come and kind of claim different claims, because there is no clear bar of what is good enough or what's the minimum you need to meet and what's the maximum and where is each company stacked according to this graph. On the robotaxi side of things, we have Waymo's reference, which kind of laid out how good you need to be in order to go driverless in cities. So I think in that way, it helped us because it created a cleaner environment in terms of the standards. So now we know that if we get to the standard, which we're tracking, we can launch driverless in the U.S., which is easier than it used to be.

I'm going to ask one more about the advanced products and then maybe turn it to Lanny to ask about robotaxis. you know you guys have talked about a growing funnel of advanced ADAS programs I'm actually going to include I know you didn't but I'm going to include surround ADAS in that because it's materially ASP expansive can you talk about the appetite from your OEMs particularly in the west for advanced ADAS solutions there had I think there was a long and slow road for adoption because partly because of I think EV roadmap issues you know where's the appetite for advanced ADAS now? Is Tesla catalyzing the industry? And, you know, maybe you could speak to your funnel

for surround ADAS through Chauffeur. Yeah. So I think surround ADAS funnel is very healthy and it's more in the category of things they need than things they want in the sense that its regulation is going to continue to push forward in new features that you need to, or new use cases that you need to account for in the next two years, three years, which means that the OEM needs to expand their ADAS capabilities in their base cars. So they need more sensors, they need a little bit more sophisticated software, and this gives a very good tailwind for surround data solutions, but still it needs to be very cost-efficient because it's going to be applied for millions of cars. Just to give some reference, our first two customers for surround ADAS accumulatively have 20 million units sourced. So it's not a small science project. This is standard fit across multiple models. With VW as an example, this goes into Volkswagen Golf and their best-selling cars. So it's a huge decision on their part. So I think in that category, I would say that the appetite is very healthy because it's supported by regulation, but also it's a cost-efficient solution, so the decision is easier. it's not like adding two thousand dollars to your car and which you know you need to prove it to justify this decision on the high end i think it's a little bit again as i said a little bit more confused because maybe some oems want to start with level three maybe they want to do level two plus plus i would if i would do a random poll across the oems i don't think that there are like one or two solutions that everyone is agreeing will be the next generation and what is the the optimal user experience. All of these are still in play, and I think the industry is kind of looking in Tesla FSD as a science, as a kind of experiment. And what can we learn from this is, on the one hand, it's an amazing product technologically, really. I'm using it every time I'm in the US. But it didn't lead Tesla to sell more cars, for example. Maybe they have increased take rate for this, maybe. But this is Tesla customers. Is this indicative for OEMs that have a different customer profile will they pay extra like the Tesla customers these are this is all in their under evaluation I would say which is why things take longer thanks

for that just to touch on your robotaxi programs before we I do want to give Josh an opportunity to talk about mentee robotics as well for your robotaxi partners like VW and Moya with uber and beep can you talk about your current path deployment and you've mentioned that you're on the path to removing the safety driver as well what are the kind of hurdles remaining for you to remove the safety driver in your cities of operation yeah so I think VW made a

couple of announcements earlier this year that are very important the first one is the operational manufacturing line in Hanover which means that there is a production line in Hanover that produces robot taxes in scale like a normal car this is a major major accomplishment for them they invested a lot of resources it means that now there is a serious production line that can produce hundreds and then thousands and tens of thousands of robotaxis at the same quality as opposed to retrofitting robotaxis thinking about costs a major cost driver is manufacturing and now we have a very scalable manufacturing we think with since then we've transitioned our testing fleets and early deployment fleet to use these cars that come out of the manufacturing line what we're doing now is basically use these cars and we net we need to do a few months of validation to kind of dot the I's and cross the T's that the cars that come out of the manufacturing line are indeed at the sufficient performance and stability quality that there is no hiccups in the series production process that's exactly what we're doing now and when we'll finish this we can start the sequence of removing like safety driver first but commercial like customers can use these cars and then no safety driver so that's all planned for the backup of

this year okay and on the topic of cost you know I think a lot of investors are trying to figure out the cost comparison between mobilized solution and say like deploying a Waymo or the cyber cab you talk about your cost advantage and then what kind of how should investors think about your ASPs from both a hardware perspective and a recurring revenue perspective as these robotaxis get

deployed yeah so I think when it comes to the system cost I'm putting aside the cyber cab because it's obviously a very cost-efficient solution it's just remains to be seen when and how it will be scaled to robotaxis referring to Waymo the current system that we have versus the current system that Waymo have is around what between 1 to 5 to 1 to 10 cost difference on the bill of materials of the system. We use off-the-shelf sensors, excluding our imaging radar that we develop, but this is a very, very cost-efficient radar. But the LiDARs are off-the-shelf LiDARs, the cameras are off-the-shelf cameras. Our compute is very cost-efficient. It costs small thousands of dollars as opposed to what others are using. So I think in the bill of material, there is a significant difference in potentially tens of thousands of dollars of difference. Of course, I cannot talk about their future generation. There's just different speculations, but it's hard to have an accurate assessment. I'm sure that they know this, and they're working very hard for multiple years to improve this. But I think we come to the market with a very cost-efficient system from day one. That has been our strategy. We knew that we won't be the first mover, but in order to carve out market share, having a cost advantage allows us to maybe come with lower price points to consumers and then create traction to our systems as opposed to the way most the volume scaling like like you talked

about earlier is really important to having just like one facility producing these cars at the same quality without having to move them around you know I was at a dinner last night where a company was talking about like a three facility manufacturing process for their vehicles like it makes a big difference that you're kind of getting these products you know right

off a normal assembly line I do want to say in one minute to take about the Robotex is I would say competitive dynamics because I think it's very important and there is a lot of confusion the there are a few driving forces but I would say that the ride-hailing companies are not necessarily counting on Waymo as a partner to be mild about it I think that they consider Waymo and Tesla's potential disruptors in the future which is why you see them especially uber placing such you know high amount of bets on different companies however what they're really looking for is access to volumes of robotaxis they really want to be able to buy ten thousand fifty thousand robotaxis and start deploying them so it's one thing when they invest in a startup and have equity in a startup to you know increase the chances of that startup to survive and maybe be an alternative in the future, it's a different thing when they are looking for an actual solution that can be deployed in one or two years. In that sense, probably the only way for them to get these cars is through BW and Mobiline. That's the only, I would say, collaboration that has plans to produce cars in meaningful scale. So I think we should just separate between, you know, investments and bets that they're doing for long term having alternatives and having negotiation leverage because they want to have 10 different companies that do autonomous driving it will help them in the future so they just you know invest in them it's a different thing when they make strategic decisions on when can we who can we partner with for 2027 28 29 deployments to have an alternative to Waymo and to potentially Tesla if they succeed this is this is the announcement on Los Angeles for example with uber which is the first market there is a like with lyft another one and there is a few more cities that probably vw will announce soon so just to see that and you know there's obviously a lot

of excitement for good reason on robo taxis um the asps are much higher but the units are much smaller i mean any timelines you can give us on when we should expect some level of materiality from that vertical for you and then i think you had talked about sort of a 50 000 asp for mobile drive at the analyst a couple years ago is that still the right number and can you walk through the split of hardware versus software in that and how the yeah

it is monetized so I would say that material impact on revenue will not be before 2028 realistically although I think that the value of creation itself beyond the financial value of having a driverless product that can be integrated in cars will be evident much before.

It depends on what you're saying is material. If material is like 3%, 4%, bump to revenue, which I think it is, then maybe back half of 27% has the potential to do that.

I was referring to like 10% or something in that area. And then on the ASP, the 50,000, it's actually a little bit higher than that, and the way it's structured is roughly a third of it is kind of an upfront payment per car, and then the rest is being paid per quarter of operations for each car. So that's kind of our, depends on if the car is actually operational, so.

Mileage based?

It is like mileage based, but the accounting of it is a little bit different, because we don't want to, we want to create some stability, I would say. So it's like a quarterly assessment of how many miles were driven, and then there is a price for that, then we get this on a quarterly

basis. Okay, thank you for that. We're quickly running out of time, but I did want to ask about the Menti Robotics acquisition. There's perhaps no better example of physical AI than autonomous driving and ADAS, and you guys made a pivot into the broader spectrum of physical AI and robotics with this Menti deal. Can you walk through the strategic rationale? What IP are you guys bringing? What are they bringing? And maybe speak to the go-to-market as well?

Yeah, so we have a minute. I wouldn't call it a pivot. It's like an expansion of what we do. I think there are many commonalities in the challenges between robotics and robotaxis. And at the end of the day, it's a robot that needs to operate in the real world, engage with humans, be safe, scalable, and functional. So far, the idea was to create a more comprehensive frontier in physical AI in a way that advancements in our AI technology can help us improve our chances to win in robot taxes and in robots. It's a new go-to market. It's not an automotive, necessarily. It's a more, I would say, earlier stage of adoption. The synergies are very, very, like, across the board when it comes to the simulation technologies and, you know, the AI architecture, infrastructure for training even hardware design many many things that mobile is excelling at mentee definitely needs and they're benefiting from this only after a few months of integration and the opposite is also true it helps the mobile I team with advanced simulation technologies and data creation technologies mentee develop and in terms of go to we're running out of time so that's it but the go-to-market is more about it's a little bit too early I think to sit to talk about this it's between manufacturing lines and logistics and maybe home use is the first market this is like the junction we're at right now okay all

right well unfortunately we're out of time Dan Nimrod really appreciate you joining us again and very much good luck thank you thank you thank you very much