Intel Corporation (INTC) Earnings Call Transcript & Summary

Well, good morning and good afternoon, everyone. This is C.J. Muse with Evercore ISI. And it's my pleasure to welcome Intel, and particularly, Erez Dagan, VP of Products and Strategy at Mobileye; and Tony Balow, Senior Director in Investor Relations at Intel. Intel and Mobileye last offered a detailed look at its strategy for ADAS autonomous vehicles at its Analyst Day in Jerusalem back in November 2019. We know they've made great progress since then. So we're very eager to hear from the team. Erez will make a few opening comments. And then my colleague, Chris McNally, and I will host a fireside chat. If you have any questions, there's a box. You can send them to us, and we'll be sure to ask them on your behalf. And with that, let me turn it over to Erez.

All right. Good morning. Erez Dagan, EVP, Products and Strategy with Mobileye. Excited to chat with you here today. Thanks a lot for the invitation. I suggest we go to the Q&A. Thanks.

Okay. Excellent. So I guess I've been tasked with starting at a very high level. Last November, you outlined a strategy with a clear focus on L2, L2++ as well as Robotaxis as well as longer-term focus on the passenger car autonomy side. So would love to just hear what has changed, what have you been able to accomplish in your overall strategy in the last kind of 10-plus months since we've heard from you.

Perfect. So the strategy is here in place. We still see this -- first of all, in the driver assistance space, we see the separation between Level 0 to Level 2 as one segment. That's driven by certain factors. There's Level 2+ segment. And there is the mobility-as-a-service or the Robotaxi deployment arena for autonomy. Robotaxi is the deployment arena of autonomy. Followed by introduction of autonomy into the consumer market. So in the driver assistance, we see lower end, we see the Level 2+ segment. In the autonomy, we see the Robotaxi as the primary introduction market and the consumer autonomy to be followed. Within the mobility-as-a-service segment, the introduction of Robotaxi, autonomy for Robotaxi. We've also, through the past 10 months, as you probably well know, acquired Moovit, acquired a company to complete the value chain -- the value nodes that were required in order to launch a stand-alone mobility-as-a-service. And we also shared that we will have a vehicle-as-a-service and ride-as-a-service as 2 derivative models. These 2 derivative models would allow us to interact and collaborate with public transit authorities on one hand through the vehicle-as-a-service model. And ride-as-a-service, and mostly designated to TNCs, which don't know how to operate the capital required for automated mobility solutions and can simply purchase rides under their own demand umbrella. This is kind of the high level of where we stand on strategic level.

That's great. I guess as a follow-up and really to help kind of think strategically about how you're attacking autonomous vehicles versus others, can you quickly describe your approach as it relates to using vision first and full true modality redundancy versus more traditional LiDAR-based simultaneous location and mapping, or perhaps unique driving policy framework like RSS versus end-to-end driving policy?

So it also has to do with the previous question in the sense that we -- in our Robotaxi approach, we separate the computer vision subsystem to be a stand-alone subsystem that operates in what you refer to as true redundancy. There's another subsystem of radars and LiDARs, which also produces a comprehensive and coherent stand-alone model of the environment. Using these 2 models of the environment, we can reach what you also referred to as true redundancy, which means that for any decision that the vehicle is going to take, we would expect it against 2 views of the environment, one produced by our computer vision system and the other one produced by the subsystem of radars and LiDARs. Now the linkage to the previous question is interesting in a sense that the same computer vision subsystem that propels our Robotaxi or autonomous vehicle solution is the surround vision solution that we are taking to market as part of Level 2+ segment. And this fact alone is based on a strategic level [ first in incubation ] that we are -- or technologies that were incubated and trickle them down into our driver assistance go-to-market. And the mapping is one other such example, where we started with designing a crowdsourced mapping solution for our Robotaxi, and we trickle it down all the way to driver assistance. Maybe the last differentiator that you mentioned is also a core differentiator here where our safety model for the vehicle, the safety concept for the vehicle is a model-based -- a model that helps us describe very precisely what is the barrier between assertive driving and dangerous driving. What sort of actions that we take may cause risks or inflict risks on other road users, and what behavior would simply be more assertive and allow us to negotiate into traffic. And that barrier line, that model that helps us clarify that barrier line, trickles all the way down to our perception what should be protected more and what should be protected less, what is safety critical and what is simply dealing with the comfort and convenience of the drive. This distinction is very strong and allows us to allocate resources into the problem in an intelligent way.

Well, that's great. I think Chris has a few questions on the ADAS Robotaxi side. So let me turn it over to you, Chris.

Thanks, C.J. Erez, and thanks so much for your participation. Look, ADAS is still your bread and butter, right, so -- and an area where you continue to dominate market share. And I think you've previously disclosed, I think, a $5 billion TAM by 2024. And last year, revenues were a little under $900 million. Can you talk about the pace between now and '24, '25? Can we see that revenue double or maybe something greater than that? Simply because we see this big opportunity from Level 2+ or Level 2++. And anything you can add about the idea of unit penetration, just chips, but also ASP increases as we move up the stack.

Okay. So there are 2 -- thanks for the question, it's a very good one. There are 2 modes through which the revenue pile is about to increase in the driver assistance domain. The first mode is much increased fitment rates of lower-end solutions, okay? We see that trend already. We already talked about it several times as well. The regulations such as Euro NCAP, GSR and many other mandates across the globes for trucks, et cetera, et cetera, are driving fitment rates faster than was previously anticipated by the market. And that's a substantial contribution to that trend that you were mentioning. The second one is the second mode of more advanced features, driver assistance features, Level 2+ mainly. And in that arena, we do see a very rapid growth. I think that has a lot to do with the consumer awareness of automated driving -- autonomous driving technologies that also leaks into awareness to advanced driver assistance. So we definitely see that on track as you denoted.

And then maybe if we just drill down into this idea of Level 2+, right, which is really an evolving -- a couple of years ago, we were all thinking about Level 3, and now it seems like Level 2+ is something that's going to make the OEMs a lot more money. How big could this be? I mean it doesn't need to be specific. But could this end up being 20% of ADAS penetration in 2024, 2025, 10 million plus units? Or is it still going to take a little bit longer for it to evolve?

We predict also based on research that's out there that we see north of 10 million units, ballpark of 12 million or 13 million units, around that. And yes, this segment of the market is really rushing very impressively forward. We can see that through our -- the market RFIs and RFQs. It's a well-established segment.

Okay. And Erez, do you think the current $5,000, $6,000, $7,000 for these sort of highway assist programs that are being charged to the customer, do you think that's the right price? Obviously, it's not something that Mobileye kind of has control over. But do you see that coming down over time? Because obviously, that's one of the inhibiting factors for Level 2+ growth. It's clear that consumers want that.

So I must admit, I am not sure at what level these systems are priced to the end consumer and whether the figures you denoted are a Level 2+ or a Level 3 price. I'm not tracking the price list, full disclosure here. But I can tell you about what we already shared about the way that we are seeing it evolving. Full -- not full, Level 4 autonomy, not Level 5, but Level 4 autonomy is going to be introduced first through Robotaxis. And then a couple of years later or 3 years later, we expect the costs of these solutions to drop to several thousands of dollars costs. And that's full autonomy. That's something that could drive in an urban scenario, doorway to doorway kind of a personal chauffeur. The other data point that we also shared in the past is that the subsystem of surround vision system is -- the affordable way to introduce Level 2+ is vision-centric, okay? Cameras are giving you a very comprehensive model of the environment, combined with a map. And these systems also, in nature, do not include expensive sensors. And if you take into account the compute and the sensors, we're talking about hundreds of dollars of costs for these systems. So again, pricing strategies and penetration strategies of OEMs, not for me to say, but I do expect that to have a -- if these are the prices currently charged, I expect it to go down as we go forward. This is -- the economies of these solutions is being tightened every day.

Great. And then one last one on ADAS, and then we just want to talk about Robotaxis in Jerusalem. Just on your recent ADAS, I mean, you essentially have such a large share. In Q1, we noticed that you added a very major Asian OEM. Would you be able to comment if this is sort of the one large Japanese OEM that you haven't worked with in the past, maybe up until now?

Unfortunately, I cannot comment at this point because of the dark period is it, the quiet period. And I would say that we see a lot of thrust in the market. As I said, it's in the -- maybe the best way to look at it is that 12 out of 15 top makers are using mobilized solutions. So we're there. We're going to continue our thrust forward in the market. And hopefully, we can share interesting news very soon.

Great. Just real quick before I send it back to C.J. So look, your Robotaxi deployment is going to happen in Jerusalem as the initial beta commercial rollout, I think, in 2022. We've become a little bit numb to the idea of these beta rollouts. Could you just talk about what can we expect for an initial deployment in Jerusalem for a commercial service? How robust will it be?

All right. So maybe it's a good opportunity to also share about our -- what we've been doing in that space for the last 10 months or a year or so. As I said, most notably, the acquisition of Moovit and joining Moovit to our force to reach that solution of full end-to-end service. We already have our Robotaxi development fleets hooked on the Moovit service application. Again, not publicly available, but as a development step and we can close the link for a full service on the technical level. On the maturity of these solutions and time to get to market, we are on track with plans as stated in the past. No setback whatsoever. We are -- in terms of what you're going to experience, if I get back to your question, simply calling out for a vehicle out of a multi-model -- application of multi-model options of transport, you can choose an autonomous vehicle as one of them. Hop on the car, get a notice when you get off the car and get a notice when the transaction is complete. Within those car, there is access to a call center in the back. There is a teleoperation center that has also the way to look into the cabin and into the surrounding of the vehicle in case anything is required to that purpose. Yes, I'm excited to share its on track, and that we're looking forward to launching it as well.

That's fantastic. C.J., why don't I send it back to you. I could talk about AV all day, but why don't you get focus back on the semi side?

Sure, sure. So yes, I really wanted to, I guess, focus a bit more on your partnership now as part of Intel. And so as you think about creating a full self-driving system and providing a full solution, can you kind of speak to both hardware and software and what Intel brings to the table to Mobileye that really is accretive to the work that you guys are doing?

Sure. So one important almost immediate value that we were able to get as part of the acquisition is a very strong team writing software and compilers that helped us open up the EyeQ5 to be a compute platform. That was a very impressive and very quick contribution that we got there. There is a lot of contributions in the area of -- at the system level, whenever it comes out from the SoC system level, we are gaining a lot of experienced developers and development workforces in Intel. Of course, on the strategic steering, we are also consulting our mother ship in that sense. And I think -- and capital as we already also executed on the Moovit deal to help us secure the path forward.

That's great. I guess digging a little bit deeper based on the Analyst Day, I guess, 10 months ago. It sounded like Intel working on other parts of the portfolio, LiDAR sensors, imaging radar, silicon photonics. So I guess, as you think about all the work being done there, maybe you could kind of walk through what's most exciting and how do you see in-sourcing versus outsourcing, looking out to -- from the current ADAS part of the world to true Level 5.