Ambarella, Inc. (AMBA) Q2 2023 Earnings Report, Transcript and Summary

Good day, and thank you for standing by. Welcome to Ambarella’s Third [Second] (sic) Quarter 2023 Earnings Conference Call. At this time, all participants are in a listen-only mode. After the speakers’ presentation, there will be a question-and-answer session. [Operator Instructions] I would now like to hand the conference over to your speaker for today, Louis Gerhardy. You may begin.

Thank you, Tewanda, and good afternoon. Thank you for joining our second quarter fiscal year 2023 financial results conference call. On the call with me today is Dr. Fermi Wang, President and CEO; and Brian White, CFO. The primary purpose of today’s call is to provide you with information regarding the results for our second quarter of fiscal year 2023. The discussion today and the responses to your questions will contain forward-looking statements regarding our projected financial results, financial prospects, market growth and demand for our solutions, among other things. These statements are subject to risks, uncertainties and assumptions. Should any of these risks or uncertainties materialize or should our assumptions prove to be incorrect, our actual results could differ materially from these forward-looking statements. And we are under no obligation to update these statements. These risks, uncertainties and assumptions, as well as other information on potential risk factors that could affect our financial results, are more fully described in the documents we file with the SEC, including the Annual Report on Form 10-K that we filed on April 1, 2022 for fiscal year 2022 ending January 31, 2022 and the Form 10-Q filed on June 8, 2022 for the first quarter of fiscal year 2023. Access to our second quarter, fiscal 2023 results press release, transcripts, historical results, SEC filings and a replay of today’s call can be found on Investor Relations portion of our website. Fermi will provide a business update for the quarter, Brian will review the financial results and then we will be available for your questions. With that, turn it over to Fermi.

Thank you, Louis, and good afternoon. Thank you for joining our call today. Q2 results were mostly as expected; revenue was down 10% sequentially and up 2% versus the year-ago quarter. CV revenue increased significantly on a year-over-year basis, driving our blended average selling price above $10. We have cumulatively shipped more than 10 million CV SoCs, including more than 20% into the automotive market, and we remain on track to achieve our prior estimate that CV will represent about 45% of our total revenue for the year. In Q2 we absorbed the worst from both, the China lockdown and the 14 nanometer shortage, and incomplete kits remained a bottleneck for many of our customers. Collectively we should see some improvement from these factors in the second half, but our recovery outlook is tempered as some customers are now reducing their inventory levels, in particular as lead times contract. In July we commenced a global roadshow to leading automotive Tier 1s and OEMs with live demonstrations of our CV3 SoC. As a reminder, CV3 is our first central domain controller integrating in a single monolithic die all of Ambarella’s core competencies; including perception processing from cameras, radars, sensor fusion and the path planning, as well as our functional safety and application software. Just one month after the SoC came out of the fab, we demonstrated full functionality across multiple live cameras to many global automotive OEMs and Tier 1s. The superior performance and efficiency, as well as the scalability of our SoC roadmap, were clearly recognized and well received, leading to many follow-on development discussions. Now, we expect to be able to share more about some of these customer engagements by the end of the year. We are also proud to announce that CV5, our first 5 nanometer SoC, entered production in Q2, at least a quarter ahead of expectations. We have three IoT customers that each purchased production volumes in Q2, and we are engaged with many other companies developing their first CV5 products. I will now provide some examples of our market development activity, where production has already, or is expected to commence this year. Toyota introduced its 16th generation 2023 Toyota Crown model featuring a two-channel digital video recorder based on Ambarella’s CV25AX AI processor. Capable of recording from both the ADAS system and the electronic mirrors, the car recorder is supplied by Japanese tier 1 Denso-Ten. BMW began shipping two dash camera models, the Advanced Car Eye 3.0 and 3.0 Pro, providing both front and rear view recording. Based on Ambarella’s H22 video SoCs, the dash cams will be sold in over 80 markets, with the Chinese version also including electronic toll charging features. In June, China-based FAW Hongqi introduced its latest generation of B-class cars. The H5 sedan includes an L2 ADAS system based on our CV22AQ automotive SoC and supplied by tier 1 Neusoft. Also in June, China-based Dongfeng introduced its Haoji SUV. The SUV includes an Occupant Monitoring System based on our CV28AQ and supplied by tier 1 BCS. And in August, Israel-based Cipia announced Chery’s SUVs will integrate Cipia’s Driver Sense DMS running on Ambarella’s CV28 AI processor. The integration of the CV28 and driver sense software utilizes neural networks to offer robust driver monitoring capabilities. I will now talk about some of the new IoT product announcements. Johnson Controls introduced its 4th generation Illustra Flex camera series, including mini-dome, PTZ and bullet models, based on our CV28S AI SoCs. Also during the quarter Ubiquiti announced its AI Bullet camera. Based on our CV25S AI SoC the 4 megapixel camera includes smart detection of people and vehicles. Korean IP-camera maker IDIS, launched four new 8 megapixel and 6 megapixel PTZ cameras based on our CV28 and CV22 SoCs. The cameras are designed for wide-area surveillance operations with the capability to identify objects and recognize faces at distances of up to 300 meters. Japan-based iPro, introduced a new 2-megapixel machine vision camera aimed at manufacturing sites and based on our CV22 AI SoC. The camera performs automatic visual inspections and supports up to 100 different object types, for example detecting the presence of cables and connectors, different colors, or the presence or polarity of electrical components. Also in Japan, JVC introduced its new PZ2510 videoconferencing camera focusing on live broadcast and recording applications such as concerts and lectures. Based on Ambarella’s CV22 AI SoC, the camera supports full 4kP60 video streaming, face detection, automatic tracking and wide viewing angle. In other IoT markets, Insta360 introduced two new products based on our H22 video SoCs. The Insta360 Sphere is an accessory for a drone and works by placing one camera above the drone and one below to create a seamless spherical image. In the smart home IoT market, Insta360’s Link is an AI-powered 4K videoconference device using a 3-axis gimbal design and powerful AI tracking, automatic zooming and framing, to ensure that the presenter is always center stage. These representative engagements, a majority of which are based on our higher value CV SoCs, provide insight into the early and continued success of our strategy. We have successfully leveraged our human viewing perception processing expertise into the larger machine sensing markets, addressing megatrends such as security, safety and automation. These machine sensing unit opportunities are incremental, and much larger than the human viewing market we have, and will continue, to serve. Furthermore, we are demonstrating we can capture more value in these sensing applications from incremental processing functions such as radar and other sensor perception, sensor fusion, planning, functional safety and application software. CV3 ties all this functionality together. So, we are confident we are implementing the right strategy and demonstrating early signs of success. Despite the current market turmoil, we will continue to drive our organic R&D investments to fully realize these market opportunities, leveraging our leadership position in the AI endpoint market. With that, Brian will now provide our prepared financial comments.

Thanks Fermi. I’ll review the financial highlights for our fiscal second quarter and provide a financial outlook for our third quarter, ending on October 31, 2022. I will be discussing non-GAAP results and ask that you refer to today’s press release for a detailed reconciliation of GAAP to non-GAAP results. For non-GAAP reporting, we have eliminated stock-based compensation expense and acquisition related costs adjusted for the impact of taxes. Revenue for fiscal Q2 was $80.9 million, slightly above the mid-point of our prior guidance range, down 10% sequentially and up 2% year-over year. As expected, supply chain related disruptions were the primary factor in the sub-seasonal performance, and both IoT and Auto revenue declined sequentially. Non-GAAP gross margin for fiscal Q2 was 64.5%, above the high-end of our guidance range of 63% to 64%. Our gross margin outperformance was driven by a higher mix of NRE revenue than originally expected. Non-GAAP operating expense for the second quarter was $44.2 million, compared to $39.8 million in Q1. This was $700,000 above the mid-point of our prior guidance range of $42 million to $45 million, driven primarily by engineering materials for new products. Our non-GAAP tax provision was $361,000, or 4.5% of pretax income, and we reported non-GAAP net income of $7.6 million or $0.20 per diluted share. Now, I’ll turn to our balance sheet and cash flow. Cash and marketable securities decreased by approximately $3 million as unusual working capital benefits in the prior quarter reversed and metrics normalized. You will remember that in the prior quarter accounts receivable benefitted from a front-end revenue skew and inventory declined. For fiscal Q2, DSO increased to 43 days from 28 days and days of inventory increased from 117 to 125. We had two logistics and ODM companies represent 10% or more of our revenue in Q2. WT Microelectronics, a fulfillment partner in Taiwan that ships to multiple customers in Asia, came in at 59% of revenue. Chicony an ODM who manufactures for multiple IoT customers was 13%. I will now discuss the outlook for the third quarter of fiscal year 2023. While we expect some improvement in the second half, from the broad supply chain disruptions, we have started to see some customers reduce the amount of inventory they’re willing to carry into year-end, in particular as component lead times contract. Our guidance, to the best of our knowledge at the current time, contemplates these challenges. We estimate our Q3 revenue to be in the range of $81 million to $85 million, or approximately flat to up 5% sequentially. We estimate Q3 non-GAAP gross margin to be between 63% and 64%. We expect non-GAAP OpEx in the third quarter to be in the range of $44 million to $46 million, with the increase compared to Q2 coming primarily from increased headcount and sales activities taking place during the quarter. We estimate net interest income to be approximately $400,000, our non-GAAP tax rate to be in the range of 4% to 6% and our diluted share count to be approximately 39 million. Ambarella will be participating in Deutsche Bank’s Technology conference on September 1st, Credit Suisse Virtual Asia Technology conference on September 6th, Citi’s Technology conference on September 8th, and UBS’s virtual Future of Mobility conference October 3rd. Please contact us for more details. Thank you for joining our call today. And with that, I will turn the call over to the operator for questions.

Thank you. [Operator Instructions] Our first question comes from the line of Matt Ramsay with Cowen.

Congratulations from me on sampling CV3. I wanted to ask if you could give us a few updates and anecdotes as to how the sampling is going with customers. Maybe how many customers have formally sampled, what those engagements have been like and sort of the early performance feedback on the part relative to what you guys had modeled and just how the software portals to CV3 are going with the customer samples? Thanks.

Right. So first of all, we demoed to maybe roughly 2 dozens of customers, including tier 1s, OEMs. I think there are three important takeaways from the demos. First of all, all of the customers are very impressed that we give a very thorough demo four weeks after the chip come back from the fab. The demo we gave was a live board with CV3 sitting in middle, and we’re taking multiple camera live video. And those live video goes into CV3, and we do our standard image processing for all of the cameras and the process of video went to both our compression engine as well as the CV flow engine. And the CV flow engine running multiple different new networks that’s important for the autonomous driving. And we -- all the video went through that neural network processing. And the end result, both in the process of video plus the neural network results are displayed on a TV monitor. So, you can imagine that this is a very thorough demo, which pretty much touch over all of the important data path of this big die. And we can do that in four weeks after the chip came back. It shows you that not only we have a stable chip, but also our fundamental software side is good enough and solid enough to give live demo. And I think all our customers are surprised and I’m glad to see that. The second thing is the most important is the performance, like you said. In fact, before we demoed CV3, we told them, all our customers, the performance expectation they have. And in fact, we hit those expecting right on, and we did surprise with our real chip. And in fact, that’s a big surprise to them because they are used to see people given the higher projection on the data -- on the people presentation and on the real chip show up, the performance a little lower. But in our case, we definitely deliver as we advertised for CV3. That’s a second takeaway. The third takeaway is power consumption. We also give our customer projections, the potential power number and we hit the power number we talk about. So, in summary, basically, we show people that we have a chip that can run multiple times CV performance than our competitors at a small portion of power consumption and -- which is a live demo, which is not a PowerPoint presentation anymore. So, I think, that’s really hit our customers, giving them a very good impression about our CV3. And the next important milestone is to provide our hardware and software SDK -- software SDK for our customer to do software development and also -- and porting the neural network. So, I think that’s something in progress, and we are engaging customers as we’re speaking. So, hopefully, that we’ll continue to be able to report more. And as I said in my script, we believe that we will be able to give you more updates on our customer engagement by the end of the year.

Thank you for all that, Fermi. That’s really helpful. And congratulations on the progress. As my follow-up, Brian, I wanted to ask about some of the customer kitting issues and your commentary about the potential for customers to want to hold less inventory into year-end. Any way to quantify what that impact is to maybe the second half of the fiscal year? And if you have any color on which geographies, which specific end markets you’re seeing that inventory coming down or folks being willing to take less inventory? Any anecdotes there would be really helpful. Thanks, guys.

Sure. Well, we gave you our outlook for revenue for fiscal Q3. We’re not going to venture out into Q4 at this point. But in terms of where we’re seeing this type of feedback, it’s with customers in various areas. So, it’s not specifically targeted to a single exposure. I think we’re in a situation where customers have been experiencing supply disruptions for quite some time, right? And those disruptions are beginning to abate to some extent. Things are getting a little bit better. But they’re also looking ahead to some of the economic risks that could be on the horizon, and the uncertainties associated with that. And so there’s, I think, a return to a focus on inventory management that maybe didn’t exist for a period of time, particularly as we head into the end of the calendar year. So, we are getting some requests for pushouts from various customers, and we’re reflecting that in the guidance we provided.

Our next question comes from the line of Joseph Moore with Morgan Stanley.

Great. Thank you. Following up on that last question, on the customer inventory stuff. When you talk about lead time contraction, is that your lead time, or is that lead time for other components? And then, I had a follow-up.

I think, Joe, we’re talking about both, but I can talk about our own lead time changes. At a peak, our lead time was at roughly 40 weeks. Now, we are talking about roughly at 30 weeks. And before COVID, our normal lead time was 24 weeks. So, I think we are gradually going back to normal, and that’s basically talking about our lead time. But I think we also noticed some of the components -- other components of our partners also are showing lead time contractions.

Okay. That’s helpful. Thank you. And then, in terms of thinking about October, I know you normally get kind of a seasonal ramp in the consumer-centric businesses, both in surveillance and other consumer, and you’re coming off of a quarter where your revenues were depressed by the foundry issue. So, are you still seeing that kind of seasonal impact, and it’s being offset by these other issues that you’re talking about, or is the impact that’s hitting you kind of more on the more seasonal products? Can you just kind of describe -- it seems like there’s a pretty big fall-off in something if you’re getting that normal seasonal ramp. I just want to make sure I understand.

Right. So, Joe, I think, first of all, those three supply chain issues we have been talking about, although it’s kind of reduced, but didn’t go away, even for second half, we believe will persist, but in a lesser degree. On top of that, we just talked about the inventory reduction at the customers. I think those two things are probably the major factors. And also, I think this is a really unique here. I think with all the supply chain situations that we’re monitoring, geopolitical situation, I have to say, I don’t think that our normal seasonality applied this year.

[Operator Instructions] Our next question comes from the line of Vivek Arya with Bank of America.

Actually, I had a clarification, and two quick questions. Brian, you mentioned there were some NRE revenues that helped to improve gross margins. I was hoping you could quantify how much in incremental NRE in Q2, and what’s embedded in your Q3 outlook?

Yes. Vivek, so we don’t really break that out specifically, but just to give you an idea, NRE for the company typically runs in the 1% to 2% of total range. And in the second quarter, total NRE was kind of low to mid-single-digits in terms of millions of dollars of revenue. And it’s going to be lumpy as we go forward. But, as we look at the near term, Q3 -- fiscal Q3, for example, we see that normalizing and returning back to kind of a typical run rate. But I think as we go out further in time and we have more activities associated with CV3 and so on. It’s possible that that starts to become a slightly larger portion of the overall revenue mix. We’re just -- we’re not quite there on a sustained level yet.

Got it. And on the inventory reduction by your customers, is that more on the automotive side? Is that more on the IoT side? Because on the automotive side, when I look at the results and outlook for most of the other automotive-exposed semiconductor suppliers, they seem to have been pretty strong, at least in the near term. Obviously, I understand that every company is supplying different kinds of component, so it might not be exactly apples-to-apples. But I’m curious, where are you seeing the somewhat softer demand signals? Is that more automotive or more IoT?

No, I think you are right. We see a little inventory reduction on the automotive customer. In fact, they are 25% of total revenue. But the balance of the business on the security -- IoT side, particularly security camera side, we do see our customers trying to reduce the inventory.

Got it. And then finally, your overall inventory kind of stands out, right, because it has now reduced for the fourth straight quarter, which is kind of an anomaly in this sector where inventories are up for most of the semiconductor suppliers. What is the right way to view the lower level of inventory on your balance sheet? Is it a signal of your demand outlook for the next two to three quarters? Is it something else? What is the right way to look at your being able to manage with lower levels of absolute dollars of inventory?

Yes. Vivek, inventory has declined in the last couple of quarters in dollar terms. But if you look at the second quarter, it actually was up in terms of days because COGS went down in the quarter. So, we went from -- I think we were about 170 days of -- 117 days of inventory in Q1, and we were about 125 in Q2. So, that’s a relatively elevated level in terms of days of inventory for the Company. But we’ve been running at a higher relative inventory level because of all the supply disruptions to try to buffer ourselves from some of those gyrations. So, if you think about it longer term, if we can get all these supply issues behind us and get to something of more of a normalized state. I would expect we should be running under 100 days of inventory in the future.

Our next question comes from the line of Gary Mobley with Wells Fargo.

I appreciate the disclosure about how your blended ASP is now trended above $10, but maybe if you can give us some baseline in terms of thinking about it on a quarter-over-quarter basis, year-over-year basis and how that compares to your unit growth.

So, I think what we can say is that since we introduced our CV product line, our ASP continued to grow, and we expect the growth will probably continue, and when we get CV3 or CV3 family of products into production, the ASP will continue to expand. I think the unit number wise, I think these are two separate issues. In fact, where we lose the majority of our unit number is with high vision, [ph] the ASP is very low, which is probably the lowest ASP business in the last several years. And we basically lost that business. And so that also helps to improve our ASP -- company-wide ASP. But I will say that overall our ASP growth will continue base because of our CVflow product line.

And I wanted to confirm that the Samsung 14-nanometer supply issue, the lingering impact in the July quarter was as you approximated $5 million, and I wanted to be clear that that’s no longer a headwind looking into the October quarter. Is that right?

Well, I think -- first of all, the $5 million in July quarter is true. And in October, I think it’s much less, but we will consider when we give that guidance.

Our next question comes from the line of Blair Botha with Arete.

I just had a question on CV5. I can see it’s now entered into mass production ahead of schedule. But just on this, could you talk us through how you see the adoption of CV5 over the next 6 to 12 months just so we could think about so the revenue curve over the next year? Do you see -- expect a sharp adoption? And could it be an important growth driver for the business? Any color on how to think about it as the hot side the revenue stream would be helpful.

Right. I think you should consider CV5 as our next-generation high-end CV4 engine. And the target application is high-end enterprise class security cameras, the enterprise video conferencing system as well as some home application -- high-end home application. And almost all of our security cameras, if they do high-end products, they are looking at CV5, and some of them, like I said, while they will take that into production this quarter and multiple will taking that into production in the second half of this year. So I think CV5 revenue, I think, will start second half of this year and ramp up next year. It definitely helps us on the ASP expansion as well as helping us to maintain our gross margin.

Okay, great. Thank you. That’s helpful. And then, I guess just as a follow-up. In terms of the two segments, IT and automotive, I know you mentioned they declined sequentially in the quarter. Could you give us any color about how to size them in the quarter as a percentage of sales? And then, I suppose just to add to that, looking at the Q3 guidance, could we expect something similar for the October quarter? Thank you.

Can you repeat the question?

Yes. So, I guess, just in terms of the two segments, if you look at IoT and automotive, I was just hoping for any more color about the size of them in the quarter, what they did as a percentage of sales? And then, just to add to that, for the Q3 guidance, could we expect something similar in the October quarter?

Yes. This is Louis speaking. So, auto and IoT, roughly the same when you consider everything included, including the NRE. And the outlook for Q3 is to have both of these areas growing sequentially.

Our next question comes from the line of Ross Seymore with Deutsche Bank.

Thanks for letting me ask a question here. On the automotive side, it seems like it’s been at about that 25% of revenue range now, given the answer to the prior question, it seems like for three quarters in a row. When do you expect that to really accelerate? I know it takes longer in the CV stuff, the majority of that’s been impacting to the benefit of the IoT side. But how do you think about the growth rates of the two segments? And specifically automotive, when should we expect either a steer step-up or a bigger ramp, given all the design wins that you’ve highlighted in the preamble, for me?

I really think that the next important milestone for us in terms of automotive revenue growth is that one major OEM take ADAS Level 2 plus car into production, right? So that’s really the key for us to provide long-term growth. I think although automotive revenue will continue to grow, but I think the step function I’m looking for is really the -- like I said, the major OEM take ADAS into the market.

Any sort of rough color of when that may occur? Is that something that you think you have the design wins, or is that going to be more CV3 dependent, so it’s a couple of years out, given the design cycles? Any sort of rough color on that?

I think CV too [ph] will give some boost and -- but like we said in the past, right, and the CV3 because of the ASP will really provide the biggest jump for us. But I think between now and the CV3 in production, you should continue to see our automotive revenue grow.

Our next question comes from the line of Tore Svanberg with Stifel.

I know you’re not going to comment beyond Q3. But just thinking about the lead times and the sort of inventory adjustment, do you think that you’ll be down to sort of the mid-20s lead times during the October quarter? Or when you talked about sort of going from 40 to 30, that process started a little bit later.

Well, I think 30 is now. And whether we’ll move to 20 24, [ph] like before, it would really depend on our supply chain, right? So I think it will take a while because there’s still, I think, on the high-end process node, probably 30 will be stable for a while. So, I wouldn’t expect that to happen go back to 20 24 immediately. However, I think the trend is there, and I think our customer is really reacting to the trend, expecting that the lead time will go back to normal soon.

Very good. And as my follow-up, can you confirm that you’re going to be updating us on your sort of vehicle revenue funnel 3.0 next quarter? And since you’re obviously not going to shed data now right now, could you at least talk qualitatively about how that funnel continues to improve?

We definitely will give you update on the 3.0 in November.

Okay. Any comment qualitatively or am I jumping the gun here?

I think we are a little -- three months too early to talk about.

All right. Very good. So, since you’re not going to answer that question, I’ll have one last one. So, could you just elaborate a little bit on the supply chain dynamics? I know you talked about the 14-nanometer but I also know there’s been maybe some potential yield issues at 10-nanometer. Now that you are in production in 5-nanometer, how do you feel the yield improvements are going on with your current supplier? Thank you.

Right. So, there are a lot of news articles about Samsung’s 5-nanometer. But I think it’s based on our experience, when we start working with Samsung 5-nanometer, yield was a problem. But now we are ready to production. And in fact, we are in production for -- in Q2. And I can say that the yield is not a problem. It can be -- continue to be improved. But I don’t think that’s a problem for us anymore because we have been working with our foundry supplier for many years. And for any advanced node, it’s a similar experience at beginning that the yield is low, and we work together with our foundry partner to improve it. And I think at this point, I won’t -- I think this 5-nanometer yield is definitely in a normal range and can be further improved in the near future.

Our next question comes from the line of Suji Desilva with Roth Capital.

You talked about the CV3 and some of the implementations there and you talked about some NRE perhaps. Can you just talk about how the design -- how long design cycle there is relative to prior products? I know it sounds more software-intensive. And whether tier 1s -- auto tier 1s would be part of getting that established through the industry? Thanks.

I think, first of all, the -- for any Level 2+ design win, I think it’s going to take probably 3 to 4 years into production in -- outside China. Inside China, maybe a little faster than that. But I think the -- from a design win momentum, I think we’re talking to both OEM and tier 1s. We think that both are very equally important. OEMs want to make a decision and try to control software, but still they need a tier 1 to deliver the final product. We need to work closely with our tier 1 partner to deliver those final products. So, I think we -- that’s why when we went out to demo CV3, we went out to both -- all the major tier 1s as well as the major OEM to make sure they see the capability and the potential of CV3. And so, when they talk each other to make a decision on the design wins that they will consider CV3.

And then, on the Oculii acquisition Radar, can you just update us on what expectations are for incremental fundamental contribution, revenue contribution or how we should think about that over the next few quarters as you integrate that in?

Right. So first of all, we didn’t break up Oculii Radar revenue per quarter, but we gave a guidance of probably $3 million to $4 million this year. I think we are close to that. And we’re going to continue to monitor the current -- the progress that we made with our current customer. And we did talk about there are a few design wins that we’re working on. Hopefully, we’ll see results in the next year, and for the ramp up revenue. I think there’s another very important milestone, I think you should track is when we can demo Oculii software running our CV3 silicon, which will probably -- the first silicon platform that can do sensor fusion with video data and the Radar data at the same time. And like I said before, that’s really important for us to provide not only just better perception and sensor fusion to our customer, but it provides huge bank savings to our customers, too. So I think that also is an important milestone that we’re trying to hit.

Our next question comes from the line of Kevin Cassidy with Rosenblatt.

Just as you’re looking at the opportunities for CV3, is there a difference between consumer vehicles and, say, robotaxis and shuttles? Is there a shorter time to market or any one that might be better suited for the CV3?

Well, I think for both, consumer and the commercial vehicle outside China, design cycle is probably similar, because even for commercial vehicle, people also want to have a auto-grade silicon hardware -- silicon and software. Inside China is a little different, and people are more willing to introduce product and test on in the market. So I think that’s a different -- two different timings. And definitely, we will try to take advantage of the shorter design cycle in China and see where we can quickly introduce some product there.

And going back to the NRE question, is Oculii part of that NRE revenue?

Yes, it is.

Okay. That’s expected to go forward, also?

I think, now we are engaging with a new customer after we acquired them. Yes, we expect that we’re going to see new revenue opportunities with Radar, also.

Our next question comes from the line of Andrew Buscaglia with Berenberg.

Into the quarter, I think there was some optimism growing around some decision-making, moving forward with some automotive OE, just given -- we saw other announcements or some adjacent technologies, specifically VW. So, what -- I understand the inventory commentary, but what about your conversations with just generally in the automotive space around decision-making moving forward? Because as you said, these cycles, you would think we should start to see some movement near term, despite what we’re seeing around the latest on the inventory side.

So, you are talking about CV3 engagement?

Yes, CV3, more so your latest technology, your latest shift that you’re -- yes, specifically, CV3.

Yes. So first of all, the CV3 engagement, like I said, we have done well with the first phase of demos, and we need to follow up with a lot of engagement on the engineering side to help customers to put software to put the new network and demo the system. So, I think that’s definitely an important one. At the same time, for future -- next -- I should not say -- we are developing derivative chips in the CV2 family so that we can address different applications. So for example, we are working on another silicon -- 5-nanometer silicon for auto grade chip, which is going to be enabling our customer to go into production in 2025 -- ‘24, ‘25. And that’s definitely the chip that we are working on, and we’re going to continue to work on different derivative of CV3 chips. And that is definitely an important investment strategy for us moving forward. I hope that answers your question.

Well, I guess that was more so in terms of being able to announce the design wins, just that we’re seeing some movement and we are seeing some decision-making move forward with some adjacent technologies. I guess, where does CV fit in there? I would think that you would start to see some movement, even despite these inventory pressures.

So, I think, like in our script, we talk about -- we believe we can give you more update in the -- before the end year of the year. There are two things we’re going to give you indication. One is we’re going -- in November, we want to give a final update, which will probably give you indication how we’re doing with the OEM in terms design wins as well as we’re going to talk about CV3 engagement that we said by the end of the end of the year, we will talk about that. So hopefully, that will continue to provide you input.

Okay. And then, maybe when you’re talking about customers around specifically the sampling CV3, is there regionally any sort of bias you’re seeing, more interest in one region versus the other, whether it be U.S., Europe or Asia, or too early to tell?

In fact, we -- among the 2,000 customers over -- I think it’s global, and we talk to almost every place, including China, Japan, Korea, Europe and the U.S. So, I think we probably talk to the target customer and we want to talk to in the last few weeks.

Yes. Andrew, just to give you some more perspective on what’s at play here. This is our first central domain controller. And so, you go through a sequential process where we had introduced the product idea to customers in the last couple of years. And as Fermi described, we’ve gone through the demonstration process, almost several dozen customers on a global basis have now seeing the board and it operating live. And then the next step that we described is providing the SDKs and EVKs where they can begin to port their own software onto the chip. And as you move through this progression of steps, a series of steps, customers will feel more comfortable to make that decision, and we’re still saying we expect to provide an update on where we’re at by the end of this year.

Our next question comes from the line of Richard Shannon with Craig-Hallum.

I think just one question for me, and following up on the topic of CV3 here. Well, I know the discussions in demos are very early stage here. Fermi, wondering what you see in terms of future discussions and competitive sourcing dynamics with large automotive customers. Do you expect them to be choosing one domain controller across the entire portfolio, or do you expect them to possibly use two or more? In other words, is this an all or none situation, or is there kind of a split situation over time?

It’s interesting because a lot of the companies that we talk to always start with -- they want to choose one because of software. But at the end, I think they all understand the risk of taking only one. So, I think it’s -- there’s a chance that some people will take two and also there is chance that someone will take one, so between these two possibilities.

Our next question comes from the line of Tristan Gerra with Baird.

You’ve mentioned that the CV5 will start ramping the second half and into next year. Should we assume that CV5 is going to be material to your revenue the second half? And is it material to your Q3 guidance, or is it just initial shipments and then it becomes more meaningful next year?

I will say that the meaningful revenue come from next year. This Q3 and Q4 will be a ramp-up time for several customers, but I think meaningful revenue probably will -- probably move to Q1, Q2 next year.

Okay. And then you’ve mentioned -- as my second question -- some inventory deleveraging across customers, both on the automotive and the IoT side. And I just wanted to get any color that you could provide specific to the China automotive market? How high are inventory levels right now? And have you seen push-outs or delays, either in terms of deliveries or new projects? How is the situation there? And how do you expect it to evolve over the next several quarters?

You’re talking about China market, right?

Yes. Automotive.

Yes, China automotive. So I think if you talk about China automotive, I will say that you need to separate consumer versus commercial vehicles. I think China commercial vehicles are facing severe problems in terms of demand, but -- which is a smaller market. On the consumer vehicle, I think inventory, we haven’t seen -- we’re still -- that market is still facing short supplies. And so that’s why we haven’t seen any revenue reduction -- sorry, the inventory reduction in the Chinese consumer automotive at this point.

And when would you expect then this supply to get back in balance? And will that create more of a slowdown on your existing revenue base to the automotive end market in China by then?

I’m not sure because I don’t know exactly what the inventory level of our current customers. For our chip, right, I can imagine that some that probably build a little bit of inventory. Some people will be conservative. So it’s really going to be customer-by-customer discussion. For example, even for IoT, not all of our customers have inventory problems. Some people are more conservative in terms of building up inventory. So the inventory reduction is more minor for them. But there are some customers that have been aggressive building our inventories throughout the last 24 months. And definitely, there is a long time for them to digest inventory. So it’s really customer-to-customer discussion. We can’t say -- it’s very hard to normalize it.

I’m showing no further questions in the queue. I would now like to turn the call back over to Dr. Wang for closing remarks.

I would like to thank you all for your time and your consideration today. And I’m looking forward to see you in the upcoming events. Thank you, guys.

Ladies and gentlemen, this concludes today’s conference call. Thank you for your participation. You may now disconnect.