Good day, and thank you for standing by. Welcome to the Ambarella Third Quarter Fiscal Year 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. And we ask that you ask one question and one follow-up question. [Operator Instructions] Please be advised that today's conference is being recorded. I would like to turn the conference over to your speaker, Louis Gerhardy. Please go ahead, sir.

Thank you, Lisa. Good afternoon and thank you for joining our third 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 third 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. 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 September 8, 2022 for the second quarter of our fiscal year 2023. Access to our third quarter, fiscal 2023, results press release, transcripts, historical results, SEC filings and a replay of today's call can be found on the Investor Relations page of our website. Fermi will provide a business update for the quarter, Brian will review the financial results and outlook and then we’ll be available for your questions. Fermi?

Thank you, Louis. Good afternoon. Thank you for joining our call today. Q3 was mostly as expected. While there are material headwinds from an industry-wide semiconductor cyclical downturn, there is no change in our very favorable secular growth outlook for the opportunity enabled by our edge AI endpoint investments. During Q3, in four key ways, we demonstrated significant progress to develop these opportunities. First of all, our positive market development momentum continues, highlighted by the November 18th announcement from Continental AG that after a multi-year evaluation, they became the first to integrate our CV3 SoCs into their ADAS product line-up. Second, our six-year automotive revenue is $2.3 billion, an increase of about 28% from the $1.8 billion announced a year ago with approximately $800 million won business. This funnel is predominately driven by our computer vision and domain controller SoCs, and it is important to note our automotive SAM over the same period is still 10x the size of this new automotive funnel, so there is plenty of headroom for share gains. Third, our content in our customers’ products continues to rise as we leverage our historical success with optimized processing for high bandwidth sensing. This is demonstrated by a total blended SoC ASP we estimate will increase about 20% this year. Furthermore, we expect our new SoCs, like CV5 and CV3 to command significantly higher ASPs. Fourth, we are on-track to reach our goal for CV to be 45% of our total revenue this year, and with a strong Q4 CV run-rate, CV revenue is expected to post strong growth in fiscal 2024 and become a larger portion of our mix. Now I will provide some examples of our market development activity. On November 18th, German automotive tier 1, Continental, announced that it will offer Advanced Driver Assistance Systems…

Thanks Fermi. I’ll review the financial highlights for our fiscal third quarter and provide a financial outlook for our fourth quarter, ending on January 31, 2023. 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 Q3 was $83.1 million, in line with the mid-point of our prior guidance range, up 3% from the prior quarter and down 10% year-over-year. Both IoT and Auto product revenue increased sequentially. Kitting issue constraints improved. However, customer inventory reduction actions resulted in sub-seasonal revenue results as we had expected. Non-GAAP gross margin for fiscal Q3 was 63.5%, in line with our prior guidance range of 63% to 64%. Non-GAAP operating expense for the third quarter was $43.5 million, a decrease of 1% from the prior quarter and below our prior guidance range of $44 million to $46 million. The lower than forecasted OpEx was aided by favorable FX impacts on our foreign spending. Q3 net interest and other income was $1.4 million comprised of approximately $800,000 of interest income plus $600,000 of other income. Our non-GAAP tax provision was $1.2 million, or 11.4% of pretax income. This was higher than our original forecast, and typical range, primarily due to taxable FX gains in foreign jurisdictions. We reported non-GAAP net income of $9.5 million or $0.24 per diluted share. Now I’ll turn to our balance sheet and Cash Flow. Fiscal Q3 cash and marketable securities increased $1 million to $199 million. DSO increased from 43 to 54 days, driven by the timing of revenue shipments, and days of inventory decreased from 125 to 124. Cash from operations was $6…

[Operator Instructions] The first question that we have today is from Matt Ramsay of Cowen.

First of all, Fermi, yourself and the whole team, congratulations on the Continental deal. I wanted to ask a few details about how that came together. It seems like even in your pre-silicon simulation results and all of those things that they had been evaluating the CV3 platform for a while, and no doubt that they were one of the early sampled customers of CV3. So, just maybe if you could walk us through how you went from simulated performance and power and capability metrics for CV3 on to sample products on to the press release that they had and sort of the scope of the wins that you might have with Continental? And then, I guess, on the detail side, you updated your automotive funnel up 28% and from $600 million to $800 million on one business. Is the -- how is the Continental deal and potential customer wins from that relationship contemplated in the changes to your automotive funnel?

Right. Well, first of all, thank you. I think that's definitely helpful. I think the Conti -- we have been working with Conti for a long time, even before we tape out and also produce our first-generation CV2 to Conti. As you can see, this is a long process because really, our claim of our CV2 performance and the power efficiency it's hard to digest if you don't see a real silicon. And with CV2 we proved to them the efficiency of our architecture, we can deliver what we promised. And -- but as you know, CV2 is good for ADAS for the Level 2 car, but for the Level 2 plus, Level 3, obviously, they need a more powerful chip, like CV3. And we started talking to them about CV3, I would say, two years ago with just a PowerPoint and our simulation. And of course, because with CV2 silicon, we can really justify how we achieve the performance and power efficiency with our CV3 silicon and plus our simulation. But the deal is really closed when we sampled our CV3 chip and the development platform to them and verified our claims on the performance and power efficiency. And I think that's the time that we really have a major breakthrough with Conti and they are totally convinced our plan and decided to move forward with this relationship. And I think -- that's one reason. The other reason is that I think we are one of the very few companies showing our -- we have a solution on the very low end of like the driver monitor assistance to the e-mirrors to a Level 2 front cameras to Level 2 plus, Level 3, Level 4. We have a silicon road map to cover the whole space that our customers want with a single software base. I think that's another reason that really, I think our customers like Conti appreciate and that help us to secure this deal. And also, like Louis said, we are expecting to continue to do more business development, and we hope we can continue to give you updates in the next quarter in terms of our CVC business development. In terms of potential design wins, I think because the Conti press release didn’t talk about it, I will leave them to answer this question. But I have to believe that with -- because Conti is such reputable and a large customer and they have credibility in front of the OEM customer that will definitely -- between Conti and us, we should be able to get design wins for our OEMs and will gradually show up in our sales funnel in the near future.

Just I guess a final clarification on my first question and then sort of a follow-up for Brian. So, just to clarify, you guys raised the automotive revenue six-year funnel from $1.8 billion to $2.3 billion and the one business from $600 million to $800 million. Is the Conti relationship contemplated in both of those numbers, one of those numbers? I'm just trying to -- if you could be a bit precise on what you guys are assuming in the funnel from that relationship? And did it materially change when the deal was announced? And I guess my follow-up for Brian, you talked about maybe underperforming normal seasonality as you go into fiscal Q1. Maybe the last couple of years have been very strange seasonally for the whole semiconductor space, given shortages and all kinds of other things going on. If you could maybe give us a little bit of clarity about what you're assuming normal seasonality is for fiscal Q1.

I will take on the first question. I will ask Brian to take on the second question. The first question is about our -- how Conti played into the two numbers. I think Conti’s contribution with the current funnel is little -- and there are some of them. But however, I think that we expect that this number of countries impact to our funnel will start showing up in the next funnel next year. And also, we expect that we don't have any Conti contribution in one business yet.

And related to your question on normal seasonality in Q1, of course, normal is difficult to define. But the way we looked at it is taking a five-year average. And in that case, the five-year average for Q1 would be a sequential decrease of 4%. And so, as we sit here today, we think that the decrease going into Q1 will be greater than that number. We don't have perfect visibility at this point to what it will be, but we want to provide as much outlook as we can at this time.

The next question will be coming up shortly. One moment, please. And that question will be coming from David Kelley of Jefferies.

Maybe to follow up on the funnel discussion. You announced a couple of pretty meaningful driver monitoring wins, an e-mirror win as well. So, how should we think about maybe framing it as the interior contribution to the auto funnel versus say the ADAS and autonomous contribution to that funnel growth?

So, first of all, as you can see, for those kinds of design wins, design cycle is much shorter than ADAS because they are not really safety sensitive. So, that for ADAS projects, we usually look at in China, more than two years development, outside China up to four-year development. For this kind of internal design, usually, you can think about a 12- to 18-month cycle that you can get into production. So that's why you start seeing more of those design wins. And that's definitely an area that we want to be really having high market share, and that's what we are doing. But that doesn't mean we don't focus on ADAS. In fact, on the contrary, I believe that we are doing well with ADAS as well as Level 2 plus. And hopefully, we can give you more design win activity in the future.

Okay. Got it. Thank you. And maybe a quick follow-up to your point, on kind of the time line of adoption, and specific to Conti, very early days, as you demonstrate with them to customers. But any early thoughts on time line to eventual revenue contribution with this partnership? Are you seeing OEMs push to accelerate development to integration and production time line faster in ADAS and autonomous as well?

Well, first of all, I will leave Conti to answer the potential revenue that they can generate with this relationship. But like we said before, any CV3 design win will take three, four years to get to revenue. And I don't think the time line has changed.

The next question that we have will be coming from Joe Moore of Morgan Stanley.

I wonder if you could talk to the pace at which that funnel can start to turn to revenue. I think it's a six-year kind of time line that you talked about. I assume that's pretty back-end loaded, given that your number is so much bigger than what you're shipping now. But can you just give us a sense for how quickly that revenue can roll in? And then, how much of that is still -- I don't want to -- be too precise, but I know there's probability weighting on some of it. Can you talk to how much of that we should think still as being a probability weighted event?

So, the methodology behind the funnel and how we discount both, one and pipeline didn't change from our prior practice. And with regards to the distribution of the revenue through the six years of the funnel, it's not linear, as you can imagine. It's exponential in shape where the sixth year is much higher than the first year. And the reason for that would be the time to revenue that Fermi was just describing in a prior question, but it's also very significantly driven by increased penetration rates of these new technologies. And then most importantly, it's actually the higher ASP of the products that we'll be selling in each of these six years. For example, CV3 contributes revenue in years 5 and 6 like we expect, the ASP per chip is much higher than it would be for, say, a DMS, CMS win with a CD25 product. So, those are the factors that caused it to be exponential in shape, not linear.

The next question that we have is coming from Kevin Wang of Stifel.

This is Jeremy Kwan calling on behalf of Tore from Stifel. I guess, maybe a question first on the ASPs. Can you give us any more color in terms of where they stand today? And I understand that once CV3 starts contributing, that's going to grow up even more significantly. But can you help us just give us more details on where they are today and where you see that? And when you said it's going to grow 20% this year, is that for the fiscal year or...

Yes. I think the comment’s for this fiscal year. It was at high single digits and now I think it's above $10 ASP right now. And the major contribution comes from the increased sales of our computer vision chip. And in fact, that's probably the biggest items which help us to continue to get a higher ASP. Now with the CV5 really getting into a ramp-up in production and we start sampling CV3, I also expect that our ASP will continue to this up trend.

Great. Thank you. And I guess, a quick question on the OpEx. I understand some of that is CES for next year -- or next -- sorry, in January. I guess, excluding CES, what kind of run rate should we think about as we look to fiscal '24? Thank you.

Sure. We're still bottoming out on our OpEx budget for next fiscal year. But I guess, the way I would think about it at this point is consider the exit velocity of our guidance for fiscal Q4, which at the midpoint was about $47.5 million. Going into next fiscal year, we're going to have additional chips in development that would cause that number to increase. So, you've got two things to consider. Number one, OpEx has been increasing throughout fiscal '23. And so, even if OpEx remained flat at the Q4 forecast level, it would be up on a year-over-year basis. And then, in addition to that, we'll have some additional spending requirements. That's about what I could give you at this point. In next quarter's call, I think we can provide some more insight.

Next question I have is coming from Kevin Cassidy of Rosenblatt.

Congratulations on the Continental win. And on the Continental win, are there expectations for them to use the scalability of CV3? Will they have a low end solution and the high end, or maybe 3 different solutions?

Yes. First of all, thank you. And also, I believe your read is right. In fact, in their press release, they talk about one of the reasons they chose CV3 family is because they can use a software structure to apply it from low end to high end product line.

I see. And is the Oculii or no -- Ambarella radar solution included in the overall design, or is it still to be determined?

In the press release, we didn't talk about radar. So, I'll keep that -- we will give you more update when we have the right to talk about the potential radar collaborations.

Okay, great. Well, congratulations again.

Thank you.

Next question I have is coming from Tore Svanberg of Stifel.

I had a follow-up question on the Continental win and, again, congratulations on that. If you could just expand a little bit on sort of the -- how the sales strategy works there? Because obviously, you're providing some functionalities but obviously, maybe some other companies will provide others. So, is this like a full reference platform that Continental is offering? If you could just add a little bit of context on how you go to market, specifically with all the parts that are part of that particular solution?

Yes. First of all, CV3 will be the domain controller for the system that we are talking to Conti about. And the goal is that there is no other major processing chips on that system. And -- but the CV3 well take in multiple different sensor modality and perform not only just the perception, but also all the higher-level software functions that the software is provided by Conti.

And a follow-up question on CV3. It sounds like you're starting to -- at least thinking about leveraging that platform into non-auto applications. I was just wondering, both from a sort of capacity and also from a timing perspective, when CV3 could start to venture into other non-auto applications. And I guess the reason I asked the question is because you obviously have a lot going on here in the auto space. So, I'm just wondering if you have enough resources to go after non-auto applications for CV3.

Right. So first of all, using CV3 in a robotic application is definitely a strategy that Ambarella is implementing. And like you said, we are a small company and we have limited resources. Therefore, we need to target and be focusing on strategic accounts. In auto, we have a few -- obviously Conti is the first one, there are a few more. But in the robotics space, definitely, our strategy is to have a target customer which has already some products in the market and as well as they really need performance like CV3 family chips, doesn't matter whether high end, low end. Because if you look at a lot of robots out there, they are -- there have multiple processing silicon in that robot today. And definitely, they can use some kind of domain controller like CV3 to integrate those functions together and provide not only the better cost, but also better system design as well as efficiency. So I think that's definitely a market we will continue to work on.

Yes, Tore, it's Louis. I'll just add. It's a good question because global robotics applications are not part of our automotive funnel, so that is something that's separate and not reflected in that funnel we've communicated.

The next question here is coming from Suji Desilva of Roth.

I'll add my congratulations on the Continental win. Quick question perhaps for Fermi. I think you hinted at this in the last -- in Tore's question there. But the software development that's involved to get CV3 into production, how much of that is the Continental on its customers versus Ambarella? It sounds like it's almost -- maybe almost all Continental. And if so, is there an incremental effort that needs to be happening for every customer that adopts CV3 thereafter, or is there any leverage from the fist win?

Right. So first of all, I think our role in this relationship is we are a strategic tier 2 to Conti, and we are providing support to the -- one of the support we need to provide is porting their software onto our CV3, which we have been talking about this strategy for many years, and we believe we are the one that really focused on this business model versus our competitors. And I think this is -- it's really about setting up a group of people helping our customers porting software. And that had plugging -- we have factored that into our headcount planning as well as our engineering development. So I think that's not a concern for us. And in fact, we expect that we need to continue to do this kind of service for all of the potential customers.

Okay. That helps, Fermi. And then a question for Brian perhaps. Brian, you guided the first quarter next year as well as this year. Are there lingering supply chain post constraint issues factored into your below seasonal guidance for 1Q, or are those kind of behind us at that point? Is that purely a statement about softening demand?

So, we think that those supply chain issues are, for the most part, behind us, and that's something that we don't anticipate to impact our Q1 revenue significantly. What we do see is continued shortening lead times and desire on the part of customers to reduce inventory levels. So, that's really the impact that we're contemplating as we think about Q1 at this point.

The next question that I have here is coming from David O'Connor of BNP. David O’Connor: Great. One from my side on the CV penetration. Fermi, you mentioned or reiterated 45% for this year. Just wondering, can you give us any update on what the exit run rate of this year would be so at to kind of baseline for next year? And I have a follow-up.

Yes. David, we haven't disclosed the run rate as well as the next year's forecast. But I can say that we expect the CV continue to become a bigger portion of our total revenue for many reasons. One is that they are enabling new markets; the other one is their new ASPs. Of course, that portion of our CV growth comes from replacing our existing video processor business. But I think that I'm very excited about the new application that we talk about that our previous video processor chip cannot address. And particularly CV3 and CV5 opportunity we're talking about today is really new opportunity we can address. David O’Connor: Okay. Got it. Thanks for that. And maybe as a follow-up, one for Brian on the sub-seasonal Q1. Brian, you talked about inventory and reduced the customers. Can we expect by -- or do you expect an exit in Q1 that inventory that you're seeing there, customers should be cleared?

That's a great question. I think it's one that we don't have great visibility to. So I'm not going to speculate on that. We certainly, in the near term, anticipate that there will be impacts within that time period, and that's why we want to provide some visibility. But in terms of when things bottom and so on and when those issues go away, I don't think we have a good feel for that at this point.

And the next question I have here is coming from Richard Shannon of Craig-Hallum.

Maybe I'll ask a question on the IoT side of the business here. Maybe Fermi, if you can talk bigger picture here over the last year, how the penetration of CV has gone into maybe both sides of the security business? And where do you think that will go over the next year or two, to be a great perspective over that time? Thanks.

Right. So, I think that's a great question. We continue to see strong IoT design wins in -- on all of the markets, we're talking about professional security and the consumer security and also others. I think -- like I said before, it's driven -- some of it is driven by new applications that we talked about today. For example, this face -- Xiaomi introduced a new lock, which is using 3D metrics to identify face. And also, we are talking about other applications like video conferencing, which is really new market enabled by CV product line. So I think on IoT space, outside of China, we are doing quite well. But at the same time, I want to highlight again that with the new U.S. regulation, we have basically zero play in China in the IoT -- particularly security camera space. I think that's something I want to highlight about, and we talk about [indiscernible] will be zero revenue for us next year.

Okay. Perfect. Thanks for that update. My second question is on Continental here. Looking at it from one angle, and you mentioned both today and in past calls regarding having a solution that really addresses all vision-related opportunities, both within and without the car here. As I look through the Continental press release from whatever, a week or two ago, I didn't see any mention of kind of indoor use cases there, and maybe I've scanned too quickly. But is that being contemplated here? Is that something you can discuss, and/or do you see other OEMs coming through the funnel here that you will have both, interior and external use cases because you have that full offering, others don't?

Right. So, for the Conti price release, they basically focus on only one application, which is ADAS. And ADAS mainly is for external use. However, I do believe that they appreciate that we offer a complete road map that goes from external to internal. And we just -- I want to say that the press release is targeted only for one application, one possible opportunity between Conti and us.

Okay. Fair enough for that update. That's all from me, guys. Thank you.

Thank you.

The next question I have is coming from Ross Seymore of Deutsche Bank.

This is DJ Sebastian [ph] on behalf of Ross Seymore. First off, thanks for letting me ask a question here. But, I was just hoping to get some color on sort of the revenue split between your IoT and auto segments, both for the reported quarter and the implied guide. Thanks.

The approximate split would be about 75-25 IoT versus automotive.

Cool. And just a quick follow-up. How should we sort of think about that going into next year? I mean, do you guys see that dynamic changing significantly, or are we still going to write sort of the same split for the foreseeable future?

Well, like I said, we haven't given any guidance for next year. But like I said, we expect auto percentage will continue to increase because the funnel that we're building. And that should be the trend which we are expecting.

Thank you for your question. And that concludes today's Q&A session. I would like to turn the call over to Dr. Fermi Wang for closing remarks. Go ahead, sir.

Thank you. And thank you all for your time and consideration. We hope to see you at CES or one of the other coming events. Thank you, and goodbye.

This concludes today's conference call. You all may hang up and everyone have a great rest of your day.