AMBA - NASDAQ

Hello, and welcome to Ambarella's Second Quarter Fiscal Year 2026 Earnings Call. [Operator Instructions] I would now like to turn the conference over to Louis Gerhardy, Vice President of Corporate Development. You may begin.

Thank you, Towanda, and good afternoon. Thank you for joining our Second Quarter Fiscal Year 2026 Financial Results Conference Call. On the call with me today is Dr. Fermi Wang, President and CEO; and John Young, CFO. The primary purpose of today's call is to provide you with information regarding the results for our second quarter fiscal year 2026. 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 based on currently available information and 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're 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. Before starting the call, I'd like to summarize our planned investor events for our third fiscal quarter. On September 3, we'll participate in Citi's Global TMT Conference in New York City. September 4, we'll host KGI Securities Bus Tour in Santa Clara. On September 16, we'll host Bernstein's Seventh Annual West Coast Semiconductor Bus Tour at our office in Santa Clara. And on September 18 and 19, Craig-Hallum will host us on the Midwestern NDR. Access to our second quarter fiscal year 2026 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. The content of today's call as well as the materials posted on our website are Ambarella's property and cannot be reproduced or transcribed without our prior written consent. Fermi will now provide a business update for the quarter. John will review the financial results and outlook, and we'll be available for your questions after that. Fermi?

Thank you, Louis, and good afternoon. Thank you for joining our call today. Our strong momentum continued in our second quarter with revenue of $95.5 million, increasing 11% sequentially above the high end of our prior guidance range of $86 million to $94 million. The second quarter results represent the fifth consecutive quarter of record edge AI revenue. Furthermore, I am proud to say the midpoint of our new third quarter and the full fiscal year 2026 revenue guidance range represents all-time record quarterly and fiscal year total revenue for Ambarella. In our May 29 earnings call, we increased our fiscal 2026 revenue growth estimate to a range of 19% to 25% or approximately $348 million at the midpoint. With a strong order book as well as our expectation for both our total unit shipped and our average selling price to increase in fiscal 2026, we are increasing our fiscal 2026 revenue growth estimate to a range of 31% to 35% or approximately $379 million at the midpoint. Needless to say, it is a very exciting time for Ambarella. Fundamentally, after a multiyear period of significant edge AI R&D investment, our broad product portfolio enable us to address a rising breadth of edge AI applications. This increased breadth not only drives our overall unit demand, but we continue to see very strong demand for our new 5-nanometer AI SoCs in both our existing and emerging edge AI markets, which is driving our firm-wide average ASP price higher. I would like to double-click on the rising breadth of edge AI applications I mentioned and focus on three applications we see as rapidly emerging for us. Portable video, robotic aerial drones and the edge infrastructure. Our edge AI revenue began in the enterprise security market more than 5 years ago and was followed by an incremental edge AI application in the smart home, automotive safety and the telematics markets, all of which are continuing their unique growth trajectories. Now this year, layer on top is the rising demand for our edge AI SoCs from the portable video market, including action camera, panorama cameras and the body-worn on cameras. In addition to portable video market, we expect to commence high-volume shipments into the robotics market by the end of this fiscal year. The unit volume in the robotic market is highly fragmented by application, form factors and the customers, but our technology, products and road map have enabled us to win one of the early high- volume robotic applications, partially autonomous aerial drones. Potable video and robotic both represent new emerging edge AI applications in Ambarella's traditional market for IoT end points. Today, we are also announcing our first win in the edge AI infrastructure with our N1-655 SoC. This win is yet another example of the expanding breadth of our edge AI business and I'm encouraged by the interest in our N1 edge AI infrastructure road map from both new and existing customers. In the automotive autonomy market, the largest subset of robotic market. We are actively bidding on OEM projects with our CV3- AD family of advanced 5-nanometer central domain controller for L2+ to L4 applications. While offering significant lifetime revenue opportunities, the lower frequency of award decisions, OEM program delays and a longer time to revenue are causing our other edge AI applications to emerge more rapidly. Nevertheless, we remain highly focused on developing this business, and we will provide updates on our progress as wins occur. I will now describe some representative customer engagement during the quarter, beginning with the two key customer design wins that validate our future vision and strategy. In a rapidly growing robotic drone market, Arashi Vision, also known as Insta360 launched the world's first 8K 360-degree drone under its new Antigravity brand. Powered by our CV5 AI SoC, this drone features dual lens on both the top and the bottom, enabling 8K 360-degree video recording. The AI capacity in CV5 is fully utilized in this partially autonomous drone and our product portfolio will enable the drone market to evolve rapidly to higher levels of autonomy. The Antigravity A1 is set to launch globally in January 2026. We are proud to see Arashi very successfully differentiate their diverse portable video and now robotic aerial drone portfolio with AI features such as new neural net image signal processing, AI editing and gesture control, leveraging our AI SOCs. A majority of Arashi's products are based on Ambarella SoCs and approximately 70% of our Arashi's shipments are exported. In the emerging edge AI infrastructure market, a global networking customer is rolling out a compact on-premises network AI clients with multimodal intelligence at the [ event ] level built on our N1-655 AI SoC. This appliance will add large language model-powered natural language search, and we were selected because of power efficiency, network bandwidth saving and low bill of material cost. This is a great example of one of the green shoots I mentioned earlier. There are several other use cases being evaluated on our N1-655 SoC. In our automotive safety, ADAS and telematics business, I would like to share some key customer wins during the quarter. Samsara, a leading provider of commercial fleet telematics solutions, has introduced its AI Multicam platform based on Ambarella's CV72 AI SoC. Samsara's AI Multicam delivers live 360-degree visibility and real-time risk detection alerts on an in-cab monitor with up to 4x auxiliary HD camera feeds. It is great design win for CV72 that demonstrates more camera inputs and advanced AI features on a single SoC. Audi is utilizing CV22FS for their left, right e-mirror functions in the E5 model initially in the China market. It enables them to provide intelligent context-adaptive viewing mode on highways, parking, turning and lane changes with dynamic image processing and display enhancement functions. Also in the mirror market, our BAIC Stelato S9 is utilizing CV22FS for their rearview electronic mirror. They note that AI-aided detection via camera input helped them cut down blind spots by up to 60%. And a leading Chinese OEM will utilize our CV22 SoC for their 8-megapixel sensor designed specifically for the Level 2 front ADAS functionality. A key capability they are enabling is small target detection and the long range. In the enterprise security segment, Honeywell in India has launched their 50 Series enterprise security cameras in 3-megapixel and 5- megapixel resolutions based on our CV25 SoCs. India is a fast-growing market with a drive for made in India products, creating new customer opportunity for us. In the smart home market, one of our long-term customers in U.S. has leveraged our H32 SoCs to build multisensory multimodal AI products available in the retail outlets today. They have built a nursery device, integrating video monitoring two-way intercom, high noise generator and air quality sensor. They have also built a garage device that features carbon monoxide and heat detection, security camera and the intercom functionality. Also in the smart home market, Netatmo launched their Indoor Camera Advance products that is built on [ S6L ] SoC in the European market. As you can see from this representative customer engagements, we continue to build design wins momentum in our existing edge AI endpoint application, and we continue to successfully address incremental edge AI applications. such as robotic aerial drones and edge infrastructure as the edge AI market breadth expands. Having shipped more than 36 million edge AI processors to hundreds of customers who have successfully ported hundreds of advanced customer AI models to our SoCs, there should be no doubt that Ambarella is a leader in edge AI. Edge AI is expected to represent about 80% of our total revenue this year. We are focused exclusively on the unique needs of the edge AI market, and we continue a rapid pace of innovation. In conclusion, I would like to summarize the key points covered today. First, we delivered Q2 results above the high end of our prior guidance, and we increased the midpoint of our full year fiscal 2026 revenue guidance by 9%. Second, the breadth of our edge AI applications we are successfully addressing is expanding as seen with our ongoing ramp in a variety of portable video applications and the anticipated production ramp for robotic aerial drones and edge infrastructure. Third, the growth of our edge AI business is also occurring with our higher-priced edge AI SoCs, supporting the anticipated growth in our ASP. Last, we are exclusively focused on the unique requirements of the edge AI market, and we remain an established edge AI market leader who continue to innovate at a rapid pace. Now John will now discuss the Q2 results and the Q3 outlook in more detail.

Thank you, Fermi. I'll now review the financial highlights for the second quarter fiscal year 2026, ending July 31, 2025. I will also provide a financial outlook for our third quarter of fiscal year 2026 ending October 31, 2025. I'll 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 and acquisition-related expenses adjusted for the impact of taxes. For fiscal Q2, revenue was $95.5 million, above the high end of our prior guidance range of $86 million to $94 million, up 11.2% from the prior quarter and up 49.9% year-over-year. Sequentially, automotive revenue increased in the mid-single digits and IoT increased in the low teens, with IoT growth led by the adoption of edge AI in portable video applications. IoT in fiscal Q2 represented slightly more than 75% of our revenue and is spread across an increasing number of edge AI applications. Non-GAAP gross margin for fiscal Q2 was 60.5%, at the low end of our prior guidance range of 60.5% to 62% due to product mix. Non-GAAP operating expense in Q2 was $53.4 million, below the midpoint of our prior guidance range of $52.5 million to $55.5 million, primarily due to lower engineering related costs associated with the timing of product development. Q2 net interest and other income was $2.2 million. Comparing to our prior guidance of $1.8 million, the increase was primarily from higher interest income. Q2 non-GAAP tax provision was approximately $200,000. We reported a non-GAAP net profit of $6.4 million or $0.15 per diluted share in Q2. Now I'll turn to our balance sheet and cash flow. Fiscal Q2 cash and marketable securities reached $261.2 million, increasing $1.8 million from the prior quarter and $41.4 million from the same quarter a year ago. Increased cash and marketable securities benefited primarily from operating cash flow associated with increased revenue, partially offset by increased expenditure on capital investments during the quarter. Receivables days sales outstanding increased from 31 days in the prior quarter to 40 days, while days of inventory decreased from 98 days to 85 days. Operating cash inflow was $5.5 million for the quarter. Capital expenditures for tangible and intangible assets were $4.1 million for the quarter. Free cash flow was $1.4 million. We had one logistics company representing 10% or more of our revenue, WT Microelectronics, a fulfillment partner in Taiwan, that ships to multiple customers in Asia, came in at 71% of revenue for the second quarter. I'll now discuss the outlook for the third quarter of fiscal year 2026. The breadth of our edge AI business is expanding with a strong unit and average selling price outlook. As a result, in Q3, we forecast revenue in the range of $100 million to $108 million or $104 million at the midpoint. Sequentially, we expect mid- to high single-digit percent growth in our automotive business with our IoT business up in the mid-teens. For fiscal 2026, we anticipate a revenue growth range of 31% to 35%. We expect fiscal Q3 non-GAAP gross margin to be in the range of 60% to 61.5%. We expect non-GAAP OpEx in the third quarter to be in the range of $54 million to $57 million, with the increase compared to Q2, driven by new product development costs. We estimate net interest and other income to be approximately $2 million. Our non-GAAP tax expense to be approximately $800,000 and our diluted share count to be approximately 43.7 million shares. Thank you for joining our call today. And with that, I will turn the call over to the operator for questions.

[Operator Instructions] Our first question comes from the line of Christopher Rolland with Susquehanna.

Congrats on a great quarter. So for my first question, I think for years, you guys pitched yourself kind of as the future of the company being automotive first, but IoT at this point has just been an incredible outperformer. I think it outperformed auto by 4x this year. So I guess my question is, are you thinking about IoT differently now? Could there be a pivot in your business where you just double down spending around IoT versus auto, lean into the development of IoT versus auto? And when might we get to a point where auto outperforms IoT? Or is this not the case just given the great interest in IoT?

Right. Thank you for the question. I think the first part to the answer is that like I said in our script that we are continuing to focus on our Level 2+, Level 4 autonomous driving, and we are working hard to continue to win design wins there. But also, I pointed out that because our other edge AI business because of design -- shorter design cycle and more available opportunity for us, we are making significant progress there. And we're going to continue to focus on the edge AI market, including both autonomous driving as well as IoT. But I want to point out that the fundamental hardware architecture between the edge AI for the IoT side and the autonomous driving side are identical. Our CV4 architecture, our image processing pipeline, our CPU investment, even on the OS side, there are huge leverage between each two. So from the OpEx expense side, the leverage is very strong. Obviously, the go-to-market strategy from the marketing side, on the sales side are different. But we are going to continue to focus on those two areas, because I still believe in the long-term autonomous driving continue -- can keep driving our strength. But as you can see in our announcement, we made a significant progress on the edge AI in the IoT side where that means we're going to put also more resources on this than before to continue to make progress and try to collect more market share in this particular market.

And yes, just maybe back to the growth rates, just a couple of things. First of all, would you expect auto to outgrow IoT next year? Or is this really going to be -- you've talked about auto and your CV3 wins, I think ramping in 2027, would we have to wait for auto to outperform at that point in time?

Right. So I think the auto will outperform IoT, where we have major design wins with OEMs like the 1 that we talk about VWK 2.0. Have we won the design? Yes, I think in the 2027-2028 time frame, we can see that, that auto growth will be -- outperform IoT. Right now, I think in the foreseeable future, that before we get any major design wins from the automotive side, IoT will continue to have a very strong contribution to our income. In fact, that our current growth, you can see that the growth rate that we got from the just IoT side significantly improved over the last few years.

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

Congratulations on the great results and outlook. You piqued my interest with the robotic aerial drones. You mentioned Insta360. Is this another trend of will there be multiple companies coming out with these solutions? And are there commercial applications like for deliveries?

Yes. So first of all, I think for Insta360, their target market is commercial and consumer, and the volume is significant compared to what we have seen in the market outside DJI. And then also that we are seeing -- definitely there is a market trend, a lot of different companies in different countries are focusing on the strong particular drone application. Now that autonomous driving on the car side become more popular technology widely available. You can imagine that the autonomous drones will become popular. And with that, that will enable many different possible applications in the near future. I think that potential trend is driving -- this is really consistent with the robotic trend that we are seeing in other applications when the autonomy become popular and become possible than the possible applications with those robots or drones become -- in the past, it was impossible not definitely thinkable. So I think that we continue to engage in multiple drone design wins activities, and we think that you're going to continue to see us to report our success in this market.

Kevin, it's Louis. Just to kind of add on to that and maybe tie it into Chris' question. This is just a great example of you've got multiple high-bandwidth sensors in a real-time application collecting data and driving a higher level of autonomy -- higher and higher levels of autonomy, just like in a vehicle, moving from L1 to L4, you see the same sort of trend beginning in the aerial drone market, and of course, other robotic spaces. And it's all happening with the same underlying AI inference accelerator, that's in common across all these markets, whether it's auto autonomy, auto safety and telematics or any of these IoT markets. So we leverage the technology across a lot of different applications.

Right. Yes, I guess your energy efficiency also is very useful if you're going to be flying something that has to have a battery and energy efficiency is really important. And just -- you have so many exciting things happening with your new designs. You didn't hear much about your process technology or moving on to the next generation. So that's still on track to moving to 2 nanometer.

Absolutely. In fact, we are -- our foundry partners continue to announce design wins not only give us a lot more confidence but also our potential customers. I think that we will continue to work on the 2-nanometer project and still remain targeted to take our customers to production in early 2027.

Our next question comes from the line of Quinn Bolton with Needham & Company.

This is Shadi on for Quinn Bolton. Congrats on the strong results. My first question is on the guidance. Your Q3 guide implies a seasonally down Q4. And given all the progress you guys have been making and the edge AI tailwinds, they're still somewhat conservative. So just want to get your thoughts on maybe the puts and takes as we think about Q4.

Right. So first of all, I think that the seasonality that we are guiding for this Q3 and Q4 is increasing the range compared to our previous year. So I don't think that should be a surprise. But if you look at that a lot of the products -- some of the products become driven by consumer cycles, that will definitely explain to you, that's why we are seeing the seasonality based on our guidance.

Got it. That makes sense. And then my follow-up is on the nonsecurity camera portion of the IoT business. How does Ambarella view this segment growing over the next few years? And at what point might the nonsecurity segment surpass the security camera segment of the IoT business.

Yes. First of all, thank you for that question. I think that's important. Internally, we're looking at that also because all of the new applications we announced today, none of them is really on the traditional security camera business. In fact, that from the drone robots to the portable video to the Edge infrastructure, those are the -- really the new market we have been talking about that we haven't shown much result until this quarter. And I think that -- I think enterprise or -- the enterprise security and home security continue to be -- the combination continues to be a large portion of -- compared to others, but I think that we do see that the growth rate on the non- security portion of the business will continue to outpace on the other side.

Yes. Just to be clear, Shadi, our security business, we expect to continue to deliver very good growth. But now you have these portable video and some of the robotics markets and other things kicking in that, as you observed, are causing our other IoT business outside of security to contribute very nice growth for us.

Our next question comes from the line of Liam Pharr with Bank of America.

This is Liam on behalf of Vivek. There's been a lot of media reports recently about M&A and industry consolidation. And I was wondering if you're able to address kind of what role you expect industry consolidation to play and what your strategy looks like if you remain independent?

So obviously, we cannot address that rumors. I think we just have no comment on that. But however, I want to point out that with today's earnings call you can see that the HAI -- the importance on the strategy side of HAI become so obvious in the market space, and with that -- we are probably 1 of few, maybe only 1 shipping 36 million units of HAI SoC so far, put us as a leader in that market. So with the combination, I really think that the rumor base is that our strength and our focus on HAI, and I think that's going to continue to play very well for us.

And then just as a follow-up. In terms of kind of going back to the IoT and auto side, clearly a strong quarter. What does the sustainability look like of these growth drivers through 2026? And where should we kind of expect more of an upside trajectory on the IoT or on the auto side?

On the auto side, I think we definitely continue to work hard to get a secured first design win on the Level 2 plus Level 3. That is really what pushed our growth trajectory beyond what we have with automotive. And with IoT, we are really growing significantly this year over last year and because -- thanks to the contribution of a few products ramping up by our customers. So we believe the growth will maintain and we're going to provide the guidance for next year. And we will definitely believe that automotive and both IoT and automotive will continue their growth trend.

Yes. Just to put a little more color on it, it's Louis. It's not like there's just a couple of markets that are contributing to the growth. 5, 6 years ago, it started for us in enterprise security and it was public and smart home, then AI video telematics and commercial fleets, certain in-cabin, e-mirrors or driver monitoring. But now more recently, in IoT, you've had portable video, which is not just 1 thing, but it's body-worn cameras, it's panorama cameras, it's action cameras, and now we're moving into robotics initially with aerial drones expected to become significant. So it's not really like are those -- are you in a couple of markets and are they going to static and how are they going to do? It's more about edge AI touching more and more different vertical applications, and that's what's been happening to the business.

Our next question comes from the line of Kyle Smith with Stifel.

This is Kyle Smith on for Tore Svanberg at Stifel. Congratulations on the strong quarter. So I think it's pretty clear that the strong revenue be in guide is stemming from tangible design wins and product momentum. But that being said, could you provide more commentary on the process that management uses could check for any potential demand pull-ins related to the tariff environment? Are you speaking directly with customers or distributors, monitoring yourself for any irregularities? Or is it kind of a mix of multiple factors?

Yes. I think that's a very important topic internally because we're all going through this industry-wide inventory correction for the last 3 years. And every time we're seeing some high growth of area. The first reaction is we better talk to customers. So in the past few years we built a relationship with all the customer and also our distributor to make sure that we review inventory every month. And then based on that, we try to decide whether that we're seeing any inventory build. So far, I think throughout the process with this internal check, we haven't seen any inventory build beyond the normal practice. And also, personally, every time I have a meeting with my peers in our customer base, one of the topic is always about supply chain and about the supply. And I got no feeling that nobody telling us that they are building excessive inventory worrying about geopolitical situation. So with that, that's just from the feedback from customers. But more internal insight, we build some kind of a checkpoint to understand, look at the customers, the older patterns and then whether that's associated with any product ramping up. So if there's any indication of actual inventory build, internally you have -- we have some data now. So far, based on all of this internal and external discussions, I think that we feel quite confident that we haven't seen any meaningful inventory buildup in our customers.

Perfect. And you mentioned a lot of really exciting design wins in the prepared remarks. I'm curious what the customer response has done to the Cooper development platform, particularly within these new and emerging markets? And are there any specific components of the platform showing outsized positive feedback?

I think, first of all, the feedback from our Cooper development platform is very positive. Not only it helped our customer to move from 1 off the chip to another chip easily because Cooper platform cover all the chips that we develop. And so that for our customers, it's really become a powerful tool for them to develop the product once and they can put -- use the same product to many different chips under Ambarella. So that's 1 of the most important thing. But because with the investment now we can easily enable our customers to play with our SDK and also play with all the new network we put into our model garden and also enabling them to learn how to use our compiler to compile the new network to our hardware. All of those features, all integrated into this Cooper platform. So -- of course, I'm not saying that is perfect, but definitely with the benefit to customers, they will continue to give us great feedback about how we can continue to improve it so that they can enjoy the platform more.

Perfect. And if I could just sneak 1 more in. Contemplating this really outsized revenue growth, do you continue to expect non-GAAP OpEx to grow at around 10% annually? Or should we maybe bake in a little bit higher OpEx going forward?

Yes. Thanks, Kyle, for the question. I think it's reasonable if you take this little bit higher than 10% is probably reasonable. I think quarter-over-quarter -- I think year-to-date, we're at about 12 -- year-to-date, we're at about 12% growth. I think we'll probably stay in that range for the full year.

Our next question comes from the line of David O'Connor with BNP Paribas.

Maybe for me, just going back on the automotive side of things again and ADAS. Through this year the L2+ adoption rates have slowed, software not ready, OEMs optimizing for price. I mean you guys have talked about this through the year. As we sit here in August, and from your recent conversations with customers, can you talk about any changes there on how they're viewing that kind of adoption on their next model? Any sign that they may be pulling it in? Or just any kind of changes that you're seeing there across the kind of -- that will help kind of frame the backdrop for potential C3 wins?

I think the scenario described continues. And we continue to see OEMs coming out bidding on the, I would say, more low end Level 2+ than higher end than that, because people -- like you said, OEM really focus on getting a proper cost than functional features. In fact, even in China recently, we start seeing a similar trend because Chinese government definitely trying to make sure that the time driving becomes safe and the safety become the most important feature. So I really think that the total trend of autonomous driving is focusing on safety and also lower end of the function performance. For example, we announced a design win on 8 megapixel ADAS in China. Let's just give you an indication that while we continue to build on all kind of different features and that we see more opportunity on the low end Level 2+ and also ADAS opportunities.

Yes, David. I mean we still see very significant lifetime revenue opportunities in the auto autonomy market for sure. But there's -- as Fermi mentioned earlier, there's a lower frequency of decisions. The market can be subject to delays like you referenced, and there's a longer time to revenue. So what's been happening is all of these other edge AI markets have more than caught up and are growing very rapidly for us now. But we still have these products and very much focused on landing these wins. It's just the frequency of them isn't as high.

That's very helpful. Maybe 1 for John. Just on the incremental kind of growth year-over-year. With the new guide, you're kind of up maybe $95 million, $100 million, something like that, for the year. Is there any way you can kind of split that out in terms of units versus ASP or content just kind of how you would break that down as kind of a percentage, half of it unit-growth, half ASP. Any kind of steer that would help us there size the kind of difference between those 2 drivers?

Yes. Thanks, David. I think what we've been seeing throughout this year as it's pulled together, our estimate is that, that growth is roughly 50-50 between ASP and unit growth.

Our next question comes from the line of Gus Richard with Northland Capital Markets.

Congratulations for the strong results. Just in the IoT market, could you give us a split between the security applications and the nonsecurity applications and which of the nonsecurity applications are growing most rapidly?

So in a nonsecurity application, I think that portable video definitely grow the fastest. And in fact, that Insta360 using a CV5 to build their next generation sports camera as well as the panorama 360-degree camera and the ASP is high and the unit number continue to grow. So that definitely is a faster-growing market. I won't be surprised if we see a lot of growth in the future, see some growth on the drone side too because the ASP and unit number growth can be significant, too.

I guess, the auto business in Q2, I think John mentioned, grew in the mid-single digits and IoT grew in the mid-teens, and that would put auto in the low 20% range as a percent of revenue in IoT the balance.

Got it. And then just in terms of the IoT business, you've got a wide diversity of applications, and I would imagine that your customers need support from field application engineers. And I'm just wondering, is that a limitation? Is that something that you need to bolster to help accelerate growth? How are you thinking about customer support in that regard?

So first of all, a unified hardware and software platform we just mentioned, is really helpful, because that means our field engineering can easily switch from 1 customer to another customer, although maybe a different application, different products, but the fundamental hardware and software are almost the same. So from that point of view, we definitely can leverage our field engineers in different applications. But you are right that our revenue growth, and when we're looking at the different customer base, we continue to add to our field engineering, which is part of our growth plan that John highlighted just a few minutes ago.

Our next question comes from the line of Martin Yang with Oppenheimer & Co.

So on the strength you called out on the portable video products, can you tell us if the strength is driven by a single key customer? Or have you expanded your customer base with new design wins with new OEMs in the past quarter?

Well, in fact, we continue to have multiple customer base in this space. But however, Insta360 definitely is the largest 1 that we mentioned because they switch from H22-based video processor only solution last year to this year's CV5-based solution, that ASP growth definitely is 1 of the main reasons we continue to see the growth from them. But we continue to engage with multiple portable video players throughout our career.

Martin, I think you're familiar with the company, but we're selling into like 7 different portable video product lines there that would include action camera, sports, panorama, but also body worn webcam, video conferencing, and now aerial drones. So it's a lot of different product categories. It's not just a few.

A follow-up question on Insta360. So in your guidance, do you assume business as usual with them without any potential impact from their ongoing lawsuit in the U.S.?

Well, first of all, yes, we look at it. And it's not in our position to make a judgment on the outcome of lawsuit and we'll leave that to the 2 parties. Our assumption is based on the POs we receive from our customers. And that's the only thing we are counting on to forecast our business.

[Operator Instructions] Our next question comes from the line of Richard Shannon with Craig-Hallum.

Let me ask a question. First 1 is on the broader edge AI opportunity. You talked about your first design win hoping to ship near the end of this fiscal year. Maybe you could describe what the pipeline looks like. Maybe describe it, even quantify a number of design opportunities and kind of the -- any maybe new applications you're seeing here versus what you described in the past.

Edge infrastructure, right?

I think, Richard, was your question about just edge IoT overall or just the infrastructure?

I'm sorry, I misspoke. Edge infrastructure. Sorry about that.

Exactly. So first of all, you know that we have been working on the N1-655 product and for a while talking to many customers. And this particular design win is our first design win that we can talk about. You can imagine that we are definitely engaging with multiple customers, new and old or existing customer with potential design wins. And you should expect that we'll continue to talk about our progress in this particular market, and the particular -- I think there are so many different type of potential appliance that people can build. But in general, you can imagine that this kind of appliance is really trying to aggregate multiple edge end points, and apply most advanced AI models on that and to provide different services. That's just in general terms to describe opportunity out there. And definitely this kind of appliance need to run not traditional computer vision part, more importantly, over a large language model or vision leverage model are probably the focus area where our customer wants.

Richard, we've talked about a SAM for this market of -- in this year, this fiscal '26 being around $125 million and in 5 years, approaching $500 million and we feel those figures are conservative. We're still learning about the market. As Fermi said this is our first design win, but we're pretty excited about the level of interest from customers, both new customers and existing customers for Ambarella.

And the success of this market will continue to drive up our average selling price.

Okay. Great. Second question here is on the portable video opportunity here and following on the questions and responses from the past couple of questions here. To what degree are these opportunities or applications more consumer-oriented versus enterprise in nature?

Well, it depends on the market. But I'd say overall, across all 7 that I just described for like Insta360, more weighted to consumer. And -- but there's still being sold into enterprise applications. For example, body worn cameras is a market that at least today is very heavy enterprise and public safety driven and that's 1 of the categories. But if you switch over to some of the other portable video markets, it might be more on the consumer side. And overall, I'd say they are weighted more heavily to the consumer side, which is 1 of the factors that allows them to get to revenue faster.

Thank you. Ladies and gentlemen, I'm showing no further questions in the queue. I would now like to turn the call back over to Dr. Fermi Wang, CEO, for closing remarks.

And thank you for joining us today. We are going to see you next time for sure. Thank you.