Ambarella Inc (AMBA) 2018 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Ambarella Third Quarter Fiscal Year 2018 Earnings Conference Call. As a reminder, this call is being recorded.

I would now like to introduce your host for today's conference, Ms. Deborah Stapleton, Investor Relations. Ms. Stapleton, you may begin.

Thank you. Good afternoon, everyone, and welcome to Ambarella's Third Fiscal Quarter 2018 Financial Results Conference Call. Thank you for joining us today. Our speakers will be Dr. Fermi Wang, President and CEO; and George Laplante, CFO. The primary purpose of today's call is to provide you with information regarding our fiscal 2018 third quarter results.

The discussion today and the responses to your questions will contain forward-looking statements regarding our projected financial results, financial prospects, our markets, demand for our solutions and the company's efforts to develop and commercialize new technologies, 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 that we file with the SEC, including the annual report on Form 10-K that we filed on March 30, 2017 for the 2017 fiscal year and the Form 10-Q filed on September 8, 2017 for the second fiscal quarter of 2018. Access to our third quarter results press release, historical results, SEC filings and a replay of today's call can be found on the Investor Relations portion of our website.

I will now turn the call over to Dr. Fermi Wang.

Thank you, Deb, and good afternoon. Our Q3 2018 revenue was $89.1 million, representing a decrease of 11.4% from the same period in the prior year. Although our revenue was lower than for last year, our non-GoPro revenue was 7.2% higher, and we enjoyed strong design win momentum in our target IP security and automotive camera markets. And we are very pleased with our execution and the progress in the development of our computer vision solutions.

During the third quarter, we demonstrated and assembled our new CV1 chip to key IP security and automotive customers. Based on our CVflow architecture, CV1 combines Ambarella's traditional strengths in high-resolution imaging for camera capturing video to be watched by humans and with the stereo vision and the neural network processing required for advanced machine vision.

In particular, the CVflow architecture is designed to support the performance requirements for perception and deep learning applications, including object recognition. As an update, CV1 has met our aggressive targets for features, performance and power consumption. And we look forward to demonstrating many of the advanced features of our CV1 chip, including in advanced ADAS systems, at the Consumer Electronics Show in January.

Beyond CV1, we remain on target to deliver our new generation of SoC solutions in the first half of next year that are based on the CV4 architecture. Leveraging 10-nanometer process technology, the SoCs will target high-volume IP security drone, VR and OEM automotive camera applications.

In the automotive market, we see opportunities in multiple applications spanning car recorders, electronic mirrors, surround view, Advanced Driver Assistance Systems, or ADAS, and ultimately, fully autonomous vehicles. Initially, our design wins and revenue are coming from car recorders that are transitioning from aftermarket dash cameras to OEM original equipment, primarily in Asia. We are now also actively engaged with the world's leading automotive mirror makers to deliver electronic mirror solutions for future generations of cars and the commercial vehicles. This includes both single-camera rearview electronic mirrors as well as a single-chip, triple-camera designs that can replace or augment rear and side view mirrors. Most major carmakers are now planning for deployment of electronic mirrors on future models, and that we anticipate revenue from these applications next year beginning in China, the world's largest automotive market.

Looking forward, we see computer vision as a critical component of our automotive solution as ADAS features will be required for many automotive applications. For example, future generations of car recorders, electronic mirrors and surround view systems may include a features such as Advanced Lane Departure Warning System, forward collision warning, pedestrian and blind spot detection. We believe Ambarella's advantages in image quality combined with powerful computer vision processing will enable carmakers to deliver the best HD video experience for human vision with the highest level of driver safety. We also believe that many future ADAS systems will require video recording to provide a record of events, further leveraging Ambarella's advanced video encoding capability.

During the quarter, Honda introduced dealer-fit drive recorders. Powered by Ambarella's A7LA video processor, the new recorders offer full HD video recording and seamlessly integrate with Honda's navigation and infotainment systems. The camera includes a parking recording mode to provide video evidence in the case of parking lot accidents and HDR processing for high-quality recording in challenging lighting conditions.

In China, carmakers BAIC and Geely both introduced new car models featuring car recorders based on Ambarella's S2L SoC. BAIC's Senova D50 includes an HD recorder with Wi-Fi connectivity to allow video to be watched on a smartphone, while Geely's GL model car recorder includes full HD video storage and a simultaneous 720p display on the vehicle's infotainment system.

In the professional IP security camera market, we see increasing requirements for advanced computer vision and video analytics across a wide range of performance and price points. During the quarter, we demonstrated and sampled our new CV1 chip to leading IP camera makers as a platform for the development of their next-generation cameras with these capabilities. Additionally, European and U.S.-based customers are now deploying cameras based on the new HEVC or H.265 video standard, following its earlier adoption in the Chinese market. During the quarter, European security supplier, Bosch, released its new AUTODOME IP 4000i, 5000i, and the 6000i camera families. Based on Ambarella's S3LA SoCs, the camera support both H.264 and H.265 video and includes up to 30x optical zoom and the full HD video at 60 frames per second. The cameras includes built-in analytics to distinguish between object types such as people, cars, bikes, motorcycles or trucks, and may also be used in applications beyond security such in business retail environment for monitoring customer activity.

And in the fast-growing connected home market, we saw the introduction of a variety of new home monitoring cameras spanning video doorbell, security spotlight and battery-powered designs. During September, Nest introduced its Hello HD video doorbell. It features a full 160-degree field view to allow people to clearly see visitors as well as any packages left on the ground. Based on Ambarella's S2L camera SoC, the Hello camera includes a high dynamic range of processing for viewing in high-contrast conditions as well as night vision. The camera can detect if a person is approaching even if they do not ring the bell and provide an alert via smartphone.

Also during the quarter, Ring announced its Spotlight Cam Wired, an HD camera integrated with spotlights. Based on our S2L chip, the Ring Spotlight provides a full 1080p HD video, a wide-angle camera lens and built-in microphone and speakers. An advanced motion detection engine with custom motion zones provides reliable and instant motion alerts if someone approaches your home.

And in the battery-powered category, Swann released its smart indoor/outdoor full HD wire-free security camera priced at just $129. Based on our S2LM chip, the battery-powered camera includes infrared night vision wide 120-degree field view and offers fast battery recharge.

In the consumer market -- in the consumer camera market, we are seeing a number of innovative new product introductions for the holiday season. Many of the new models include support for 360-degree video, advanced image stabilization and software applications that will address simplifying content creation and video editing.

In October, California-based startup, Rylo, introduced its small form factor 360-degree camera selling for $499. Featuring dual wide-angle lenses and using Ambarella's H2 SoC, the Rylo captures 360-degree video in full 4K resolution and creates panoramic 6K resolution still pictures. The Rylo app includes advanced stabilization technology and object tracking to follow a person or object to keep them in frame.

Also during the quarter, San Francisco-based startup, REVO, began shipping its ARC camera, which has a gimbal integrated into the rear of the camera to keep it level with the horizon. Featuring Ambarella's A9AC SoC, the camera shoots 4K video and 12-megapixel stills. The REVL editing app automatically finds the best parts of your videos, and it can also play back data captured by the camera, including height, speed and how many rotations you have made.

And in August, Insta360 introduced its Insta360 ONE model that can be used as both a stand-alone camera and as an iPhone accessory. Priced at $299 and based on Ambarella's A12 SoC, the camera can shoot 4K video and take 24-megapixel stills. Its free-capture feature enables user to shoot first and point later by first capturing 360-degree video and then panning and zooming on the phone to create 1080p director's cut.

Also during the quarter, GoPro introduced its Fusion 360-degree camera. Based on dual Ambarella A9SE SoC, featuring its cable, we're shooting 5.2K video at 30 frames per second, taking 18-megapixel spherical photos and capturing 360-degree sound. Selling for $699, it also integrates Wi-Fi, Bluetooth, GPS and voice control.

And earlier this month, we were excited to see that Ambarella-based VR camera has arrived at International Space Station. The Vuze VR camera from Israel-based HumanEyes supports both 360-degree and 3D video and will be used to film a cinematic virtual reality experience as part of National Geographic's upcoming series, "One Strange Rock."

And in the drone market category, market leader DJI released its new Zenmuse X7 camera based on our H1 chip. The Zenmuse X7 is a compact Super 35 camera with an integrated gimbal made for high-end filmmaking that delivers ultra-HD video and outstanding image quality. The X7 boasts very high sensitivity to light, shoots 4K video at up to 60 frames per second and support both H.264 and H.265 video standards.

In our last earnings call, we described the innovative "Snap" drone from U.S.-based Vantage Robotics featuring a collapsible magnetic 1.37-pound design with protected rotors and Ambarella-based 4K camera.

During the quarter, we were pleased to learn that the news network CNN has received a first-of-its-kind waiver from the FAA to fly the Snap drone over crowds of people for aerial video capture and news gathering. CNN's successful waiver application was the product of over 2 years of research and testing on the part of CNN and Vantage Robotics.

In summary, we are continuing to invest in leading-edge technology to expand our market opportunities both in current markets and in a broad range of new automotive applications. We are very pleased with our progress in delivering the next-generation of computer vision solutions and look forward to meeting with many of you at the 2018 CES show to demonstrate their capabilities.

I'll now hand over to George for more details on the financials and our guidance for Q4.

Thank you, Fermi, and good afternoon, everyone. Today, I will focus my review on the financial highlights for the third quarter of fiscal 2018 ended October 31, 2017, as well as the financial outlook for Q4 of this fiscal year. During the call, I'll discuss 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 for Q3, we have eliminated stock-based compensation expense adjusted for the impact of taxes.

Our Q3 2018 revenue of $89.1 million represents a decrease of 11.4% from the $100.5 million of revenue in the same period in the prior year. GoPro revenues in the quarter declined to $18.4 million compared to $34.6 million in the same period in the prior year. Q3 non-GoPro revenue grew 7.2% from the previous year, finishing the quarter at $70.6 million.

In Q3, we saw strong year-over-year growth in IP security followed by in-line growth in auto and non-GoPro wearable markets, partially offset by a moderate year-over-year decline in the drone revenues. In IP security, year-over-year growth was led by strong consumer product builds for the North American holiday selling season. Solid revenue increases from the OEM automotive video recorders plus in-line aftermarket shipments resulted in good year-over-year growth in the auto market.

Drone revenues in the quarter were down from last year, most likely due to DJI's low-end Spark drone impacting higher-end drone sales, in addition to continued weakness from Tier 2 consumer drone customers as we discussed last quarter.

Non-GAAP gross margin for Q3 was 64% compared to 63% in the preceding quarter and 66.3% in the third quarter of the prior year. Gross margin improved from the previous quarter due primarily to a better mix of products in IP security, particularly in the consumer home monitoring market, as well as an increase in sports camera revenues.

Non-GAAP operating expenses for our third quarter were $27.2 million compared to $25.4 million for Q2 of this fiscal year and $24.2 million for Q3 of the prior year. OpEx increased from the previous quarter primarily as a result of increased chip development costs, partially offset by a decline in the provision for annual executive bonuses.

Non-GAAP net income for Q3 was $25.7 million or $0.75 per diluted ordinary share compared with non-GAAP net income of $38.4 million or $1.11 per ordinary -- diluted ordinary share for the same period in the previous year.

The non-GAAP effective tax rate in Q3 2018 was 14.6%. In the third quarter, the non-GAAP earnings per diluted ordinary share were based on 34.4 million diluted shares as compared to 34.6 million diluted shares for Q3 of fiscal year 2017.

Total headcount at the end of Q3 was 689 compared to 680 at the end of the previous quarter with about 81% of employees dedicated to engineering. Approximately 71% of the total headcount is located in Asia.

We ended Q3 with cash and marketable securities of $414 million, adding $26.6 million of cash from operations in the quarter. During the quarter, the company repurchased 269,540 shares at an average price of $47.41 for a total consideration of approximately $12.8 million. The total number of shares purchased since the inception of the repurchase programs in June 2016 is 1,433,137 at an average share price of $50 for a total consideration of $71.7 million. As of October 31, 2017, there was approximately $35 million available for repurchase under the repurchase program through June 30, 2018.

Total accounts receivable at the end of Q3 was $47.2 million or 48 days sales outstanding. This compares to accounts receivable of $38.4 million or 49 days sales outstanding at the end of the prior quarter.

Net inventory at the end of Q3 was $21.1 million or 55 days sales -- days of inventory compared to $17.7 million or 63 days of inventory at the end of Q2. Accounts receivable and inventory remain in line with company targets.

We had 3 10%-plus revenue customers in Q3. WT Microelectronics came in at 50.6% of revenues compared to 53.7% in the third quarter of the prior year. GoPro represented 20.7% compared to 30% last year, and Chicony came in at 11.9% this year.

I would now like to discuss the outlook for Q4 of fiscal year 2018. We expect revenues for the fourth quarter ending January 31, 2018 to be between $68 million and $72 million, representing a decrease of between 22.3% and 17.7%, respectively, from Q4 of last year. In the quarter, we expect GoPro revenues to decline to approximately $13 million from $30.2 million in the same period in the prior year. We expect solid year-over-year revenue increases in IP security and auto OEM markets, offset by a moderate decline in the drone market. We continue to expect a shortage of high-speed memory to impact revenues over the next few quarters, particularly in the early part of next fiscal year. Although we have included an estimate of the expected impact of potential memory shortages to our Q4 revenue guidance, it is not yet clear to what extent revenues will be impacted in future quarters.

We estimate Q4 non-GAAP gross margins to be between 62% and 63.5% compared to 64% in Q3 of 2018 and 67.2% in Q4 of the prior year.

As forecasted, we are seeing the increased mix of China professional security revenues having the largest negative impact to margins in the quarter. As compared to the previous year, some of the reduction in margin can also be attributed to the decline in sports camera revenues.

We expect non-GAAP OpEx in Q4 to be between $28.5 million and $30 million, with the increase from Q3 primarily coming from increased chip tape-out fees and engineering headcount increases. We expect the non-GAAP effective tax rate to be approximately 16% in the quarter. We estimate our diluted share count for Q4 to be approximately 34.8 million shares.

At this point, not much has changed as it relates to our expectations for fiscal '19. Non-GoPro revenues are estimated to grow in the range of 10%, which assumes growth in security, auto and wearable markets, including non-GoPro sports cameras, offset by lower revenues in the consumer drone market. Although we expect some revenues from GoPro, we lack clear visibility and therefore are unable to provide guidance at this time. Next year, we expect gross margins to move into the high end of our target margin range of 59% to 62% during the first half, while we estimate OpEx to be between $120 million and $130 million.

I would like to thank everyone for joining our call today. And now I will turn it back to the operator to manage the Q&A session.

(Operator Instructions) Our first question comes from the line of Kevin Cassidy with Stifel.

In the 360-degree camera adoption, what kind of unit volumes are you seeing for that? It's a new category. And maybe can you compare it at all to sports cameras?

Well, Kevin, I think this is still an early stage of the 360-degree. You can see that most of the designs we announced today, they just started shipping. We are conservative on the volume, but I think that we -- it's hard to compare to sports camera because this is a new market. So I would say we are conservative, but we still continue to develop technology for this market because we believe that, given time, this market can be a meaningful business for us in the near future.

Okay. And on the deployment of the e-cameras -- or the e-mirrors, sorry, e-mirrors for automotive, can you give us any -- you said it will be next year. Can you qualify that a little tighter?

Yes. First of all, we announced Gentex, but also this time, we mentioned that we have our other design activities with several OEM and Tier 1s at this point. And we continue to believe that at the second half next year, later next year, we'll have a -- starting seeing revenue from the electronic mirrors.

Our next question comes from the line of Matt Ramsey with Canaccord Genuity.

Fermi, I was pleased to hear that you guys did sample CV1 in the quarter. Maybe you could give us a little bit more detail about, a, which market segment that you were able to sample the product into and maybe a representative set of customers there? And second, how the initial reception of the product, particularly on the software side, was received by the customers that you were able to sample to?

Right. So we -- the 2 focus markets for us is automotive and the IP security, and we just demoed -- and the sample, like I said, demoed a sample to the key customers in those markets. And you can imagine that some of the key customers from those markets that already received -- seeing our demo and received our SDKs. I think it's still early because we just sampled a few weeks ago. It's still very early stage. But I have -- I can say that we see a strong interest in CV1 and our CV technology, particularly due to the -- our unique features, performance, power consumption and, probably more importantly, the road map our CV family we talk about. And in fact, in this time, we also talk about we expect to start sample to our key customers early next year for the next generations CV SoC based on 10-nanometer chips. So I think with that -- those features, performance and the road map, I think a lot of customers are very interested to continue to evaluating. We expect to see some -- hear feedback in several months.

Got it. That's helpful. Just as a follow-up on the automotive market because it seems like one of the very exciting spaces for CV in the long run. Maybe you could comment a little bit, Fermi, on revenue opportunity in the automotive space between where we are today and when the CV family really ramps in the automotive space. I know you have opportunities in e-mirrors and surround view and a few other areas. But I think it would be helpful for us to understand the magnitude of revenue growth over the next calendar year in automotive even before the CV contributes materially to the sales pipe.

It's hard for me to give you a revenue growth forecast, but like I can answer that question in a different way. When we look at the total available market for us, our current market, including professional security and the connected home market, we are talking about roughly $1.6 billion based on the public -- research number that published by other firms. But if you add auto including electronic mirrors, surround view, recorders, ADAS and autonomous car, in several years, our total available market size can grow up to $5 billion. So basically, that's where we look at our potential growth from, and that's how we decide that we're going to continue to invest heavily on the security camera as well as auto because they've become 2 biggest market opportunities for us. And in terms of the near-term revenue, it's really about video recorder and e-mirrors because those 2 things we have a lot of activity. Like today, we talk about Honda and also, on e-mirrors, we talk about we have several design wins that we're working on right now. And so nearer term, I think that recorders will continue to be strong in Japan and in China, and e-mirror is an international opportunity for us. And long term -- longer term, I want to mention that ADAS becomes a road map for e-mirrors, for recorders, for surround views. It's not a separate product anymore. We continue to see that the front camera-based ADAS system will continue, but all of the current video system for human vision, like e-mirrors, recorders and surround view, will add ADAS function into it as a road map. So I think we are very well positioned to serve the human vision side of video products for automotive customers and use that as a foundation to [put] our computer vision technology into those the same customers in the future. And it's definitely our ambition to continue to be a supplier (inaudible) ADAS and autonomous car in the future, and that's why we develop CV and CV families.

Our next question comes from the line of Charlie Anderson with Dougherty & Company.

Fermi, you referenced in your script sort of increasing requirements, if I heard it right, for IP security cameras for computer vision. So I wonder if maybe you could describe that a little bit more in terms of the content available to you. I imagine there's different features that someone versus others -- other prices established in the market in terms of what people will pay for CV. Any commentary there would be helpful, and then I've got a follow-up.

Right. So I really think that what we heard is that, first of all, CV is coming to security cameras both for the professional and the connected home. But I think everybody is looking for a single-chip solution that can do very good video processing that we have today and a very good computer vision, and there's no such products today in the market. And a lot of people, our customers are using 2-chip solutions or using servers to do computer vision at this point. So I really think that our focus is going to be trying to provide most price efficient but at the same time powerful in terms of the performance of features on the video processing computer vision on a single-chip and to leverage our current position with the current IP security customers. And then that will add our ability to compete in this market. But more importantly, with any chip like with that kind of performance and features, we expect that the ASP will be higher. Gross margin will be -- is meant to be same place on the competition, but we definitely expect that there's upside on the ASP side for us.

Great. And then as a follow-up, George, I was wondering, I think you sort of spelled out 10% non-GoPro next -- growth next year. I think you previously talked about 10% to 15%. Is that delta the difference from the memory shortages that extend in the first half of next year and if you can maybe quantify that? And also, did they have an impact in Q3 and in your Q4 guidance, to what extent?

Yes, first of all, I think it's the range of 10%. So it could be above that also. I didn't mean to make it a significant difference, but there are, I think, a couple variables next year. First of all, the transition to the new rev rec next year could have an impact in the early part of the year to our reported revenue. In addition, in the early part there is the exposure to the high-speed memory. And I also think, at this point, we're being very cautious on the consumer drone market until we see how the rest of the holiday season progresses.

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

George, just wondering if you could give us any more thoughts on GoPro. I think a couple of quarters ago you had thought that, that revenue was going to tail off. With the release of the HERO6, it looked like the GoPro revenue has been stronger here in the third quarter and the guide for the fourth quarter still has a healthy amount of GoPro. As you look into fiscal '19, do you see sort of a continued revenue from GoPro cameras? Or does the second half revenue kind of represent somewhat of a last time buy for that GoPro business? How should we be thinking about GoPro next year?

No. I don't believe the last -- Q4 will be the last time buy. I believe there will be additional revenues next year. It's difficult at this point to determine the actual quantity and timing. We believe there'll be very little in the early part of the year due to inventory levels coming out of the holiday season. But we do believe after that, that we'll see some revenues particularly, probably, as Q2 and early Q3 and then wind down.

And is that for the Fusion? Or are there other products that you're seeing in the GoPro?

Well, we understand the HERO5 will wind down next year at some point, so it will be the Fusion and the HERO5 at this point.

Great. And then just coming back to the CV family, I think in the past you had talked about applications also in autonomous drones. It sounds like you're currently limiting sampling to the automotive and security markets. Do you expect to sample into the drone market at some point in the future? Or are there reasons why you may be sort of deemphasizing CV family for drones going forward?

No. In fact, we do plan to sample to the drone market. In fact, you should come to CES to see our demos. But at the same time, the whole drone market has one big customer, which is DJI. So that's where we need to focus on. And we have been talking about that DJI has been well known that they are doing their own silicon development. So how to sample, how to deal with that is going to be an important strategy move for us. And that's why we never talked about this, but I think DJI continues to be a very important customer for us. We're definitely going to show them our technology and continue to build relationship from there.

I think one thing on that is just priority here of rolling out the software for each market. We see a big -- a much bigger demand in the security market than we do with the drone. So there's just a bit of timing also on getting the resources allocated.

Got it. And then just, Fermi, I think in response to Charlie's question on the security market, you sort of referenced the need for single-chip CV plus video processors. Can we assume that the 10-nanometer version of the CV family member that you're [taping] out first half next year, will that merge both the traditional computer vision plus the video encoding? Or is it too early to talk about features for that second-generation solution?

No, I think that's a good assumption.

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

I wanted to first hit on the CV side of things and talk about the existing markets, both the IP security, probably primarily, and then, to a lesser extent, the drone side. As you guys think about the feature set that you're going to bring to it, you've talked many times about the ASP being higher. But the value proposition, do you think that's going to be something that really allows you to gain share? Is it going to expand the market because the features that the drone or the IP security cameras can drive? So really, is it a share gain story for you guys? Or is it a powerful enough technology that you could actually expedite the growth rate of the industry you're addressing itself?

My belief is -- let's take the different markets one at a time. For security camera, I think CV plays a major role to really reshape this market. And I do believe that CV will increase the size, the total available market for security cameras both for professional and the connected home. So from that point of view, I think the market will grow. Not to mention that we also believe there will be a refresh cycle coming up when the CV becomes at a reasonable price and a good feature and performance. So that refresh cycle plus the growth of the market is definitely going to help us. Plus, I think the ASP definitely go up because of the function features. For drones, consumer drone, I think it's -- we don't see -- continue to have a strong growth in that market because the regulations and also the price point. But I do believe when the function and performance and the features become available and the regulations become eased for the professional commercial drones, that will boost the total available market for us. So I really think that for consumer drone, we need to focus on maybe a couple of important customers, and DJI obviously #1, and also for commercial drones in the future. So we are doing technology. Fortunately, the technology that we build for CV definitely directly applies to drone also. We talk about stereo processing. We talk about object detection, about planning -- route planning. They're all technology directly usable for the current drone market. So I really think that it's really about how we allocate resources and building software for the near-term biggest opportunity. That's why we continue to talk about security and automotive. But drone, we never say -- we're definitely still keeping our radar, and we want to get back there at the right point.

Great. That's helpful. And then for my follow-up, I just wanted to hit on gross margin with a 2-part question, both of them for you, George. Talk a little bit about -- clarify a little bit and add some color to the first half commentary about hitting the high end of your 59% to 62% range. What are the pluses and minuses that gets you there? And kind of any color in the back half of the year? And then the second part of the question is I know the ASPs are higher with the CV chip that you talked about. But do you expect gross margins to be accretive or dilutive in the CV family in general?

Well, I think in the first half we'll start to see the impact of the China business increasing and the sports camera decline. So that's why I'm forecasting that to drift into the high end. As far as the second half is concerned, right now it's actually looking, what I would call, flat. No real degradation because I think the mix forecast at this point is not [unsimilar] to the first half. So I'm reasonably comfortable with that. On the CV side, I'm not expecting much impact to margins at all this year. I think the volumes would be not really material to the overall margins. I think going forward, I believe they'll be accretive particularly in certain markets like security and, potentially, drones. I think we still have to sort out the various opportunities we have in auto because there'll be different competitive environments, for instance, in electronic mirrors versus surround view and, later on, ADAS. So I think those margins are still open for analysis in the future.

Our next question comes from the line of Jagadish Iyer with Summit Redstone.

Two questions for me, first on the CV chip, particularly you talked about the applications related to automotive. What is the realistic time line given the long qualification cycle? And given that you are sampling currently with your existing products, when do you think that realistically you can get revenues from the CV chip for the automotive market? And then I have a follow-up.

Right. So for the automotive market, like you said, design cycle is long, but I think there are 2 portions of it. In China, the design cycle is shorter than other places. So we have seen that. So -- but you are right. If you -- let's say -- let's take and say electronic mirror that we have designs already and we talk about revenue, non-CV electronic mirrors revenue starting late next year, we believe that the next generation of electronic mirror design will have CV in there and that could be 2 years after that in China, maybe even longer for that. So you're absolutely right. Automotive CV is a low end -- from a revenue point of view is low end for us. But however, all the designs we need to be win today for the 3-year rate of revenue. So that's why you have to start working on those designs today. But at the same time, I -- we definitely realize that by working with Asian vendors, especially in China, we do see a near-term opportunity there.

Okay. Then as a follow-up I have, if you look at the home -- consumer home monitoring market, I was wondering if you have any thoughts on how big is it going -- it's going to be and kind of a growth rate. And how should we think about competition here?

First of all, I think that market has continued to grow. We have been talking about the consumer market has been growing for 2 years now. We continue to see strong growth. It's hard for me to quote a growth rate because we haven't seen any market research on that. But based on our current customer feedback, we think that all the customers are seeing strong growth on their revenue and their products. So I optimistically believe that this market will continue to grow for another couple years. In terms of how big it is, again, there's no really market research for that. But if you look at the popular camera vendors are selling multiple cameras into the home today. And if you're considering how much -- how many homes are in U.S. and the penetration rate, I think we're still in probably the early phases of the connected home market, and we expect that this market can continue to grow maybe 2 to 3 years following. Especially when CV becomes a reality for connected home cameras, that will, again, boost the usage because it becomes a lot more convenient and user-friendly for the users of those cameras.

So is there a dollar number that you can kind of throw out there for the home monitoring market just to give us some kind of a context like whether it's $20 a home, or some kind of a number that we can take and then go from there?

Well, our chip ASP is high single digit, mid- to high single digit in that range for each camera. And if you say the 2 to 3 -- you have to do your own calculation, but that's what -- where the ASP is for us for those cameras.

And our next question comes from the line of Joe Moore with Morgan Stanley.

I wonder -- it sounded like GoPro was stronger than your expectations or at least looks that way in the January quarter, and I guess that's kind of surprising in the sense of them talking about HERO5 not selling as well. So I guess, a, is that the case? And do you understand why? And then b, does that mean that there's some other part of your business in the January quarter that's kind of maybe less than you expected that that's offsetting?

Actually, the GoPro business is pretty much in line with what we expected just because of the order processing timing. And that was why my comment earlier that we expect the early part of next year to be slow for GoPro due to probably excess inventory similar to what we've seen the last 2 years.

Okay. Great. That makes sense. And then I guess with regards to this longer automotive time line on the CV chips, I'm kind of curious, when you sort of look at a business where you're iterating the chips multiple times a year, you're putting out multiple generations and you have these long design cycles, how does that work? Is CV1 sort of vying for a model year several years out type of wins? Or is CV1 to sort of established a development platform so that they can evaluate the future iterations of the chip?

Right. So I think -- let me give you an example. For example, when we start doing electronic mirrors, we use our consumer chip called A9 to demo the function and features and powers and to get people interested and to start moving on the designs. And we secure design wins, too, by showing them auto-grade chip for A9AQ. So that's the process we're going through, and we believe we need to win through similar things in the auto side, so that CV1 will be the evaluation platform for them, but we need to follow up with another chip which is auto grade for them to go into production. But without CV1, look, if there's no chip that can evaluate your performance and your claims, you won't be able to even get the discussion with them. So that's the cycle we're going through. But however, the 2 to 3 chip a year is not all for auto. You can imagine that for the IP security camera or for the home security, the performance requirements for CV is much less than auto. So you need a different chip for that. So you can imagine that we start doing chips for different markets, one target for auto, one target for the security and maybe there's other things in between. So we are developing a family become targeting particular applications. I think that's a requirement for the -- continue to play in multiple different markets.

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

In the fiscal '19 guidance for 10% year-over-year, can you talk about whether you think the IP security market or the auto market would grow faster on a relative basis and the puts and takes there?

Well, right now I think the auto market is probably going to be our fastest-growing market next year just on a percentage basis, but it is off a smaller base. The next growth area is probably going to be the home security market. So I think that will, again, be very strong next year. In professional security, I think it will grow next year but probably more in line with market because you're going to have ASP erosion there, unit increase and ASP erosion.

Okay. That's helpful. And then the auto -- just to clarify, the auto revenues for fiscal '19, are you assuming any CV revenue in there or not? And when will you be in a position to announce customer wins for CV?

Right now there's no CV revenue in the auto market forecast. And I would say in CV, probably the second half, as we've talked about, we'll be able to talk about customer relationships. Potentially or hopefully before that, we might be able to talk in the auto space about relationships with CV.

Okay. That's very helpful. And maybe one quick question on the security market. What's the mix now of your revenues of professional versus consumer? Has consumer come up the curve where it's half-half? Or it's still small relative to the professional bucket?

Yes, it's still less than half, but it is expanding. It's well over 1/3 of the security -- total security now.

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

My question is a follow-up on the IP security market kind of assumptions built in for fiscal '19. Just following up on the previous question, I'm wondering if you could tell -- give us the thought process on how you think about China versus the rest of the world, whether there's any sort of realistic or noticeable bump coming from HEVC outside of China, and then any thoughts on relative share.

I still think that the Chinese market is growing faster than the international sales and also that Hikvision (inaudible) continue to be very, very strong. So I don't think -- I don't expect that to change. In terms of HEVC, that was a Chinese product based syndrome. But hardware, like we said, Bosch starts shipping HEVC products this quarter, and we also knew that a few other international system companies will follow with their HEVC offerings. So HEVC will become quite the international standard for security camera. However, to our knowledge, that royalty situation with HEVC is not -- has not been resolved yet. So that's still a big question for our customer, how to handle it. But nonetheless, we start seeing because Hikvision (inaudible) has been aggressively pushing HEVC and everybody else is following up.

Yes, just one other comment on the overall market. We do -- we did see, I think, strong revenues in professional security outside of China this year also. So that did contribute to some of the growth, and we are expecting, next year, to see the same.

And that concludes our question-and-answer session for today. I'd like to turn the floor back over to Dr. Fermi Wang for any closing comments.

Thank you, and I would like to thank all of our colleagues for their outstanding job to deliver computer vision technology on time, and also thank you, all of you, for joining us this afternoon. I'll see you next time. Bye.

Thank you. Ladies and gentlemen, thank you for your participation in today's conference. This does conclude the program, and you may now disconnect. Everyone, have a great day.