Ambarella Inc (AMBA) 2017 Q4 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Ambarella fourth-quarter and fiscal-year 2017 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 and welcome to Ambarella's fourth fiscal quarter 2017 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 2017 fourth-quarter and full-year results.

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 that we file with the SEC, including the annual report on Form 10-K that we filed on March 25, 2016, for the 2016 fiscal year and quarterly report on Form 10-Q for our third quarter of fiscal 2017 that we filed on December 9, 2016.

Access to our fourth-quarter results press release, historical results, SEC filings, and a replay of today's call can all be found on the Investor Relations portion of our website.

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

Thank you and good afternoon, everyone.

I am pleased to report that our Q4 fiscal-year 2017 revenue was $87.5 million, representing an increase of 28.8% from the $68 million of revenue in the same period of fiscal-year 2016.

For the full year ending January 31, 2017, our revenue declined 1.9% to $310.3 million.

During the year, we continued to make progress in diversifying our markets and the customer base.

We enjoyed significant revenue growth from fast-growing home monitoring and the drone markets and the built consumer momentum in new OEM automotive applications, such as video recorders.

We also entered full mass production with our first 14-nanometer SoCs and continued to raise the performance bar with the introduction of full 4Kp60 Ultra HD SoCs in our first automotive-qualified solutions.

A the recent CES 2017 show, Ambarella introduced the widest range of new products in the Company's history including the A9AQ automotive SoC family for electronic mirror and surround-view applications.

H3, the industry's first 8-K video SoC and the H22 4K SoC family for drones, VR, and wearable cameras.

For Ambarella, A9AQ is an automotive AECQ 100-certified 4K single-chip solution for single- or multi-camera electronic mirror surround view and video recording systems.

Its LED flicker mitigation technology dramatically reduces artifacts introduced by LED headlights and traffic signs, while its multi-exposure HDR processing provides a superior visibility in low-light or high-contrast conditions.

The A9AQ's powerful processing performance enables customer applications, including advanced driver assistance features, user interface, and wireless networking.

Ambarella also introduced its B6 family of 4K Ultra HD capable serializer/deserializer or "SERDES" companion chips to enable high-performance multi-camera applications based on the A9AQ SoC.

At CES, Gentex, the automotive industry's leading supplier of rear vision solutions, demonstrated its new A9AQ-based hybrid three-camera system that streams video to a rearview mirror integrated display.

Gentex partnered with Ambarella on the project because of the A9AQ's support for high-quality video in multiple viewing modes, providing automaker customers with a sophisticated, customizable, high-performing camera monitoring system.

Also at CES, we introduced the industry's first 8K camera SoC, H3, targeting high-end drones, as well as new generation of video cameras, including 360-degree and virtual reality cameras.

H3 video SoC supports 8K Ultra HD H.264 video at 30 frames per second and also enables 360-degree cameras with dual 4Kp60 video capture.

It includes 10-bit HEVC high dynamic range video processing and features Ambarella's advanced electronic image stabilization to generate smooth 4Kp60 video output under challenging conditions, including during high wind and with high degrees of rotation.

We also introduced the H22 family of 4K SoCs, bringing 4K Ultra HD video to small form-factor forms, including portable "selfie" models, wearables, and the VR cameras.

The new H22 SoC is capable of including 4K Ultra HD video at 60 frames per second, using the H.265 or H.264 video formats, and it supports advanced electronic image stabilization up to 4K resolution.

The on-chip 1.2 gigahertz quad core ARM Cortex A53 CPU provides significant processing power for customer applications, including drone flight control of advanced analytics, video stitching, and wireless networking.

Fabricated in 14-nanometer process technology, the H22 offers very low power consumption, enabling 4K cameras with small form factors and extended battery life.

In addition to continuing Ambarella's leadership in high-quality, low-power video and image processing, we are seeing opportunities for further differentiation as computer vision capabilities are required in IP security, drones, and automotive applications.

In OEM automotive markets, the requirements for advanced driver assistance systems, or ADAS, represents a significant opportunity and it requires a combination of advanced image processing and powerful computer vision processing.

To address these opportunities, Ambarella has been investing in the development of advanced computer vision technology and in early February taped out our first computer vision chip, our third chip to be based on 14-nanometer process technology.

We are extremely excited by the new applications it will enable, especially as we combine the new chip with advanced algorithms and software, leveraging the automotive experience of our VisLab team.

We will continue to further invest in computer vision technology, automotive solutions, and advanced 10-nanometer process nodes, as we believe these efforts will provide us with significant additional opportunities for success in the future.

Now I'd like to update you on a few of the customer highlights from the last quarter.

In the home monitoring market, we are continuing to see significant growth, with new product categories being created as cameras are integrated into home alarm systems, baby monitors, and outdoor lights.

Ambarella's advantages in terms of video quality, low light imaging, low power, and smart video analytics are enabling the creation of innovative new products in each category.

During CES, NETGEAR introduced its Arlo Baby, a new baby monitor based on Ambarella S2Lm SoC.

The Arlo Baby is an all-in-one smart baby monitoring camera with a full 1080p HD video night light, rechargeable battery, and air sensors.

It includes smart analytics and is able to distinguish a baby crying from other sounds, and it will notify parents only when interesting motion has occurred.

Also at CES, Ring announced its Floodlight Cam, a motion-activated security camera with built-in floodlights, siren alarm, and two-way audio.

Based on S2lM, the Floodlight Cam provides smart object and facial detection, customizable motion zones, and smart room panning, while being controlled remotely from the Ring smartphone applications.

In the professional IP security camera market, Ambarella continued to see increased adoption of its latest H.265 camera SoCs, primarily in China.

During the quarter, HIKVision, the world's largest security camera supplier, started shipping its new value-added series of H.265 camera based on Ambarella's S3LM IP camera SoC family.

In the wearable sports camera category, Chinese camera maker YI Technology announced its newest product, the YI 4K+ Action Camera, the first action camera to offer 4Kp60 video recording, priced at $299.

In addition to 60 frames per second 4K video recording, the camera includes electronic image stabilization, live video streaming, and extended battery factory life.

This is the first action camera to be based on the Ambarella H2 SoC processor.

In the automotive market, Ambarella has traditionally provided camera SoC solutions for video camera recorders or dash cams for aftermarket sales.

Additionally, we are continuing to see accelerating adoption of camera video recorders by automotive OEMs, both as original equipment and as dealer-fit options, primarily for the China and the Japan markets.

During the quarter, Japanese automotive giant Nissan introduced its Full HD video recorder based on Ambarella's A12A automotive camera SoC.

The camera includes a four-megapixel sensor, high dynamic range processing, and the integration of video recording with navigation display and parking functions.

In the automotive retail market, US-based startup Waylens introduced its "Horizon" 1080p60 automotive camera system.

The Horizon system allows drivers to easily capture, edit, and share video in real time with its mobile app.

Based on Ambarella's S2 SoC, the camera also allows drivers to record their vehicle's acceleration, performance, and to compete with other drivers by sharing performance data remotely.

In summary, as we look forward to 2017, we see continued growth in IT security, home monitoring, drones, and automotive camera markets.

We believe that our continuing investment in computer vision technology, automotive solutions, and advanced process nodes will all further expand the opportunities available to us.

Now I will hand it over to George for more details on the financials and our guidance for Q1 and the fiscal-year 2018.

Thank you, Fermi, and good afternoon, everyone.

Today, I will focus my review on the financial highlights for the fourth quarter of fiscal 2017 ended January 31, 2017, and then review the financial outlook for Q1 of fiscal-year 2018 that ends on April 30, 2017.

In addition, I will discuss our preliminary expectations for the full year ended January 31, 2018.

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 Q4 2017.

For non-GAAP reporting for Q4, we have eliminated stock-based compensation expense adjusted for the impact of taxes.

Our Q4 2017 revenue of $87.5 million represents an increase of 28.8% from the $68 million of revenue in the same period for the prior year.

In the fourth quarter, wearable cameras demonstrated strong year-over-year growth, led by GoPro.

GoPro's Q4 revenues increased from approximately 18% of revenue in the prior year to 34% this year.

Excluding GoPro revenues, the other camera markets grew 7% year over year, offset by a 63% or $2.1 million decline in the infrastructure revenues.

We saw increases in year-over-year revenues in the auto, drone, and nonsports wearable markets in the quarter.

Year-over-year growth in security was flat, with solid consumer IP security growth offset by a decline in the professional market.

In addition, the infrastructure market revenues were down substantially in the quarter, compared to the prior year due, to difficult comparable numbers for the prior year and continued decline of the older H.264 encoder market.

For the year ended January 31, 2017, revenues declined 1.9% to $310.3 million.

GoPro total revenues, including sales directly to GoPro and their ODMs, is estimated to have declined from approximately 30% of revenue in the prior year to 24% in the fiscal year-end 2017.

Non-GoPro camera revenues increased by approximately 8.4%, partially offset by a decline in the infrastructure revenues of 27%.

Non-GAAP gross margin for Q4 was 67.2%, compared to 66.3% in the immediately preceding quarter and 64.1% in the fourth quarter of the prior year.

Gross margins reflected the benefit of approximately $850,000 from the recovery and sale of inventory previously written down due to prior-period yield loss in the manufacturing process.

We don't expect any material amounts of inventory recovery to impact Q1 margins at this time.

In addition, we benefited from a strong mix of products across all markets, especially in professional security, generating margin upside.

Non-GAAP operating expenses for fourth quarter were $25.4 million, compared to $24.2 million for Q3 of 2017 and $22.9 million for Q4 the prior year.

Q4 OpEx increased from the previous quarter, due primarily to higher compensation costs related to additional headcount and adjustments to year-end bonuses.

Non-GAAP net income for Q4 was $32 million or $0.92 per diluted ordinary share, compared with non-GAAP net income of $21.6 million or $0.64 per diluted ordinary share for the same period in the previous year.

The non-GAAP effective tax rate in Q4 2017 was 4.9%.

In the fourth quarter, the non-GAAP earnings per diluted ordinary share are based on 34.6 million diluted shares, as compared to 33.7 million diluted shares for Q4 of fiscal-year 2016.

Total headcount at the end of Q4 was 669, compared to 658 at the end of the previous quarter, with about 82% of employees dedicated to engineering.

Approximately 72% of our total headcount is located in Asia, primarily in Taiwan and China.

We ended Q4 with cash and marketable securities of $405.4 million, adding $32.3 million of cash from operations in the quarter.

Total accounts receivable at the end of Q4 was $38.6 million or 41 days sales outstanding.

This compares to accounts receivable of $41.4 million or 37 days sales outstanding at the end of the prior quarter.

Net inventory at the end of Q4 was $20.1 million or about 68 days of inventory, compared to $23.3 million or 59 days of inventory at the end of Q3.

Accounts Receivable and inventory remain in line with Company targets.

WT Microelectronics, our Asia logistics supplier, represented approximately 48% of revenue in the quarter, compared to 63% for the same period in the previous year.

During the quarter, we continued to ship a significant quantity of chips directly to GoPro in support of the launch of its HERO5 products.

As a result, sales directly to GoPro represented approximately 32% of our revenue in Q4 of fiscal 2017, compared to 1% for the same period in the prior year.

To avoid any misunderstanding, please note that revenues generated from GoPro in both years include only direct shipments and do not include chips shipped to GoPro's various ODMs, either directly or through WT Microelectronics, for production of their camera products.

I would now like to discuss the outlook for Q1 of fiscal-year 2018.

We expect revenues for our first quarter ending April 30, 2017, to be between $62.5 million and $64.5 million, representing an increase of between 9.3% and 12.8%, respectively, in Q1 of last year.

We are anticipating strong year-over-year revenue growth in IP security, both professional and consumer, as well as solid growth in automotive and nonsports wearables, partially offset by a decline in the drone revenues.

The decline in the drone revenue is due primarily to difficult comparisons from the prior year, resulting from timing of new product launches.

High inventory levels of our solution at GoPro and their ODMs at the end of the prior quarter will significantly impact GoPro revenue in Q1.

Non-GoPro revenues in Q1 of fiscal-year 2018 are expected to grow between 13.1% and 16.8%, excluding estimated revenues to GoPro and their ODMs in the quarter of $3.4 million in fiscal-year 2017 and $1.7 million for fiscal-year 2018, respectively.

We estimate Q1 non-GAAP gross margins to be between 63% and 64.5%, compared to 67.2% in Q4 of 2017 and 64.6% in Q1 of the prior year.

As I mentioned earlier, we are expecting a minimal additional gross margin impact in Q1 from the recovery of previously written-off inventory.

In addition, we are anticipating the mix of China security revenues will be dilutive to margin in the quarter.

We expect non-GAAP OpEx in the first quarter to be between $26 million and $27.5 million, with the increase from Q4 primarily coming from increased headcount costs associated with new chip development.

We expect the non-GAAP effective tax rate to be approximately 11.5% in the quarter and estimate our diluted share count for Q1 to be approximately 34.7 million shares.

For fiscal-year 2018, we are expecting revenues to be up 3% to down 3% as compared to fiscal-year 2017.

The current guidance assumes growth in IP security, drones, nonsports wearable, and automotive market revenues, as well as the early ramp of revenues in the new virtual reality camera market.

For wearable sports cameras, the estimate assumes GoPro revenues, including shipments to their ODMs, decline to between 3% and 6% of total fiscal-year 2018 revenues, from approximately 24% in fiscal-year 2017.

The lower revenues from GoPro assume first-half weakness due to excess inventory at GoPro and their ODMs of Ambarella chips exiting fiscal-year 2017.

In addition, based on our current level of interaction with GoPro on product development activities, we believe GoPro is incorporating a competitive chip for at least one new mainstream camera product, which will negatively impact our revenues in the second half of the year.

Excluding sales to GoPro and their ODMs, fiscal-year 2018 revenues are estimated to grow between 20% and 32%, based on the annual guidance I've given.

We feel our estimate of GoPro revenues for the next year takes into account the most likely shipment scenarios for fiscal-year 2018.

For fiscal-year 2018, gross margin should move into the high end of our target margin range of 59% to 62%, while OpEx should increase 12% to 14% over fiscal-year 2017.

The increase in OpEx is directly related to expansion of our development efforts in support of computer vision technology for all markets, as well as substantial investment targeted at the auto OEM markets.

We understand this additional investment is at the expense of the bottom-line growth that is essential to the long-term expansion of Company revenues and markets.

No shares were repurchased by the Company in Q4 under the $75 million share repurchase program that we initially announced in June 2016.

Approximately $55 million remains available for repurchase of shares under the program through June 30, 2017.

We will continue to evaluate the opportunity to repurchase shares for the remainder of the purchase period.

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).

Kevin Cassidy, Stifel.

Thank you.

Thanks and congratulations on the strong quarter.

I wonder if you could spend a little more time on the drone revenue.

Was that what the upside was in the fourth quarter and you're saying it's some of the downside on the first quarter?

Can you say a little more details of the dynamics?

I think if you look at drone revenue, as I've said recently, it's pretty volatile.

It's based on product launches.

When I look year over year in Q1, last year we were delivering a lot of inventory into customers for product launches which happened either late Q1 or into Q2, so therefore we had a pretty strong Q1 and Q2 last year.

So I think a lot of year-over-year Q1 relates to tough comparables the last year.

Q4, though, if you look backwards in Q4, it was pretty much in line with our expectations.

It was up a little bit over the prior year.

As you know, there was a couple launches during the second half of last year which helped contribute to that.

Kevin, this is Fermi.

I think also George mentioned in his script that, if you look at an annual basis, we still expect the drone is going to grow from last year to this year.

Okay.

Maybe just to expand on that, are you seeing much activity in commercial drones?

No, in fact we don't.

In fact, we continue to see a lot of activity in the consumer drone and I think that we are still a major supplier to all of the consumer drones in the market today.

We haven't seen much on the commercial side yet.

Okay.

And just one other question, George, I had concern on some consumer inventories coming into the beginning of the year.

I know clearly GoPro was going to be a problem.

What's your view on the home security camera inventory levels?

I think from what we see with all our major customers particularly, it's fine, both inventory at the manufacturers and in the channel.

Charlie Anderson, Dougherty & Company.

Thank you for taking my questions.

George, I wondered if it's possible to be a little bit more specific about the annual guidance in terms of the revenue percent by segment.

You gave GoPro.

I wonder if you could offer anything more on the other statements or maybe if you want to rank order the growth by some of the areas, that would be helpful, and then I've got a follow-up.

Yes, we've done a lot of work over the last month or so trying to develop numbers for GoPro, particularly in the prior years, in order to really be able to break out GoPro going forward.

It's a lot more difficult with the other markets.

I can give you some sort of general direction of the markets.

Probably the highest dollar growth should come from our traditional markets this year, like IP security, and that's both professional and home.

Drones, nonsports wearable should be strong for us this year.

We should see a nice contribution from the new markets, like virtual reality and hopefully the auto OEMs segment, but these are off a relatively small base and they are new markets, so it's a little more difficult to give exact numbers.

In addition, we do expect to see growth in our non-GoPro sports revenues this year.

I don't know if I caught drones there.

What was the outlook on drones for you, George?

That was in one of our higher dollar growth areas, so I said it at the beginning.

So that should be good growth for the year, but, again, I just want to emphasize it's going to be volatile on a quarterly basis, but we still feel pretty good this year about the drone growth.

Perfect.

And then as it relates to GoPro and this competitive solution, it is you guys' understanding that this is a unique situation with them or is it a new merchant competitor that could potentially compete in other sockets that you guys have currently?

Just any color there would be helpful.

In fact, from a competitive landscape point of view, I do not believe that has changed in the last three months.

In fact, Hisilicon remains to be our largest competitor, and with other sports camera customers or in any other markets, we do not see a change in the competitive landscape.

Okay.

Great.

Thanks so much.

Quinn Bolton, Needham & Company.

Trying to follow up on Charlie's question, and obviously there will be a lot of concern about this second sourcing at GoPro, so can you just talk to us a little bit about is this a pure video processor, a sort of C company?

Is this a company that might have 4G LTE capability as part of their solution that you guys obviously don't supply?

That's the first question.

And then, the second question is to the extent that this competitive solution was selected by GoPro, what's the risk that it could be adopted by some of your security camera or drone customers?

Again, is there a feature that was particularly attractive in wearables that might have allowed for the selection that would not be applicable to some of these other markets?

In terms of the specifics of that solution, I think that should be a question for GoPro.

However, we do believe that it's very unlikely that an app processor type of chip can be used in a GoPro form factor, due to the power consumption limitation that we talked about in the past, so I really don't think that's a processor type of chip.

However, you talk about whether that chip can be used in any market, like I just said in a previous question, we have not seen the competitive landscape change in any other market or with a customer in the sports camera market.

So Fermi, reading between the lines, does this sounds like it's an ASIC solution that was selected by GoPro that, because it's an ASIC, would not have applicability to some of the other markets?

Quinn, I think that's really a good question for GoPro.

We are not at liberty to talk about it.

I think, Quinn, just to follow up, from our perspective we really haven't seen any change in the other markets or customer environment from a competitive standpoint.

I think that's the key to the answer, really, as far as we probably want to go with it.

I understand.

I appreciate the sensitivity, so I do appreciate the additional color.

I wanted to move over -- this is my second follow-up question on the automotive market.

I know that the sort of aftermarket has been the vast majority of automotive revenues historically, but over the last few quarters you've announced OEM design wins with Nissan, with Toyota, with Mercedes.

Is the OEM or dealer install segment of automotive, is that becoming a meaningful percentage of total automotive revenue or is that still, say, in the single-digit percentage of the overall automotive bucket?

Although we are being very pleasantly surprised in the number of manufacturers adopting our products and how fast they grow, the revenue grows in the past, that's still off a very small base, so I don't think that's significant revenue, but we do believe that if the trend continues, it will become significant revenue in the future.

I think you can assume it's -- last year was obviously in the low single digits, and I would expect it to at least move up quite a bit and maybe even reach some of the high single digits, maybe high to mid single digits as a percent of revenue next year on its own, and then you would add in the retail aftermarket products.

So, overall, the auto space should grow nicely next year.

And sorry, just a clarification there, George.

The OEM segment alone would be mid to high single digits, so if revenues are just a little over $300 million, we're talking $20 million or so just direct to automotive OEMs, or is that high single-digit percentage of the overall automotive revenue for the Company?

I would say closer to mid to high single digits, so somewhere in between that, of total revenue is for the auto OEM business, so it's coming up nicely next year.

At least that's our early indications.

It is early on in this market and there's a lot of people adopting it, but we don't have a lot of marketing history yet, so I think over the next couple of quarters we'll be able to become more specific on what we think there.

Appreciate it.

Thank you.

Jagadish Iyer, Summit Redstone.

Thanks for taking my question.

You had showcased about the rearview and the side view mirrors at CES.

I just was wondering in terms of how much more traction have you had and how should we be thinking about the dollar opportunity here as far as volumes, if some of the auto OEMs start to embrace this approach towards viewing through the rearview as well as the side view?

And then, I have a follow-up.

So from a market point of view, we started talking to a potential customer several years ago about this, and then we started talking about it publicly probably 12 to 18 months ago.

And at that time, it was basically a brand new idea.

However, at this point, I think we got a lot of feedback from the customer that this camera-based rearview mirror will happen, and in fact, we see a lot of design win activities and there are a lot of people trying to issue RFQs.

So, although we cannot talk about the potential customers yet, but I think from a market point of view we do see that develop quite well in the last 12 months.

In terms of the revenue, we've continued to think it's a 2018 type revenue, and it's just because any auto OEM deal will take time because of the qualification process.

You see that we announced our A9AQ chip and you'll see we announced our B-6 ship, which is a SerDes for that market, so our solution is ready.

It's really how fast our Tier 1 and ODM customer can take that into production.

I think that's the question.

Fair enough.

And then the follow-up question is on the consumer security camera.

You had highlighted some, again, at the CES show.

I was just wondering what kind of traction have you had, because given the attractive price points and the capability that it has in terms of the software and things like that, so I was just wondering what kind of momentum are you seeing particularly in the US?

In the US, we continue to see strong growth opportunity for us and we continue to have design wins in that market.

You can see that.

We just talked about Nike, as well as Ring, in today's scripts and among other things we continue to see our other customers doing well in this market.

So we are optimistic about this market and that the percentage of growth will be significant, but, however, because it's still off a smaller revenue base as last year.

So from a dollar-sign point of view, it's still small.

However, the momentum is there and we believe this market will continue to develop in two directions, like we said before, both in the retail side, as well as on the service provider side.

Ross Seymore, Deutsche Bank.

Hi, guys.

Thanks for the color that you gave on the second sourcing.

I know that's a sensitive topic, so let me take a single customer out of the equation and just ask how you approach the performance versus price dynamic.

You have a significantly higher gross margin than most people expect the Company to be able to achieve that's playing in the market you're in, and that's an admirable trait.

When you lead in performance, which you clearly do, how do you determine whether keeping a socket but lowering the gross margin is the trade-off that you'd like to make versus letting that socket go because the gross margin pressure is getting too steep or what have you?

I guess that trade-off between those dynamics, can you talk a little bit about how you strategically focus on that?

I think in the past we talked about two things.

When we manage a company, we are managing on two things.

One is revenue growth; the other one is gross margin.

So, like you say, there are some times that these two conflict with each other.

However, like we showed in the past, even when the gross margin suffered in certain markets, like our low-end IP security camera, we still compete there, so we -- our trade-off is always trying to grow the revenue as much as possible and we are in a market that we think we have a unique offering we try to get some margin, and the balance is really what we are trying to achieve opportunistically.

So like you said, when we are managing the Company, both the revenue growth as well as gross margin are important.

But, then, it's a judgment call that we have to make from design win to design win.

As a follow-up on the gross margin side, if we bring your guidance into it, George, saying at the high end of your historical range, you guys have kind of crushed the high end of your historical range for a number of years.

I realize this year benefited a little bit from selling written-off inventory, but what would bring the gross margin down from the level that you did in fiscal-year 2017 to that -- I guess the 62% would be the high end of your former range that you are talking about or is that just some implied conservatism?

No, I don't think it's implied conservatism when you look at the change in market mix.

Obviously, we had reasonably good margins in the wearable sports camera market, and our lower margin business or our lowest margin business is our China security business.

So if you look at decline in the sports camera, the China security business becomes a much bigger part of our business, and that is on the lowest end of our margin profile.

So I think this year, this change in mix, which we are basically forecasting at this point, would bring that margin down as we progress through the year.

I think with a small company like us, with some of these volatile markets, there's always events that could take place, for instance, like very strong growth in the drone market a year or so ago, which helped on the upside.

But at this point, based on where we see the individual markets going, I think it's reasonable to assume we'll start moving towards that high end.

And then, I guess my last question on the OpEx side, and correct me if I'm misheard you.

I think you said for fiscal 2018 up 12% to 15% year over year was the OpEx growth, is that correct?

12% to 14%, yes.

12% to 14%.

Is that sort of increase, it seems like it's a bit of a one-year focus.

Maybe the comparison versus the revenue growth would be the one-year focus, but longer term can you remind us how you try to manage OpEx versus the revenue growth?

And, again, I know this year has some unique attributes to it, but if we think beyond that, what's the model you try to emulate?

I think there's a bit of a change going on.

Historically, we very much focused on revenue growth and OpEx growth as sort of a joint budgeting effort.

I think now we have had some headwind we're going to have in the near term on revenue, but we feel strongly that our investment, particularly in computer vision, as well as the auto OEM market, has to proceed as we originally planned.

So I think either over this year and maybe possibly next year, you will see that continued investment.

Now, obviously, we want to keep pushing the topline up, but I don't think at this point we have any intention of slowing down that investment.

Those two areas are extremely important for us, particularly as we go forward in years three, four, and five out there, so we will emphasize that and we'll try to make it clearer to you as we go forward if there's any change in that position.

If I can just sneak in one quick one, hopefully quick, the CV product that you taped out in February, Fermi, when do you expect some revenues from that product?

Let's take a stab at that first.

I think we will start sampling this, I would say, early second half this year and hopefully that will go to the IP security and drone market first and we can demo with auto later.

Revenue wise, I think it's definitely next year.

Great.

Thanks, guys.

Brad Erickson, Pacific Crest Securities.

Hi, thanks for taking my questions.

First, I guess, on the last socket, can you just give us a sense between high end/low end kind of where you are losing that socket with your largest action camera customer?

And then, just curious if when we think about the holiday next year relative to your fiscal year guidance, just wondering if you can unpack sort of what you've excluded in terms of the mix from that largest action camera customer for the upcoming holiday next year.

Thanks.

Let me take that second part first.

I think the assumption right now is, based on their call, they've announced the HERO6, which I think is going to be their mainstream product, so obviously what I've quoted is for our revenues would assume that those type of mainstream products we won't be in.

So as far as the rest of the product line, there is significant inventory there that they can utilize through the year and that's really impacting our revenues.

That doesn't really define what their revenues would look like because of their inventory levels.

Got it.

And then just on the drone front, I think you guys are actually down year over year in the upcoming quarter, but relative to the full-year outlook, it seemed like the growth is still really strong.

Does that snap back right away in Q2 or is that more like a second-half phenomenon?

Right now, I think you'll see -- right now, we see a little bit more seasonality this year, potentially, so end of Q2, moving into Q3, and early Q4 would probably be the stronger periods, but, again, we don't have clear visibility on all the launch cycles.

In other words, we understand the development, but when they can get the drones launched is difficult for us to forecast.

That's why I mentioned earlier that we would expect volatility on a quarterly basis, but assume pretty good growth for the year.

Got it.

Thanks.

Richard Shannon, Craig-Hallum.

Hi, guys, thanks for taking my questions, probably just a couple ones for me.

I noticed in your qualitative description of your revenue guidance for the year I think you were talking about some VR launches.

I'm wondering if you can give us a sense of the scale you are thinking about, how late in the year, how broad of a customer base, et cetera, that you are implying in there.

We are still watching the virtual reality market.

We have announced several customers previously and we continue to see design win activity, and in terms of the market development, I think it's still at a very early stage and it's still hard for us to tell how big the market is, so while we continue supporting our key customers for this market, we are still conservative on the revenue outlook for this.

So we talk about that we do expect to see some revenue growth out of VR this year, but it's going to be off a very small revenue base.

Okay.

That's kind of what I thought; just wanted to clarify on that one.

I guess my other question is if you can kind of give us -- kind of peel back the layer on the China security market.

It seems like you're expecting a little bit better environment this year than perhaps last year, although perhaps it will be kind of a headwind on the gross margin side.

Fermi, if you can give us a sense of how you see the overall revenue stream you're developing as you look at it by unit, unit market growth, competition, and any other dynamics going on as it relates to China would be great to hear.

I think for the security camera point of view, I think Chinese particularly, I think the one major difference between last year and this year for us is we get our midstream HEVC product in production with HIKVision Daiwa and we didn't play in that market last year.

Well, we are going to play this year, so that would help us to get some market share and also grow our revenue.

And from a market point of view, I think the whole market continued to grow, but really consolidated very quickly, and when we started in this business, we were talking about 200 companies.

Now I think it's only probably the top 10 really matters in this business.

In China, Hikvision/Dahua are the only two really driving the whole business and that's [strictly within chance].

In terms of total unit number, we see all kind of numbers, but we do believe the market continued to grow last year and, surprisingly, the analog HDTV was the biggest growth factor last year and we started playing in that market with Dahua, so we were tapping into that market, also another growth area for us for this year.

Okay.

I think that is good for me.

Thanks for all the detail, guys.

I'll jump out of line.

Kevin Cassidy, Stifel.

Thank you for taking my follow-up.

On the computer vision products, can you say how many customers you'll be sampling and the range of customers?

For this kind of a product, we need to be careful out who we sample to, so we are going to send it to our strategy customer first.

You can see that for the drone and IP security.

You know our customer base.

They are our first priority to sample to.

Of course, there will be more in the future, but for the very first release, it will be our current strategy [promise].

And maybe as a follow-up, what product differentiation do you think you'll bring to the market?

We haven't talked about this publicly yet, but when we start assembling this chip, we will start talking about our features in the stack to the public.

Okay.

Thank you.

Suji De Silva, ROTH Capital.

I just wanted to understand the -- at CES, you showed the 8K product for the sports camera market.

I'm wondering when that dovetails with your customers' roadmaps?

Is it a situation where as that comes into the market you expand your differentiation again versus competitors who are starting a second source here?

We missed a part of that, Suji.

Was it what the impact the CV what have on the sports camera market?

No, not CV, just 8K, the transition to 8K from (multiple speakers)

Oh, 8K (multiple speakers).

I think 8K is going to be a high-end product only for the drone and maybe sports camera in the long term, so I think from a customer point of view, you should expect only the guys who want to sell a high-end camera start using 8K, so I won't say that's an immediate high-volume market for us, but that's definitely something we start seeing people asking for.

Not really just for 8K spec, but like I said we have started talking to a lot of customers which requires putting multiple cameras into chip, and the 8K capability allows us to, for example, put in two 4K P60 cameras combined into a 360 surround-view camera.

That's another possibility to use that chip for.

So, 8K, if you just look at 8K spec, that's limited.

Like if you expand it to different markets which requires multiple camera input, then the application becomes much bigger.

Okay.

That helps.

And then on the drone market, it seems like it's a little harder to call that out as far as linearity wise.

Are manufacturing constraints still affecting your drone customers here, George, and have those passed, or if not, when do those pass?

I think -- the constraints have completely past.

What we're seeing is it is difficult to manufacture these and I think throughput in the factory is probably still slower than maybe the drone manufacturers would like, and I think most of them are curing that by increasing overall capacity.

So I think it will improve going forward.

I still think both the ability to launch drones, as well as getting them manufacturable, is still going to be an issue for a while, but I think we've worked those into our numbers this year.

We have a little bit more experience there.

And currently, I think the best-selling product in the market is DGI Magnum Pro, and late last quarter we did see some backlog putting to order, but I think they will clean the older backlog.

If you order today, you can get receiving the product very quickly.

So my assumption is that DGI figured out how to address their manufacturing process.

So at least for this high-volume application, the Magnum Pro, there should not be any delay in the near future.

Okay.

Thanks, guys.

Thank you.

And at this time, I would like to turn the call over to Dr. Fermi Wang for closing remarks.

Thank you, everybody, for joining us today.

I want to give a special thanks to our colleagues for their continued dedication and great support.

Goodbye for now.

Thank you.

Thank you.

Ladies and gentlemen, this concludes today's conference.

You may now disconnect.

Good day.