CEVA Inc (CEVA) 2016 Q4 法說會逐字稿

Hello and welcome to the CEVA fourth-quarter and year-end 2016 earnings conference call. (Operator Instructions). Please note, this event is being recorded.

I would now like to turn the conference over to Richard Kingston, Vice President, Market Intelligence, Investor and Public Relations. Please go ahead.

Thank you. Good morning, everyone, and welcome to CEVA's fourth-quarter and annual 2016 earnings conference call. I'm joined today by Gideon Wertheizer, Chief Executive Officer of CEVA; and Yaniv Arieli, Chief Financial Officer of CEVA. Gideon will cover the business aspects and the highlights in the quarter and year, and provide general qualitative data. Yaniv will then cover the financial results for the fourth quarter and year, and also provide guidance for the first quarter of 2017 and general data for the full year.

I will start with the forward-looking statements. Today's conference call contains forward-looking statements that involve risks and uncertainties as well as assumptions that, if they materialize or prove incorrect, could cause the results of CEVA to differ materially from those expressed or implied by such forward-looking statements and assumptions. Forward-looking statements include our financial guidance for the first quarter of 2017, including the degree of confidence in guidance, and general market outlook and revenue drivers for 2017; expectation of a favorable licensing environment, including provision and imaging, LTE IoT services, Bluetooth and 5G; optimism about the licensing pipeline and customer ramp-up schedules; and our ability to capitalize on LTE migration.

The risks, uncertainties and assumptions include the ability of CEVA's signal processing IPs for smarter connected devices to continue to be strong growth drivers for us; our success in penetrating new markets, specifically non-baseline markets, and maintaining our market position in existing markets; the ability of new products incorporating our technologies to achieve market acceptance; the speed and extent of the expansion of the 3G, LTE and 5G network, Bluetooth 5, and the IoT space; customer ramp-up schedules and the impact on royalty revenues; the effect of intense industry competition and consolidation, global chip market trends, and general market conditions and other risks relating to our business, including but not limited to, those that are described from time to time in our SEC filings.

CEVA assumes no obligation to update any forward-looking statements or information, which speak as of their respective dates.

That said, I would now like to turn the call over to Gideon.

Thanks, Richard. Good morning, everyone, and thanks for joining us today. 2016 was an outstanding year for CEVA, with all-time record high revenue of $72.7 million, up 22% year-over-year with great momentum in terms of customers' adoptions and deployments of our advanced technologies. This evidences our successful transformation into a leading one-stop-shop IP house for wireless broadband and Internet of Things-related technologies.

This impressive execution could not otherwise have been achieved without the relentless work and the perseverance of our 280 employees worldwide. I would like to take this opportunity to thank them for all this great performance.

For the fourth quarter, we delivered another quarter with strong financial results, reaching a record high total revenue, a record high royalty revenue, and record high non-GAAP EPS. The licensing environment continues to be robust, with $8.3 million in revenue reported for the quarter derived from 15 deals concluded, of which 14 were for non-handset baseband application, and four of which were with first-time customers that recently joined our more than 60 active customers. Geographically, four of the agreements were in the US, and 11 were in Asia, including Japan.

The fourth quarter was particularly strong for our vision product line, with eight deals concluded with customers targeting next-generation smartphones, automotive ADAS, smart surveillance camera, mirrorless camera, and more. We are also extremely happy that ON Semi, a tier-one image sensor provider in the automotive ADAS space, joined our customer base. We are looking forward to collaborate with ON Semi in powering intelligence camera that surround the car.

Our vision portfolio, which incorporates our leading XM4 and XM6 DSP with a comprehensive list of vision algorithm, and our reputable CDNN2 deep learning software, continues to set new milestones in innovations and customer engagements. We now have more than 30 design wins for our imaging and vision technologies.

Royalty revenue derived from third-quarter shipments was strong due to combination of LTE and Bluetooth shipment growth on seasonal tailwinds, as well as from one-time catch-up payment of approximately $1.1 million from a handset baseband customer for prior quarters. Overall royalties reached $12.9 million, up 60% year over year, and the highest quarterly royalty revenue in the Company's history.

Let me take the next few minutes to provide you with a summary of the underlying drivers for our business last year and the perspective for 2017.

On licensing, in 2016 we concluded 49 licensing agreements, 45 of which were for non-handset baseband application and use cases; and the balance of four deals were with large incumbents in the handset baseband space. 17 of the agreements signed in 2016 were with first-time CEVA customers. These design wins align very well with our diversification strategy beyond cellular baseband and the market we target.

Our non-handset baseband category contain technologies and products for imaging and vision, LTE IoT, Wi-Fi, Bluetooth audio, and base stations. Those technologies are centerpiece in every IoT device, and enable devices to become smarter via processing of advanced algorithm such as deep learning, machine vision, and voice assistance; and be wirelessly connected via long-range wireless, such as LTE and 5G, or via short-range wireless such as Bluetooth and Wi-Fi connectivity. This possesses an opportunity in tens of billions of units in the coming years for which we can address with multiple royalty [billing call] per device.

In addition, our product line for the base station market, along with our long-standing relationship with tier-one OEM expose us to lucrative business model in the infrastructure market, and uniquely positions CEVA as an end-to-end technology provider for the 4G and 5G mobile broadband space.

As we look into 2017, we are expecting continuation of the favorable licensing environment, driven by the following market dynamics. First, the growing adoptions of advanced computer vision and machine learning algorithms in smartphones, IP security camera, ADAS, and new camera modalities such as action camera, 360-degree camera, VR and AR. Our XM4 and XM6 DSP platform offers higher performance, lower power consumption, and are accompanied with value-add software and neural network framework that significantly reduces the barrier to entry into those new markets and applications.

Second, the emergence of LTE IoT services -- LTE IoT is a composition of multiple low bitrate LTE standards, among which are Cat-NB1 and Cat-N1, aimed to enable billions of very low-power wirelessly connected devices to communicate with each other and with the cloud. The LTE IoT usage model is applicable for broad markets such as automotive, smart meter, [welders] and smart home. According to the recent Ericsson Mobility Report, it is expected that cellular IoT will exceed 1.5 billion devices in 2021, at 27% CAGR between 2016 and 2021.

CEVA is already engaged with customers in this space, in particular in relations to China Mobile, who is taking an early lead in proliferating devices and services around LTE IoT. Our latest DSP, CEVA X-1, addresses they LTE IoT modem workload extremely efficiently, as well as being able to process value-added applications such as voice, GPS, and sensor fusion.

Third, the acceleration of Bluetooth design resulting from the new Bluetooth 5 standard. Bluetooth 5 represent a quantum industry shift for Bluetooth, offering up to 4X longer range and 2X higher data rate versus its predecessor, Bluetooth 4.2. Developers can take advantage of it for new use cases in welders, smart home, beacons, and audio over Bluetooth's low energy.

Also in conjunction with audio, we believe that Apple AirPort, its new Bluetooth-based earbud, will drive other handset OEM to follow and to displace the wired earbuds equipped with every smartphone sale with Bluetooth-based earbud. Our leading Bluetooth offering, along with our unique expertise in audio [audibility] put us in an excellent position to benefit from the volumes expected. We already have half a dozen Bluetooth 5 design wins, and expect our first Bluetooth 5 customer to announce its product very shortly.

Fourth, we are expecting 5G design and licensing activities to expedite. While the standard is about two years out, operators are looking to deploy pre-standard products as early as 2018. 5G also has unified network supporting multiple services, among which are 20 gigabit per second peak data rate for mobile broadband; low latency for mission-critical use cases in autonomous driving and medical; and large capacity for the billion IoT devices expected to be connected to the 5G network.

Our CEVA-XC and CEVA-X product line offer our customer multiple and scalable options to address this market opportunity.

Turning to royalty, 2016 was an extraordinarily strong year with remarkable 49% annual royalty revenue growth on the back of almost 1.1 billion CEVA-enabled product shipments. In handset, approximately 865 million phones were sold with CEVA IP, of which 510 million were smartphone, and the balance was feature phone.

LTE annual shipment was in particular strong, reaching a new record high for us of 226 million units, and representing a 223% unit growth on a year-over-year basis.

Our diverse technology for baseband processing allows us to address customer needs for all different market segments, starting with the [stream sides] of the market of local feature phones widely used in India and other emerging economies; growing 3G and LTE meter in smartphone, and all the way to flagship models. We also experienced accelerated growth in shipments in the second half of the year in regard to our non-handset product with 26% year-over-year unit growth, reaching 210 million units, and more than 71 million of them just in the fourth quarter of 2016.

Looking into 2017 on handset, based [fairly via] research, the market is forecasted to grow annually by 6% to exceed 2.1 billion handset shipment. We expect continued migration to LTE, which benefit us as EPs carries higher royalty ASP. India, in particular, is expected to be strong in LTE migration. In the third quarter last year, seven out of 10 smartphones shipped were LTE-enabled, according to IBC India. We are also monitoring the recent initiative of Reliance Geo, a tier-one operator in India, to launch LTE feature phone based on spread on chip. This benefit Geo in terms of spectrum efficiencies and standardization around LTE chip rather than our data chips. Feature phone currently has 56% market share in India, and any migration to LTE will benefit us in terms of ASP.

On the non-handset baseband, we are set to benefit from the deployment of our vision processor in smartphones to enable differentiated camera features such as night vision, digital video stabilization, zoom and more. The ASUS ZenFone 3 Zoom, a [major end] smartphone which was announced at CF, is enabled per our vision DSP. More smartphones and SKUs featuring our vision DSP are set to be introduced shortly.

We also expect ramp-up during 2017 of multiple DSLR camera, smart surveillance camera, and automotive ADAS dish cameras.

In the Bluetooth and Wi-Fi segment, we expect another growth year in unit shipments as new devices as use cases reach the market, particularly in fitness, [healber] and healthcare applications.

In conclusion, 2016 was a great year for CEVA, with outstanding financial performance and strong execution and customer engagement. Our strong competency developed over the years, as a result of our disciplined investment, enable us to innovate and put us at the forefront of technologies required for smart and connected devices. We are determined to continue strengthening our competencies, and working closely with our customers to capitalize on the numerous opportunity in front of us in the smart of connected world.

Finally, I would like to take this opportunity and thank again for our employees, suppliers, and investors for their support, and wish you all happy and prosperous year.

With that said, I'll now turn the call over to Yaniv, who will outline financials and guidance.

Thank you, Gideon. I'll start by reviewing the results of our operations for the fourth quarter 2016. Revenue for the fourth quarter was $21.2 million, as previously announced. This would also be our fourth consecutive quarterly all-time high record revenue achievement. This was 32% higher on a year-over-year basis, and 19% higher sequentially.

The revenue breakdown is as follows. Licensing and related revenue was $8.3 million, reflecting 39% of our total revenue, 3% higher as compared to the fourth quarter of 2015, and above our plans and expectations. Royalty revenue was $12.9 million, reflecting 61% of total revenue, an impressive increase of 60% on a year-over-year basis, and the eighth successive quarter that we delivered year-over-year royalty growth. Fourth-quarter royalties include a one-time catch-up payment of approximately $1.1 million from a handset baseband customer, relating to prior quarters.

Quarterly gross margin was 92% on US GAAP basis and 93% on non-GAAP basis. Non-GAAP quarterly gross margin excluded approximately $67,000 of equity-based compensation expenses.

Our total operating expenses for the quarter were just below the mid-range of our guidance at $13.4 million. OpEx also included aggregated equity-based compensation expense of approximately $1.5 million, and $0.3 million for the amortization of acquired intangibles of RivieraWaves. In total OpEx for the quarter, including these two items, were $11.6 million, at the mid-range of our guidance.

US GAAP net income and diluted EPS for the quarter were significantly up at 126% and 118% to $5.2 million and $0.24, respectively. This compares to net income and diluted EPS of only $2.3 million and $0.11 for the fourth quarter, a year ago. Our non-GAAP net income and diluted EPS for the fourth quarter of 2016 increased 98% and 88% year-over-year to $7 million and $0.32, respectively. Non-GAAP net income and diluted EPS for the fourth quarter of 2015 was $3.6 million and $0.17, respectively.

These figures for the fourth quarter of 2016 and 2015 exclude equity-based compensation expenses, net of taxes of $1.5 million and $1.2 million, respectively. And the impact of amortization of acquired intangibles of RivieraWaves, net of taxes of $0.3 million and $0.2 million for these two quarters in both years.

Other related data -- shipped units by civil license fees during the fourth quarter of 2016 were an all-time high at 343 million units, up 24% sequentially and 35% for the fourth-quarter shipments of 2015. Of the 343 million units shipped, 271 million units, or 79%, were for handset baseband chips, reflecting a sequential increase of 24% from 218 million units of handset baseband shipped, and 34% increase from 202 million shipped a year ago.

Non-handset volume shipments also increased 20% sequentially and 41% year-over-year. The increase is due to a record high quarterly Bluetooth shipments in Q3 and continued ramp-up of audio and voice products powered by our DSPs. The quarterly handset baseband royalty ASP was flat sequentially, but increased 23% on a year-over-year basis due to a higher volume mix of smartphones as compared to feature phones. Our overall quarterly corporate blended ASP was about flat sequentially, but increased 18% year-over-year due to this product mix.

Some other interesting annual data -- our annual smartphone shipments, both 3G and LTE, increased 72% on a year-over-year basis, and grew by north of 200 million units to reach a record high 510 million phones; while feature phone shipments decreased by about 100 million units to reach 356 million phones for the year.

Our total shipments grew respectively by 17% year-over-year to reach 1.1 billion devices, which equals to approximately 34 CEVA-powered devices sold each second in 2016. These unit shipments represented significantly annual royalty revenue increase of 49% year-over-year, reaching an all-time record high of over $40 million for the year.

Annual handset baseband royalty ASP was up the second year in a row, over 30%, due to more favorable mix of smartphones. And our annual blended ASP across all our products on an annual basis was up 27% to $0.038 per unit.

As for the balance sheet items, as of December 31, 2016, CEVA's cash, cash equivalent balances, marketable securities and bank deposits were approximately $156 million. In 2016, we paid $2.2 million as part of our last contractual commitments from acquiring RivieraWaves.

Our DSOs for the fourth quarter was a bit above normal, at 65 days, but down from the prior quarter; and after we received the September 30, 2016, outstanding payments that we discussed about at the last conference call.

Regarding our share buyback program, in 2016 we only purchased during the first quarter of the year approximately 180,000 shares for an average price of $19 per share, and for approximately $3.4 million. We have approximately 300,000 shares left in the program. And we have overall concluded north of 5.7 million shares for an aggregated $87 million throughout our buyback program lifespan.

During the fourth quarter, we generated $12.5 million of cash from operations. Depreciation was about $400,000. Purchase of fixed assets were about $360,000. And at the end of last year, our headcount was 278 people, of which 220 were engineers.

Overall, post- fiscal year 2015 financial achievements, we continue to demonstrate excellent and new, enhanced milestones in 2016, leveraging on revenue growth, new customers and markets, as well as disciplined expense control. In addition to growth in total revenue, we generated non-GAAP operating income increase of 87%. We generated non-GAAP EPS increase of 75%. And we had a solid annual free cash flow from operations of $14.5 million. Backlog at year-end is at its new record high. The deal pipeline is robust. And royalty momentum appears to continue in our favor.

Now for the guidance. I'll start with some general data in relation to the full year. On licensing, as Gideon described, we are experiencing a healthy demand across the entire range of products we offer. While licensing revenue can be lumpy due to the timing of deal closures, we believe we have the fundamentals to increase our annual licensing revenue guidance up from the $32 million level that we experienced in the last two years to a new higher range of $34 million to $36 million for this year.

On royalties, we believe that 2017 will be another growth year. In handset baseband, we are forecasting growth with a higher mix of smartphones versus feature phones, primarily driven by the expansion of 4G LTE in India and other emerging markets, and the introduction of new flagship smartphones in the second half of the year.

In non-handset baseband, we are forecasting noticeable growth in royalty revenue as our vision customers are expected to ramp production, and as our strong presence in the Bluetooth market continues. Overall, we forecast 10% to 20% annual growth in royalty revenues at this stage.

On OpEx, with our increased confidence in our licensing revenue, which is reflected in our guidance and in line with Gideon's remarks earlier concerning our disciplined investments to innovate and to reinforce our leadership, we plan to increase our investment in R&D in vision, LTE for Internet of Things, 5G, and next-generation Bluetooth product lines. Our OpEx increase is mainly associated with headcount, employee-related costs that typically begin with a new calendar year, outsourcing costs, and EBA tools for these new product lines.

The increase will be in the region of $5 million for the year. It should be noted that approximately one-third of this growth is attributed to lower government grants for our R&D efforts. Equity-based compensation expenses is also forecasted to be higher this year, due to ongoing employee-related retention efforts.

Before I deliver our guidance for the first quarter of 2017, I would like to note that we have already managed to conclude a large portion of our licensing business for the quarter. And although we have not yet received a number of royalty reports, due to the Chinese New Year holidays, we believe we have good visibility to project the royalty revenue for the quarter. And, as such, we are guiding Q1 with a high level of confidence.

Now for the guidance for Q1. Revenue for the first quarter is expected to be in the range of $20.5 million to $21.5 million. Gross margin is expected to be approximately 92% on a GAAP basis, and 93% on non-GAAP basis, excluding equity-based compensation expense.

Overall OpEx should be higher in the first quarter compared to the rest of the year, due to the timing of the Israeli R&D grant programs for this year. Q1 2017 OpEx is expected to be in the range of $14.5 million to $15.5 million. And of the expected total operating expenses for the quarter, $1.9 million is expected to be attributed to equity-based compensation expenses; and $0.3 million to amortization of acquired intangibles. So our non-GAAP OpEx is expected to be in the range of $12.4 million to $13.4 million for the first quarter.

Net interest income is expected to be approximately $500,000. Tax rate for the first quarter is expected to be a bit lower than the norm, and the following quarters as well, at approximately 11% for non-GAAP and 14% for GAAP.

Share count for the first quarter, approximately 22.5 million shares. And that will bring us to US GAAP fully diluted earnings per share in the range of $0.18 to $0.20 per share; and non-GAAP EPS forecasted, excluding the aggregate $1.9 million of equity-based compensation expenses, net of taxes and amortization expenses, of $0.3 million, non-GAAP EPS is expected to be in the range of $0.27 to $0.29 per share.

Operator, you can now open our Q&A session, please.

(Operator Instructions) Matt Ramsay, Canaccord Genuity.

This is Logan Bender dialing in for Matt. Congratulations on a great 2016. A couple quick questions. Looking to 2017, licensing growth is diversifying in the pipeline really nicely. But the growth will primarily be driven by baseband in the near term. So how should we think about the moving parts of the 2017 LTE unit growth with a full year of Intel at Apple, and some strength in China with Leadcore; but a bit of uncertainty at Samsung with the roadmap there? So could you talk a little bit about that, please?

Sure. So as we mentioned in the prepared remarks, and as we have looked at the business in the last two years, today we divide our handset business into two segments: the feature phone and the smartphones. Smartphones increased dramatically last year from 300 million to 500 million units. And within those 500 million units, we were able to triple our LTE content from 70 million, if you remember back in 2015, to north of 220 million in last year, 2016.

When we look into 2017, first we believe that those 500 million smartphones will continue to grow for us. And if we open up specifically the breakdown of the LTE of 4G within those smartphones, at this point of time we believe that those 220 million LTEs should grow north of 300 million for the year. So that would be, again, the positive contributor both from LTE, the higher ASPs, our bigger market share as we go along. And that's the positive momentum that we see on the handset space.

Okay, great. And then a really quick question about some of the royalty trends, some positive news there. Is there anything else, any more granularity we can add to the royalty unit trends into 2017 with the LTE royalty and it gradually coming down? So could we get a little bit more there? Anything else?

In our prepared remark, we spoke about the non-handset category of product. And we mentioned specifically two products that we believe will get into noticeable production this year. I should say this. One is the vision, and that's a very interesting trend because it's associated with smartphones. We see OEMs and a few semiconductors like Rockchip that are not the SoC companies, but they find a space in the smartphone to provide the chip that its main purpose is to enhance the stream and the video quality. And as you know, if you take the iPhone 7 Plus with the dual sensors, these are the things that people are looking to add into the 2017 model. So vision is an area that we see growth. It is something that we didn't have in 2016.

The other area is Bluetooth. And specifically, Bluetooth 4.2 and Bluetooth 5 powered the second half of the year. This is something -- again, in 2016, the first half was kind of a pose because there was a transition to more advanced Bluetooth technologies. We are going to see a unit grow in 2017. And then, really, the end of the year we expect to see base station. And this is a lucrative business, as I mentioned.

Okay, great. Very helpful, guys. Thanks, and congrats on the year.

Joseph Wolf, Barclays.

Just a housekeeping item of sorts; but also, of course, the new products in the non-handset -- as you announce some of the -- or as you get deployed in some of these vision products, that was going to be reported in the non-handset business. And is that accretive to ASPs versus the handset LTE business that you're selling right now? Or how does that fit into the ASP mix?

So, yes, it is going to be reported in the non-handset stuff. As we grow, we probably -- maybe we could open it up and give a little bit more color. But I'm sure we will give some milestones as we start counting those units. ASP is probably twice as high in vision than the average of the Company, or up to twice as high, I would say. So it's a nice incremental add-on. We just need to look for the unit volumes.

Not all of them are handsets; some are. And those specifically could be higher volume. Some could be consumer devices, the SLR cameras and the like. So you don't sell tens of millions of them, so we have to just count the units as they start hitting the market. And hopefully, that -- and we are planning for that segment to be an interesting growth driver in royalties for the first time this year.

Okay, thank you. And as you look at the -- you gave guidance, if I heard you correctly, for 10% to 20% royalty revenue growth for the year. But the number that you gave for the -- is that correct?

Yes, that's correct.

The LTE number you just gave is like 36% growth. So was that a conservative number? What are the puts and takes as we think about that revenue number with some of the opportunities here, which all seem -- which, generally speaking, seem to be pointing towards higher ASPs? Or is the Bluetooth growth at lower ASP going to be good for unit growth but less so for royalty growth?

Yes. That's always the $1 million question, sort of the mix between all these product lines. One are very high ASPs; the other could be much higher volume, but could be $0.005 to $0.01 in the lower ASPs. So it's always difficult for us, especially in the beginning of the year, to have and manage to guide and build a model from all these moving pieces. We are trying to do it sort of bottoms up, so we are looking at growth with the higher ASPs; we are looking at growth with more units and lower ASPs.

At the end of the day, maybe the average ASP does not move that much. It moves a little bit. It depends, again, on what will be the stronger contributor. But overall, from a dollar perspective, and when we had some of the new markets which are not only handsets, at least at this start of the year we are looking at 10% to 20% growth in dollar contribution of royalties; and another around 10% growth in the licensing part of our business, which is not common practice for us from just organic efforts.

Okay. And then finally, are there any -- there's a lot of articles right now about the Alexa product, which would seem to be a product that you guys could be in. Are there any flagship consumer products were CEVA has been designed in that we can talk about? And as you look at some of these devices and some of the auto and vision or specifically in the auto market, are the design wins that you have right now integrated product in the manufacture of the auto? Are they still mainly add-on accessories that customers buy afterwards? And does the relationship with ON put you into that long-term design cycle where we won't see revenues for a few years, but we will have a design win before that?

Let me answer you. You basically asked two questions. The Alexa product is -- it is a class of product that is called voice assistant. Now, the idea with Alexa, and as well as with Google is that they enable an interface to their cloud and open up for everybody to develop all sorts of product. So I should say that in the quarter, we signed our sales deal to this class of product. This is associated, not with Amazon; it's associated with Alibaba. As you can imagine, in China you have similar trends and in much larger scale. So, yes, we are involved in the voice assistant area, and we have already customer in this respect.

That said, your question about the voice assistant; with regard to the ADAS area, we mentioned the deal with Sony, and we have other deals. This is not an add-on accessory. We do have another customer that is in what is called the aftermarket. This is what you refer as an accessory.

ON Semi and the other customers that we have are targeting to the ADAS in autonomous driving, what people are calling Level 2 and above; 2 and 3 are the near-term, and 4 and 5 -- these are the fully autonomous. And this is still a way to go. But we are speaking here on product, and certainly the core collaboration with ON Semi will give us even more credibility and entry point into the mainstream of the market.

Also at CES, we announced a few other players like Rockchip and Novatek, which are also players in the aftermarket devices that is associated with your question. So the design wins are there. We need to see now the volume production kick in.

Thank you, guys.

Gary Mobley, Benchmark.

Congratulations on a strong year and strong finish to the year. I had a question about the continuation of the increase in deferred revenue. So here we sit at the end of the fiscal year at $6.3 million. If I'm not mistaken, that's a 50% sequential increase over what was a pretty substantial increase in Q3. And so I'm wondering if that has to do with the subscription-like license deals that we talked about last quarter; or if there were additional subscription-like license deals concluded in the fourth quarter; and then, as well, any other contributing factors that spike in deferred revenue and backlog.

No; that's a quick answer, because I think you answered your own question quite remarkably well. It mostly has to do with the subscription deal that we signed back in Q3. And the other two deals that we talked about last quarter that are more work-related, so sometimes we get the payments ahead of revenue recognition. But this is exactly the reason that, here and there, we have these -- a little bit of interesting or different deals. And of course, that helps us also with our guidance, with our confidence that we talked about Q1 in general and the licensing for the year, which is not trivial for us that was -- at least in the past, was not. And that gives us stronger confidence. But the reasons are exactly as you describe.

Okay. Just a housekeeping question: what kind of tax rate do you expect for the full-year 2017?

Non-GAAP, a bit lower than last year, probably around 12% up to 13%. First quarter is going to be a bit lower than that, and it gradually increases over the quarters.

Okay. So I'm having a hard time dumbing down the royalty expectation for 2017 to be in a range of 10% to 20% growth. And I say that because if you are going to get to $700 million and $900 million non-mobile handset baseband royalty units by 2018, we have to see something offsetting that. And so I'm wondering if you are standing by that $700 million or $900 million royalty forecast for non-mobile handset for 2018, or any change to that range as we sit here today. Comments -- the insight would be helpful.

I could start. It's a good question. And I think we are today trying to give qualitative data for 2017. We don't have the model like that $700 million to $900 million. We've done a model, but it was more of wishful internal targets for us. I think we are not yet giving out a model for next year from a volume perspective, nor from a dollar perspective.

This is something that we will probably do throughout the year, and maybe for 2018 and maybe for 2019, a little bit longer-term, when we get a little bit more facts. And I think what we are waiting for are two things. One is the non-handset, really to see those units and dollars flow in, and the volumes in order to substantiate those 700 million to 900 million units. Or, on the other hand, the dollar figures, especially in 2018 from the base station market, which may be are small units in the overall mix of that $700 million; but in a dollar perspective it's much, much more significant.

And I don't think at this point we are giving any specific numbers of how many hundreds of millions we'll want to power in two years. But I think throughout the year we'll try to give a little bit more color; and maybe even move out a year, a little bit further out, as we get more confident from the volume perspective, less from the dollar perspective.

I think what we are expecting to see in 2017 set a precedent for what we are going to see in 2018, what will be the contributor. So cell television, and I mentioned -- I think somebody else asked me about smartphone. This is a market that expected to take volume direction. And Yaniv mentioned base stations, and of course the Bluetooth 5. And I mentioned the Apple AirPort, the type of product which we are very optimistic about the trend of moving from wired earbud to Bluetooth earbud.

So these are products that basically could be done in designs today and could be in high volume in 2018. (multiple speakers) So we have good understanding, and stand behind our forecast for 2018, in light of the fact that we already see those elements going in or flourishing in 2017.

All right, thanks for the response, guys. Again, congratulations.

Matt Robison, Wunderlich.

Congratulations, Yaniv. Nice to see the cash flow, and especially since it looks like you've got some room to continue it with a further DSO reduction, especially if your quarter is already largely in place on the licensing side. Hopefully, it's front-loaded a bit.

I wanted to first ask you -- you mentioned this ramp of non-baseband this year. That would normally imply lower ASPs. But some of this vision stuff -- my understanding is it's higher ASPs. Do you expect that -- the royalty ASP to go up, or stay about the same, or go down in 2017?

And then, Gideon, you mentioned base station. I really wasn't thinking about base station being a factor until 2018. Do you think that's going to come in this year?

And then the other -- lastly, what should we think about in terms of timing for the vision and other application-oriented functions to start to show up this year?

Let me start with the first one. We told you in the beginning that we are not sure what the mix is. It may end up more or less the same. Again, it depends on the volume. We ended 2016 with the record high Bluetooth devices, 50 million in the last quarter, and another 21 from these other devices, mainly audio related. And it depends how fast, with all these new products that Gideon mentioned, where it will take us by the end of the year. LTE will add higher ASPs. Vision will add higher ASPs. But these may be offset by lower ASPs from connectivity and audio.

I don't think that the targets for us is to increase or decrease the average ASP of the Company, but just to be able to tackle many more markets. And we get royalty revenues from many more customers, which was not the common practice for us, three, four, or even two years ago. Still, the bulk of it is coming from handsets. So I would, for now, maybe keep it flat, maybe slightly higher, but not too much. And we will see how it goes with the quarterly ASPs.

Just a quick answer to your base station. So far, what we see is a good progress with our base station customer. And we think that by the end of the year, we will start production. The exact amount we don't know yet; but DC is going very well, I should say.

Thanks, guys.

Suji De Silva, ROTH Capital.

Congratulations on the strong 2016 sales here. As you look at the licensing pipeline in 2017, you talked about strong (technical difficulty) diversification in 2016. Do you expect similar new customer interest in 2017 to continue to diversify in a similar magnitude or fashion?

I mentioned the full anchor product that will drive, in our opinion, the 2017 licensing. Three of them are non-handset; these are vision, Bluetooth 5 and the LTE IoT, which is in China. It is going very well; very fast traction there. In the baseband side, the way we see it is -- and we see today is 5G. People are starting pulling the trigger on 5G designs. They understand that this will happen eventually, and this is really long cycle. So we had, last year, 5G customer. And I'm putting aside the base station, because in base station [relative] customers for 5G. But in terms of handset and what is called user equipment, this is now moving.

Okay, great. That's helpful. And then you talked about vision being a nearer-term opportunity. How would you size the unit opportunity there and how do you talked about the relative ASP of vision for you guys?

Yes, sure. Let's start. The ASPs should be probably up to twice the average of the Company's ASP, anywhere from 1.5 to 2 times. That would be an average norm in high volume. We have a few sockets and handsets which we may sit along next to a Qualcomm Snapdragon, for example. If we talk about a handset, it is going to be in few SKUs; that could be in the millions; 10 million, 20 million, if it's a high runner, it's a good seller type of phone.

We have two or three different DSLR cameras. If you look at an annual base, these are probably up to 1 million; or, if it's a super camera, up slightly more, but in hundreds of thousands. And we have few of them, so that ramps up, or potentially ramps up to, again, higher volume.

And the third is the surveillance. Surveillance is a bigger market opportunity, but it's the first time for us. So again, we need to -- as soon as we are designed in the product, you will see them -- they are richer than the month quarterly reports that we look at new products. You saw it in our IR presentations, of course. And we will count the units as we go along. Because I don't think that ahead of time we have a clear crystal ball how much our customers are planning to ship, or at least in the internal part, because this is relatively new products with us inside. So one step at a time. But we are forecasting quite a few millions of dollars both on the unit volumes, as well as the dollar contribution. And we will add them up as we go along.

And then maybe lastly on the vision part, can you talk about the competitive landscape in the sense of why this functionality would necessarily be integrated into an application processor, why you think it's a sustainable opportunity there for you guys?

The vision is a developed area. And right now, if you take the smartphone, which is now getting a boost in terms of -- and I think if we stop, it's because of iPhone 7 Plus came out with the dual sensor, and it showed a benefit.

What makes us optimistic is that the next step should be neural network. Right now, people are speaking on this on roadmap terms, but we have the product and a clear advantage. I mentioned in my prepared remarks, our software on top of it. I think today in the software side for neural network, other than Nvidia no one is having similar comprehensiveness in neural network that we are having. And I think this is where we don't think we have an edge versus competitors. It is more trying to put just in hardware code and less software. And then our approach is we are compressing one, where we combine hardware and software.

Okay. Thanks, guys.

And this concludes our question-and-answer session. I would now like to turn the conference back over to Richard Kingston for any closing remarks.

Thank you. Thank you all for joining us today and for your continued interest and support of CEVA. We will be attending the following upcoming events, and invite you to meet us there. We will be at Mobile World Congress in Barcelona from February 27 to March 2. We will be attending the 29th Annual ROTH Conference in Dana Point, California, March 13 to March 15. Please visit the Investors section of our website for further information on these events and other events we will be attending. Thank you and goodbye.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect your lines. Have a great day.