Hello, welcome to the CEVA third quarter 2016 conference call. All participants will be in listen only mode. Should you need assistance, please signal a conference specialist by pressing the star key followed by 0. After today's presentation, there will be an opportunity to ask questions. To ask a question, you may press star then 1 on your touchtone phone. To withdraw your question, please press star then 2. Please note that 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, Mr. Kingston.
Thank you. Good morning, everyone, welcome to CEVA's third quarter 2016 earnings conference call. I am joined today by Gideon Wertheizer, Chief Executive Officer at CEVA, and Yaniv Arieli, Chief Financial Officer at CEVA. Gideon will cover the business aspects and the highlights from the quarter and general qualitative data. Yaniv will cover the financial results for the third quarter and provide guidance for the fourth quarter of 2016 and the full year. I will start with the forward-looking segment. 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 fourth quarter of 2016 and the entire year, future revenues to be generated from the expanded portfolio license agreement, and optimism about extending such a licensing model to other customers, optimism about the licensing pipeline and customer ramp-up schedule, our ability to capitalize on emerging market opportunities, including cellular IoT, machine learning, non-baseband, LTE, and 5G, and resolution of our accounts receivable balance issue by year-end.
The risks, uncertainties, and assumptions include the ability of the CEVA signal processing IPs for smarter connected devices to continue to be strong growth drivers for us, our success in penetrating new markets, specifically non-baseband markets, and maintaining our market share in existing markets, the ability of new products incorporating our technology to achieve market acceptance, the speed and extent of the expansion of the 3G, LTE, and 5G networks, Bluetooth 5, and the IoT space, customers' ramp-up schedule and the impact on royalty revenue, and 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.
With that said, I would now like to turn the call over to Gideon.
Thank you, Richard, and welcome, everyone. We are pleased to report a very successful third quarter with record high revenue, strong execution in licensing, and continued positive royalty trajectory, which led to our highest non-GAAP quarterly EPS in more than four years. In addition, we have, in reach our technology leadership, announced two new DSP-based platforms targeting the fast-growing spaces of machine learning and IoT. Total revenue came at a record high, $17.8 million, up 10% year-over-year. Licensing and licensing-related revenue came at $7.5 million, with 13 licensing deals signed, including comprehensive technology portfolio agreement, the first of its kind in CEVA history. Of the agreements signed, four were for CEVA DSP cores and platform, eight were for connectivity products, and one was the portfolio agreement that I just mentioned. 10 of the agreements were for non-handset baseband applications, and six were with first-time CEVA customers.
Geographically, four of the deals signed were in the U.S., seven in the APAC region, and two in Europe. We ended the quarter with record high bookings, primarily due to two large agreements with market leaders, which I will elaborate on shortly. Royalty revenue came in at a record high $10.4 million, up 36% year-over-year and 8% sequentially, driven by new rollout of handset baseband units and growing shipments of non-handsets by our customer, which reached an all record high of 59 million units. The third quarter licensing dynamics highlight the strengths of our diversified portfolio and the strategic benefit it offers to our customer roadmap. Let me take the next few minutes to discuss three deals signed during the quarter in more detail. The first deal is with a tier 1 handset OEM with in-house baseband design capabilities.
It is our first licensing for 5G mobile broadband to be deployed in next generation smartphone, connected car, [inaudible ], and more. While mass deployment of 5G is about two years out, leading operators such as Verizon, AT&T, and Korea Telecom are pushing their ecosystem to expedite solution despite lack of full consensus on the 5G standard. Our customer is targeting the 2018 Winter Olympics in Korea for its product, which will be the first large-scale event where 5G will be widely used. Our CEVA-XC DSP software-defined radio platform benefits our customer both in terms of shorter time to market and the flexibility to convert to the full 5G standard once it is ratified, without the need to re-spin a costly chip design.
On our last earnings call, we commented that we have a stronger than normal pipeline with a number of customers looking for comprehensive engagement to take advantage of our technology portfolio and unique expertise. In this respect, we are happy to announce that we have concluded two such agreements during the quarter. The first is a portfolio license agreement, the first of its kind for us, with a very large semiconductor company amounting to a few million dollars over the next few years. Under the agreement, the customer has a predetermined annual spending account that can be utilized for chip designs enabled by our off-the-shelf portfolio of baseband, vision, audio, and connectivity IPs. Revenues will be recognized on an annual basis, or once the customer's actual spending exceeds its yearly amount.
We are extremely proud that this key customer has decided to adopt our portfolio of IPs and believe that this business model can be extended to other tier 1 customers who are looking for a one-stop shop for technologies that are at the center of every smart and connected devices. The second important agreement that we signed is with a large reputable OEM for hearing instruments, a space that is expected to grow with aging population. As part of the agreement, we will enhance our Bluetooth technology to support stereo audio streaming from a smartphone or smart TV into hearing devices. Revenue recognition of this deal will be tied to the progress of the design work performed by CEVA. The potential of high-quality stereo audio over Bluetooth extends far beyond the hearing instrument market.
Apple's AirPods, that was recently announced, addresses a similar use case and underlying technology within its in-house W1 chip. We believe that Apple initiative to displace the traditional cordless earbud with a wireless one will drive other smartphone vendors to follow suit, which creates significant market opportunity for us to capitalize on our innovation in this space. On the technology leadership front, we recently announced two new DSP-based platforms that address the performance, cost, and low power needs for next-generation machine learning and cellular IoT applications. The CEVA-XM6 is our sixth-generation image and vision technology addressing the performance and power requirement of deep networks and advanced camera processing for next-generation smartphone, automotive ADAS, surveillance, and more. Powered by new CEVA-XM6, a combo vector and scalar DSP engine.
The platform include an array of value-added technologies, including special purpose co-processors to boost the performance of deep networks and image warping processing. Along with comprehensive computer vision software and our highly acclaimed neural network software framework, CDNN2, which streamlines the development required to add deep learning capability to any embedded device. The CEVA-XM6 platform offers 8x higher performance versus its predecessor, the CEVA-XM4, and up to 25x better power efficiency for deep learning workloads than NVIDIA Jetson TX1 GPU. The second product we announced is a lightweight DSP for cellular IoT, a market forecasted to add more than 1 billion new connections by 2021, according to the latest edition of the Ericsson Mobility Report, which covers services such as smart home, smart utilities, asset tracking, wearable health and environmental monitoring, and more.
The CEVA-X1 DSP is optimized for the stringent low power and low cost modem requirements of the 3GPP IoT standards, such as LTE Cat M1 and Narrowband IoT, as well as the upcoming 5G standard. It also functions as a processing hub for other IoT-related standards such as Wi-Fi, Bluetooth, Zigbee, VOIP, positioning, and sensing workloads. Its unique architecture consolidates DSP and CPU instruction set architecture, allowing for a single core to handle both the modem and the protocol stacks, which save cost of adding separate CPU cores. We believe that cellular IoT and machine learning technologies enable unparalleled efficiencies for consumer and industrial spaces. It will be a key driver for many new products and usage models for mass deployment.
We are highly optimistic of the value proposition that the CEVA-XM6 and the CEVA-X1 platform offer. We will work diligently with our customers to leverage those opportunities. Moving to royalties. Our royalty revenue growth trajectory continued in Q3, with an increase of 8% sequentially and an impressive 36% year-over-year. This was driven by growing shipments of smartphones powered by our DSP and a record non-baseband unit shipment. In respect to non-baseband segment, we are also experiencing a good progress with new customers design and production ramps. The following is a short update on new customer products. Rockchip, a leading Chinese fabless semiconductor, recently introduced two chips enabled by our CEVA-XM4 vision DSP. The RK1108 and the RK1608. The RK1608 is expected to go into production with tier 1 smartphone OEMs shortly.
The RK1108 is the SoC targeting machine vision application for a range of markets, including 360-degree action cam, car DVR, surveillance, and drone. Both of these chips become available just nine months after licensing our CEVA-XM4, illustrating the maturity of our IP and our experience in supporting customers. Silver Spring Networks started production of its Gen5 platform, which is based on our CEVA-XC DSP, with volume gradually building up into 2017. This production ramp is at the back of significant supply contracts with major U.S. and U.K. utility companies. Actions Semiconductor of China launched recently the ATJ2167 SoC, high-end audio chip enabled by our DSP. It enables power-efficient solutions for lossless audio decoding, thus providing immersive audio experience for the mass market of wireless speakers. Fujifilm has announced that it will release the Fujifilm X-T2 mirrorless camera.
It uses our vision DSP for color reproduction for still and video and with high ISO and low noise. HMicro, a California-based company targeting high volume clinical and industrial IoT application, recently announced together with STMicroelectronics the industry's first single chip solution for clinical-grade, single-use disposable smart patches and biosensors. The product, WiPoint, aims to displace about 5 million wired sensors that used in hospital for vital sign monitoring annually, with an advanced wireless disposable sensor based on our RivieraWaves Wi-Fi IP. HMicro is about to enter high volume production with leading medical component suppliers shortly. China's Espressif Systems , a supplier of Bluetooth and Wi-Fi chips, has announced the ESP32, a low-cost Bluetooth and Wi-Fi combo chip. The chip is powered by our RivieraWaves Bluetooth dual mode technology and targets a wide range of IoT use cases.
In summary, we have delivered robust results with record high revenue, our highest quarterly gross margin in more than four years, and new strategic engagements, including our first portfolio agreement. Our broad product portfolio covering connectivity, vision, and audio IP appeals to many customers looking to be part of smart and connected world. Our CEVA-XM6 and CEVA-X1 put us at the forefront of emerging machine learning and LP IoT spaces. On royalties, we continue to expand in the baseband space and are encouraged by the progress in the non-handset baseband space, both with regard to ongoing shipments and production ramps. With that said, I will turn the call over to Yaniv for third quarter financial and fourth quarter guidance.
Thank you, Gil. I will start by reviewing the results of our operations for the third quarter of 2016. Revenue for the third quarter was $17.8 million, our third consecutive quarterly all-time record high achievement. This was 10% higher on an annual basis and 4% higher sequentially. The revenue breakdown is as follows: licensing and related revenue was $7.5 million, reflecting 42% of total revenues, 13% lower compared to the comparable quarter in 2015, but in line with our plans and expectations. Revenue of $10.4 million, reflecting 68% of our total revenue, an impressive increase of 36% on a year-over-year basis, and the seventh successive quarter that we delivered year-over-year quarterly royalty growth. Operating margins were 92% both on US GAAP and non-GAAP basis. The non-GAAP quarterly gross margin excluded approximately $65,000 of equity-based compensation expenses.
Some operating expenses for the quarter were lower than expected and also lower as compared to the first two quarters of the year, mainly due to the magnitude and timing of the research and development grants payment that we received from the Office of the Chief Scientist of Israel. Overall, we recorded OpEx of $12.6 million at the lower range of our guidance. OpEx also included an aggregated equity-based compensation expense of $1.5 million and $0.3 million for the amortization of acquired intangibles of [inaudible]. The total operating expenses for the third quarter excluding these two items were $10.8 million, below the mid-range of our guidance and the lowest quarterly OpEx for the year.
Taxes for the quarter, GAAP and non-GAAP, were around $1 million, a bit higher than the norm due to a one-time tax expense relating to a tax court ruling associated with uncertain tax positions relating to prior years. US GAAP net income for the quarter was $3.4 million, quite similar to last year's comparable quarter of $3.3 million. The diluted EPS was $0.15 for the third quarter this year and $0.16 for the same quarter last year. Non-GAAP net income and diluted EPS for the third quarter of 2016 increased 10% and 9% year-over-year to $5.2 million and $0.24 per share, respectively. Non-GAAP net income and diluted EPS for the third quarter of 2015 were $4.7 million and $0.22, respectively.
These figures for the third quarter of 2016 and 2015 exclude stock-based compensation expenses, net of taxes of $1.5 million and $1.2 million, respectively, and the impact of amortization of acquired intangibles of VoiOs, net of taxes of $0.3 million and $0.2 million for the quarter in both years. Other related data. Shipped units by CEVA licensees during the second quarter of 2016 were 278 million, up 23% and 23% from last year. Of the 278 million units shipped, 218 million units or 79% were for baseband chips, reflecting a sequential increase of 14% from 192 million units from baseband chips shipped and 22% increase from 179 million units shipped a year ago. During the quarter, two customers started mass production of handset baseband. One is incorporated in a premium smartphone, and the second into a mid low-end type smartphone at a significant volume.
The latter have our content is more limited. As such, bears lower ASPs compared to our normal rates. In the non-baseband volume, shipments increased dramatically, 76% sequentially and 29% year-over-year. The increase is due to record high quarterly Bluetooth shipments in Q2 and a continued ramp-up of our audio voice products powered by DSP. The quarterly handset baseband royalty ASP declined 8% sequentially, but increased 15% on a year-over-year basis. This is due to a higher volume of smartphones as compared to feature phones. Our overall corporate blended royalty ASP declined 12% sequentially, but increased 10% year-over-year due to our product mix. As for the balance sheet items. As of the end of September, our cash equivalent balances, marketable securities, and bank deposits were approximately $145 million. Our accounts receivable balance in quarter end was unusually high, approximately $17 million.
Due to one of our large customers' internal changing of its financial and legal entity, which affected the timing of payments to us. We expect this issue to resolve itself by the end of the year. During the third quarter, we generated $3.3 million net cash from operations. Depreciation was $0.4 million, and purchase of fixed assets were $1.3 million, mainly for platform tools for our DSP. At the end of September, our headcount was 281 people, of which 223 were engineers. Now for the back. According to a new report from research firm Strategy Analytics, smartphone unit shipments increased by 6% in Q3 2016 compared to the respective quarter in 2015.
Based on preliminary royalty reports from our smartphone customers for the third quarter shipments, we expect to significantly surpass the market, anticipating more than 100% growth in comparing the third quarter of this year to last year in regards to unit volume. This momentum, along with the continued progress in non-baseband shipments, are set to deliver another record high in royalty revenue, up more than 40% from Q3 of last year and more than 40% over our annual royalty increase for 2015 over 2016. On licensing, as Gideon noted, we are benefiting from having a broad portfolio of IPs to license and our customers' appreciation of these technologies, and expect fourth quarter licensing activity to be in line with recent quarters. Our guidance for the fourth quarter of 2016. Revenue for the fourth quarter is expected to be in the range of $18.5 million-$19.5 million.
This is, again, the highest quarterly revenue guidance in the company's history. Gross margin is expected to be approximately 92% on both GAAP and non-GAAP basis. Overall expenses should be a bit higher than prior quarters, but on an annual basis, within the range of OpEx guidance we set at the beginning of the year. U.S. GAAP operating expenses are expected to be in the range of $13 million-$14 million. Of our expected operating expenses for the fourth quarter, $1.5 million is expected to be attributed to equity-based compensation and $0.3 million to the amortization of acquired intangibles. Net of these two items, non-GAAP OpEx is expected to be in the range of $11.1 million-$12.1 million. Net interest income is expected to be approximately $500,000 for the quarter. Tax rates on a non-GAAP basis, 13%, and on GAAP basis, 18%.
Share count for the fourth quarter, approximately 22.4 million shares, and on an annual basis, approximately 22 million shares. U.S. GAAP fully diluted earnings per share is expected to be in the range of $0.15-$0.17. Non-GAAP EPS forecasted, excluding equity-based compensation and amortization of intangibles, net of taxes, is expected to be in the range of $0.24-$0.26 per share. Overall, for 2016, we are forecasting revenue growth of approximately 18%, which will contribute significantly to our earnings. Both non-GAAP net income and fully diluted EPS are forecasted to grow more of 60% on an annual basis year-over-year. Ed, you could now open the session for Q&A. Ed, operator?
Thank you. We will now begin the question-and-answer session. To ask a question, you may press star then one on your touchtone phone. If you're using a speakerphone, please pick up your handset before pressing the keys. To withdraw your question, please press star then two. At this time, we will pause momentarily to assemble our roster. Our first question comes from Gary Mobley of Benchmark. Please go ahead.
Hi, guys. Congratulations to a strong finish to the year, or what is expected to be. A housekeeping question to start out with. What were the 4G royalty units in the third quarter?
Third quarter, we have 62 million units, which is in line what we said in the earlier conference call last quarter. The surprise comes for the fourth quarter. We anticipate around 80 million LTE units next quarter.
Okay. On that note, I'm assuming that any sort of disruption in what has previously been your top customer with respect to battery issues and whatnot, is going to be more than offset by new avenues of growth, such as another high-end smartphone. Am I summarizing that correctly?
You're right, Gideon. It's all of the above. As you can see from the guidance, things are going well in our side.
Okay. All right. With respect to the revenue recognition for this new portfolio agreement, the way it was described, it sounds as if there's going to be one quarter out of each year in which there will be a significant amount of revenue recognized. Is that going to create, in the future, a lumpy fourth quarter as that licensee either does or does not exceed some predetermined threshold for revenue recognition?
The annual basis is not calendar basis. This is a deal that we signed during the third quarter. As we said, the revenue recognition will be annually. That means that the second quarter of next year, either by then we recognize a part of that revenue or maybe earlier, if that customer reaches its annual spending account, then maybe we could recognize that earlier than the one-year length. The one year is the maximum for the next couple of years, and if they take different technologies that Gideon talked about, then we could potentially recognize it a bit earlier. I'm not sure if you're looking for lumpy third quarter or second quarter in the next couple of years, but for sure, that adds nice backlog color to the second quarter in the next few years.
Okay
hopefully, a bit earlier.
Okay. I noticed the deferred revenue almost doubled sequentially. Was that a function of the recently launched CEVA-X1 and XM6 before they're generally available, or is it a function of some of these new portfolio license agreements?
No, the latter. This is exactly a good example of those deals that we signed. We got paid for some of them, or partially paid, and we haven't yet recognized revenue because of the different deals, either this one, the portfolio item that we talked about. The other big one that Gideon talked about is more service-oriented, and we haven't recognized a big portion of that deal because they only had an initial upfront payment, like most of these licensing activities. That, over the next couple of quarters, will be used as part of our revenue recognition.
Okay. Last question for me relates to 5G market share. You've been humming along here for a couple of years at a mid-30% baseband market share. It sounds as if, as it relates to 5G, you're dominant in baseband processing in base stations. It sounds as if you're growing your footprint on the mobile handset side with respect to 5G with some of these portfolio agreements and whatnot. At this point in time, would you expect your overall baseband share to increase in the future as we transition to 5G and maybe any additional detail with respect to a 40% threshold, a 50% threshold commentary would be helpful.
Yeah. Gideon, about 5G, you said it right. The 5G footprint that we have so far until this deal that we announced today is in the base station. This was the first 5G mobile broadband, which take us to the handsets and automotive and all the things that 5G is targeting for when it comes to high bandwidth. I'm not talking about IoT. The standard for 5G is not yet ratified. People are, and there are not that many companies, but those that have strong technology base are rushing to come to the market with a pre-standard 5G, and our customer is rushing on this one. Going forward, in terms of going beyond our current market share is 36% now, and going forward, we have about 2.4 billion chips going to handset in a year. We are at about 900 million this year.
We are expecting, going forward, because of the emerging of course, 5G, and because of the movement in the mid-range and low-end, where we are strong, to improve our market share.
Okay. Thank you, guys. That's it for me. Thanks.
Thanks, Gary.
Our next question comes from Josh Wolf of Barclays. Please go ahead.
Name change thrown in there. Good morning, guys. I have a question on the, if you look at the non-Bluetooth Unit count that you gave. When you look into that number and the growth there, I think you mentioned audio and voice, do you see specific areas? Is that an eclectic group in that number right now? I know it's a smaller number, or is there anything in there that could potentially be a large blockbuster product, which is in the early stages, and we should see a big ramp-up in 2017 with what you're already selling?
Yeah, I think it's a combination of products. Right now, out of the $60 million, by the way, next quarter, we are looking at the new record high of probably about $70 million. Annualizing that, we're just shy of a $300 million run rate for now. The big driver in the last two years with Bluetooth is the more mature product and the first product [out slum] with the AirPods. The second wave comes from different markets. Audio, we're seeing a lot of interest around and pickup around all always-on type of functionalities, with speakers, and things like that. We have few Wi-Fi deals that should be ramping up next year. I think the most exciting one is a newcomer, is the vision platform.
Gideon mentioned quite a few new names like Rockchip, like others that are now the Fujifilm guy in camera, in different vision applications. I think this is the highest data speed in average for these non-baseband type of devices that we have on our hand, and it's a brand new market opportunity of things picking up. I think these are the three elements other than Bluetooth, they're the most interesting and exciting opportunity around the vision.
Okay. A second question. If you look at the handset market, the baseband market, there used to be specific seasonality where the fourth quarter of the year was the strongest, so you saw that in your first quarter, then it kind of shifted because of customer mixes back to the third quarter, so your fourth quarter was the strongest. The fourth quarter guidance is for significant growth. Where do you think we are, given your new customers? You talked about the large premium customer. Where do you think the seasonality for the handset market is right now? Should we expect a fall from the first quarter, or is there growth in the first quarter?
It's hard to say because today in the handset market, we don't have really seasonality in terms of the calendar year, pre-Christmas, post-Christmas. Each company has its own timelines for introducing new models. Some of them are taking different tactical approaches, to make things even more complicated, we are in a market share game. We are entering into and expanding in the smartphone market, we cannot fully associate it with seasonality. The trend, and that was in focus, the trend is clear. We are expanding in the baseband where we want to expand, this is the smartphone space.
All right. Thank you very much.
Thanks, Josh.
Our next question comes from Matt Ramsay of Canaccord Genuity. Please go ahead.
Good afternoon, guys. Thank you for taking my questions. Gideon, there's obviously a lot in the press and a lot going on at Samsung, which is obviously one of your key partners for 4G and it sounds like for 5G going forward. Looking into next year, maybe you could talk about some of the dynamics from a royalty perspective of what you think could be driven out of Samsung's LSI semiconductor division. You have, obviously, new mid-tier products, potentially some brand disruption and product disruption at the high tier, and some new opportunities with them longer term, but MediaTek winning a couple of sockets there for the first time. There's a lot of moving parts there, and it's a question I get a lot from investors, so any commentary you might have on outlook for that particular customer next year would be really helpful. Thank you.
Yeah. Matt, in general, we cannot comment on 2017. It's too early. The trend that we expect next year is to grow smartphone units. This is the baseband side. Of course, in the non-handset baseband, we see the trend, we see the trajectory in the last quarter. We expect this to continue next year and even after that.
No, that's fair enough. It's early days. I understand that. You made some good commentary, I think, in the prepared remarks about progress with Rockchip, I think I'd be really interested to hear a little bit more about progress on launches of devices that collect royalties from other chipset vendors in China. I know Leadcore has been a partner in the past, I believe Leadcore is licensing some of its modem technology to some other semiconductor groups with smartphone vendors in Asia. I think folks obviously focus on Intel spread from Samsung, but I know there's some other dynamics in play in China, and if you could give us an update on the progress there, that would be helpful. Thank you.
Yeah. Matt, it's a good question what you are asking. I think what we are encouraged to see in the vision side is the smartphone evolution. Smartphone is kind of behind in adopting vision, but if you take iPhone 7, for example, the plus version, where they have a dual camera to get better focus and zoom capabilities, you need a vision processor for this kind of workload. There are people speaking at the conference of 360-degree cameras in smartphone, that's something that you need a vision processor like we offer. Rockchip is one example. What they do is basically a chip that will be a companion chip to the big SoC, whoever it be, Qualcomm, MediaTek, and this will be a co-processor dedicated for the camera. Rockchip is one thing.
We mentioned Fujifilm, we have few other customers going into all sorts of innovation regarding the camera. On processor, we mentioned, on the connectivity company that people may not be familiar because they are not active in the U.S., but they are in China, quite known, and pushing all sorts of consumer products, whether it's wireless speakers, and there's a lot of innovation about Just think about the Amazon Echo. There are in China, tons of companies now doing similar product. Think about wearable devices. All these things, the good thing about us is that our technology base is generic enough and agnostic enough to get all those end products, and this is where we see the design wins and the production wins.
That's helpful. Then just last one from me, and I'll jump back in the queue. Yaniv, you guys have talked about for a while, a target of non-baseband royalty units out to 2018. I believe 800 million was that number at the midpoint. Obviously that guidance was given a couple of years ago and things are moving around. Good to see that number inflecting higher in terms of the units in the quarter and in the guidance, but a progress report towards that number would be great. Thanks, guys.
Sure. If we start rounding up the year, it is a bit early, at least from the royalty perspective and the unit perspective, what we have seen is maybe two or three interesting indicators. First of all, from a volume perspective on baseband, we are probably around 14%-15% higher than last year, with a very interesting chunk of smartphones, probably close to half a million devices, compared to 300 million last year. That is a big increase in the smartphones compared to the feature phone. In the non-baseband units that you asked about, if we were about 167 million last year, we are looking at 20% growth for this year, putting us just short of 200 million. If I look at the fourth quarter and annualize it, the number is much better than 200 million.
If you annualize 70 million, this is more or less the number that we are forecasting for the next quarter, and annualize that, 280 million, puts us in a run rate, or starting run rate for next year, which starts to close the gap towards 800 million. As we start closing, when we put this out, it was 2016, three years out, we have not touched the target, but I think there are very interesting things happening with us in both of them. Even more important on the dollar royalty contribution in 2019, when we will have also the base station royalty kick in. We are slowly closing the gap. We are not yet 100% sure, as Gideon mentioned, about next year volumes and dollars. We will talk about it and do our homework towards the next earnings call as usual.
From a unit progression, I think we are doing quite well this year, both in non-baseband and in the baseband itself.
Thank you very much.
Great. Thank you.
Our next question comes from Suji Desilva of ROTH Capital. Please go ahead.
Hi, Gideon. Hi, Yaniv. Congratulations on the strong results here. Can you talk about where we are in the emerging markets upgrade cycle for LTE smartphones, China, India, and just some indication of the expectations in 2017? I know you can't talk much about 2017, but some thoughts there would be helpful.
Yeah. This is a good question. The interesting point is China is completely LTE central. We don't see too many 3G going there. India, what we understand that all the majority of the phones going there, even though they don't have the network, are LTE phones. We see companies in China trying to build a more low-cost LTE than today, in general, that's a good trend for us because whenever we go to a smartphone and whenever it's an LTE, it drives our ASP higher.
Okay, great. On the overall plan at ASP for the company, should we still think of Bluetooth units predominating and the mix impact the next several years, or will that stabilize?
We never like to say that the ASP is the target or the working tool. It's really a consequence of the volumes in the different products. I don't think we have a goal for ASP. We do have a goal for volume, we do have a goal for different markets, and if they all work out and each one has a different blend ASP, whatever comes in average for the company, this is what we could report. I don't think we have a model that takes it out, or this is a target market for us. It's just more of the outcome of the different, and I think we talked about different volumes in these different markets and the progress that we are seeing this year.
Of course, with [LPIC] kicking in much stronger this year, more than 3X volume from last year, we had a very positive momentum from the ASP perspective. When we add vision, and we had some of those other elements that we talked about in the call, it all gives a different flavor, but it should increase the royalty, and it's a positive effect for us.
Great. My last question, really, these portfolio licenses you've achieved, can you just review for us the significance of them just so it's clear for us? Thanks.
I'm not sure I understood the last question, sorry.
Oh, I'm sorry. The portfolio licenses, if you could just speak of the significance of achieving those.
The portfolio licenses. Yeah. To explain the merit of the idea, how this is working out?
Yeah. What the customers are seeing that drove them to take the whole portfolio?
I think this company that knows us, that did not use the variety of IP, and we looked for a way to simplify the uses of IP. Instead of negotiating each deal separately, and instead of having a lot of infrastructure and a lot of legal teams and accounting teams and business teams deal with us, we tied a much bigger portfolio licenses to then expand with all the different technologies, baseband, vision, audio, connectivity, Wi-Fi, Bluetooth, you have it, and it's sort of prepaid by the corporate. Each division that needs an IP could take it off the shelf in a very simple activation code. That's the merit behind it. We are trying to build an easier access and cleaner access to our technology.
If it does work out for us over the next couple of years, hopefully, we have more products, more chips, and more markets that this large company could utilize our technology and generate new royalty streams for us. This one, this is more strategic. This deal is strategic from a strategic standpoint because other than, I need the processor now. I need the DSP for this specific project, and that's it. It's more like a futuristic way. You take few years ahead and say, "I don't know exactly what I want to do, but I want this product to be available for me so I can use it for my next generation and next-next generation product." It's more like a few years horizon for collaboration with the customer.
Thanks, guys.
Thank you.
Our next question comes from Matt Robison of Wunderlich. Please go ahead.
Thanks for taking my question. I just wanted to ask you about, you mentioned two, I think, if I heard you right, you mentioned two new mass market products, one being low-end. Is that low-end product an LTE product?
We mentioned two, Matt. This is Gideon. We mentioned two handset or baseband product. Both are in high volume. One is a premium handset maker. Another one is a high volume company in the mid-low range product. These are the two new royalty payers and volume productions.
Just asking if the second one, the high volume, lower end, if it's LTE.
It's an LTE front, but our content there is, as we mentioned we prepare now, is lower. Meaning that we are dealing with legacy part and not necessarily with LTE.
Thanks.
Thank you.
Our last question comes from David O'Connor of Exane BNP Paribas. Please go ahead.
Yeah. Good morning, gentlemen. Thanks for taking my question. Maybe the first one, Yaniv, for you on the portfolio licensing deal. What's your expectation when you look across your range of customers at the moment for those type of deals reoccurring? We should be thinking more on an annual basis, or is it going to be a lot, is it more on a kind of every two, three years that those type of deals come along? That's my first question. Second question maybe, going back to, you spoke, Gideon, on the vision co-processor for cameras. It's going to be implemented as a kind of a co-processor chip in the first generation.
I'm just wondering when you look at that over time, is that going to be something that's going to be integrated into the ISP longer term, or is it going to sit there as a co-processor in the next couple of years? That's my second question. Maybe, if I could get in one more. A lot of talk at the moment around the vision systems in ADAS and into automotive. I'm just wondering how your discussions trending with kind of auto OEMs or any auto suppliers at the moment. Thanks.
The problem is when you ask too many questions, it's like a pistol. We forget the first question.
I remember.
Yeah. It's like food. When you smell and feel it, you get the appetite.
Yeah. Okay, that's fine. I remember now. With regard to the portfolio agreement, I think the way to look on the portfolio agreement is not on the licensing line. The licensing line is predetermined and well-defined on dollars, on the amount, on the revenue recognition. The benefit is that the other customer that going to use us or plan to use us with a lot of our products, and the way to look into this one is the potential for royalties coming from this transaction going forward. This is the benefit regarding the portfolio.
I think the question was: Do we want this to be repeated once a year or once a few years? Of course, as I mentioned, with the food comes the appetite, and we think it could be a great opportunity, especially with large companies. It doesn't work with a startup. It doesn't work with a first-time image design win because they don't know yet how many chips they're going to have or how successful that product is. There's much limited with consolidation in the space. There is more limited type of companies, but for sure, this is something we'll try out with others as well.
Hard for us right now to commit if this is one year or one is a couple of years, but there's no doubt that the sales people are aware of this deal, we are aware of such a deal, and as we mentioned and talked about here at length today, we see a lot of advantages for that. We'll try to copy it. Can't promise, at least today, that this is something that we'll have for sure next year as well.
Let me now, because we are almost out of time, I want to respond to the vision question because this is an important question. As you pointed out, currently it's a coprocessor that stays in between the ISP chip and the SoC, whether it's a Qualcomm, MediaTek, we don't know exactly. Going forward, this technology could be either in the SoC, be integrated in the SoC and depends on customer choice, could be integrated into the sensor chip or could be a coprocessor. A chip kind of in between all these things. We are agnostic in regards to all the different locations there, and for us at this stage, what encourages us is the use cases that customers are coming out of how to use vision processor and deep learning to improve vision, improve the camera performance or the camera sensing. Hello, are we on?
Mr. O'Connor, did you have any more questions?
Yeah, just my follow-on question on the ADAS, the vision systems into ADAS and auto. That was my third question.
Yeah.
What kind of discussions are you having with auto OEMs or suppliers, and how is that trending? Thanks.
Yeah. Sorry for forgetting this question. This is also an important question. When it comes to ADAS automotive, autonomous cars, this is a different market than the consumer that we are used. The timelines are different and the acceptance criteria is different, but it is very dynamic area. We have a plan that we are following in trying to engaging customer. There are only a few companies. One of them started to ship this quarter. It's in aftermarket and OEM market, but it is a shipment there. We are optimistic about because we see the fruit.
This does conclude our question and answer session. I would like to turn the conference back over to Richard Kingston for any closing remarks.
Thank you everyone for joining us today and for your continued interest and support for CEVA. We will be attending the following upcoming conferences and invite you to join us there for an update. The Bernstein Technology Innovation Summit on November 7th in New York, the second annual ROTH Technology Corporate Access Day on November 16th in New York, and Barclays Global Technology Conference on December 7th in San Francisco. For further information on the event and other events that we're hosting, please visit our investor section of our CEVA website. Thank you and goodbye.
The conference is now concluded. Thank you for attending today's presentation. You may now disconnect.