Good morning, welcome to the CEVA, Inc. third quarter 2018 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 zero. After today's presentation, there will be an opportunity to ask questions. To ask a question, you may press star then one on your telephone keypad, to withdraw your question, please press star then two. Please note this event is being recorded. I would now like to turn the conference over to Richard Kingston, Vice President of Market Intelligence and Investor & Public Relations. Please go ahead, sir.
Thank you, Rocco, good morning, everyone, welcome to CEVA's third quarter 2018 earnings conference call. I'm joined today by Gideon Wertheizer, Chief Executive Officer of CEVA, Yaniv Arieli, Chief Financial Officer of CEVA. Gideon will cover the business aspects and the highlights from the third quarter and provide general qualitative data. Yaniv will cover the financial results for the third quarter also provide qualitative data for the remainder of 2018. I'll start with the forward-looking statement. Please note that today's discussion 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 and full year 2018, optimism about a strong momentum, CEVA's ability to capitalize on trends associated with wireless-based connectivity and NB-IoT product, 5G healthy licensing environment and demand for CEVA's product, optimism about sustained growth in non-handset baseband product lines and customer production ramp-ups, positive forecasts from IC Insights and Ericsson Mobility. For information on the factors that could cause a difference in our results, please refer to our filings with the Securities and Exchange Commission.
These include the ability of CEVA signal processing IPs for smarter connected devices to continue to be strong growth drivers for us, the traction with edge technology for AI, our success in penetrating new markets and maintaining our position in existing markets, the ability of new products incorporating our technologies to achieve market acceptance and offset the maturity of the handset market, the speed and extent of the expansion of the 5G network and wireless connectivity, artificial intelligence, LTE, IoT, the IoT space generally, our ability to execute more broad portfolio license agreements, customer ramp-ups and schedules that impact on royalty revenues. CEVA assumes no obligation to update any forward-looking statements or information which speak as of their respective date. In addition to the financial results prepared in accordance with the generally accepted accounting principles or GAAP, we will also present certain non-GAAP financial measures today.
CEVA's management believes that in addition to using GAAP results in evaluating our business, it is also useful to review results using certain non-GAAP financial measures. Investors and potential investors are encouraged to review the reconciliation of non-GAAP financial measures with their most direct comparable GAAP financial results, which can be found in the earnings press release issued today. A copy of today's press release for the third quarter ending September 30, 2018, and the related financial tables and management commentary, which were included in our current report on Form 8-K filed today, also can be found on the investor relations portion of our website shortly. Before handing the call over to Gideon, I would like to remind you that CEVA adopted the new revenue accounting standard known as ASC 606 as of January first, 2018.
Under the new standard, our royalty revenue represents what our customers shipped during the third quarter of 2018 for our best estimate for such shipment. The numbers stated on this call for the third quarter are based on ASC 606, unless otherwise stated. However, as our Q3 2018 financial results are not directly comparable to our Q3 2017 financial results, which were reported under the old revenue recognition accounting standard, known as ASC 605. We will also provide you today our Q3 2018 financial results as reported under ASC 605 to allow for an apples to apples comparison on a year-over-year basis. We will have this dual reporting approach throughout 2018 as required by the Financial Accounting Standards Board. With all of that said, I will now hand the call over to Gideon.
Thank you, Richard, and welcome everyone. Our third quarter revenue came in line with our expectations, demonstrating strong royalty revenue recovery and solid execution in licensing. Total revenue was $21.4 million, of which licensing and related revenue was $9.8 million, and royalty revenue was $11.6 million. Royalty revenue reflected strength across the board with a step-up in contribution from a wide deployment of our advanced DSP technology within recently launched flagship smartphone and the continued growth of shipment in the non-handset category as new and existing customers roll out new product into the market.
The strengths of our technology portfolio led to another quarter of sustained licensing revenue with 13 agreements executed across multiple end markets included in the two strategic ADAS space. China in particular show good dynamics with nine agreements out of the 13 deals concluded in the quarter with vibrant Chinese semiconductor companies targeting variety of IoT devices for consumer and industrial application. According to recent analysis from research firm IC Insights, China-based fabless IC firms are expected to account for 19% of the global total of new place foundry sales in 2018, up from 9% in 2016 and from about 13% last year. We are experiencing, in particular, strong demand for our Wi-Fi, Bluetooth, and cellular product as wireless IP expertise is scarce in China and must have for IoT product.
Let me take the next few minutes to provide you with foresight on licensing landscape, elaborating, in particular, on the wireless space. The opportunities and the customer engagement we have experienced are authentic indicators of product trends and the precursor for new royalty revenue stream. According to the latest Ericsson Mobility Report, short-range connected devices are expected to increase from 6 billion in 2017 to 17 billion in 2023, and long-range IoT from 0.6 billion to 2.4 billion, a total install base of 20 billion by 2023. Short-range IoT devices are primarily smart home and office product using Bluetooth, Wi-Fi, and Zigbee. Long-range IoT include devices that connected to the internet via cellular IoT, standards such as NB-IoT. CEVA is a prime supplier of wireless connectivity technologies for IoT. Our offering is comprehensive, vertically integrated, and include both short-range and cellular technology.
We are benefiting currently from the proliferation of new consumer-oriented wirelessly connected product like headphone, smartwatches, smart speaker, and range of wirelessly connected home appliances and home entertainment devices. The scope of short-range wireless connectivity extend beyond consumer and peer-to-peer communication to indoor navigation, asset tracking, and control, which almost double the total addressable market. With our NB-IoT solution, our market reach extend to emerging verticals such as automotive, smart cities, industrial, driven by national initiatives like Industry 4.0 and Made in China 2025. To foster the NB-IoT opportunity and development, we launched recently our second generation full solution, the Dragonfly NB2. Dragonfly NB2 complies with the latest 3GPP standard Release 14 and is software upgradeable to 5G. It incorporate in one platform all the different disciplines required for NB-IoT solution, including our latest CEVA-X1 DSP, hardware acceleration block, GPS, protocol stack, and software and RF IP.
The Dragonfly NB2 platform reduces dramatically the entry barriers for companies looking to get into the cellular IoT space, enabling them to focus on their excellence rather to develop cellular competencies, which are difficult. As for 5G, the scale of activities of operator and equipment deployment and launches of services significantly expedited in the last few months, indicating a meaningful transition play similar to how 4G was. The FCC published recently a 5G action plan with the mission to place the U.S. at the forefront of developing and deploying 5G technology. Dubbed as 5G FAST Plan, it contains policies to streamline spectrum and infrastructure build-out, the key hurdles for fast deployment.
In a recent earning call, Nokia CEO said that they see excellent order intake reflecting growing market demand, which implied a 30% growth in backlog compared to the beginning of the year, and also mentioned that AT&T selected Nokia as one of their 5G suppliers. He also commented that the deployment of the cost-efficient ReefShark chip enabled by our DSP platform will pave the way for higher operating margins for Nokia network. ZTE, using our technology, achieved successful test results from the third phase of national testing. The first phase is seen as the last step before ZTE can proceed with its 5G commercialization plan. The Ministry of Industry and Information Technology, MIIT, commented that 5G devices will be ready for commercial use in China in 2019.
As we stated in prior calls, CEVA is in a unique position to capitalize on 5G, both at the base station RAN and within the devices. 5G revolutionizes the network architecture and design. It requires much denser base station with as many as four to eight times more cell tower per square kilometer in comparison to 4G. On the device side, 5G enables new usage models in car manufacturing, health, and more. We are experiencing a solid licensing pipeline composed of both incumbents that did not use our technologies for LTE, and newcomers that can make use of cellular technology to reduce time to market. In summary, the third quarter financial results and business execution reflect a healthy demand for our products and successful production ramps by our customers.
We continue to expand our design wins and pipeline, capitalizing on the rapid proliferations of wirelessly connected IoT devices, in particular in China. These wins are across multiple verticals and pose high volume-based royalty opportunities. Together with our 5G end-to-end offering, we are a one-stop shop to any newcomer or incumbents for wireless technology. Our sensing and AI technology for computer vision voice are complementary to our wireless technologies and provide us with increased content and cross-sale opportunities. On wireless, we return to sequential growth to the successful launch of the latest smartphone product line from prominent OEM and the sustained growth in shipment within our non-baseband product line. With that said, let me turn the call over to Yaniv to discuss our financials and guidance.
Thank you, Gil. I'll start by reviewing the results of our operations for the third quarter of 2018. Revenue for the third quarter based on ASC 606 was $21.4 million. The revenue breakdown is as follows: Licensing and related revenue was $9.8 million, reflecting 46% of our total revenue, 30% lower as compared to the third quarter 2017 all-time record high. Royalty revenue was $11.6 million, reflecting 64% of our total revenue, a decrease of 8% on year-over-year basis compared to $12.6 million for the third quarter actual shipment that were reported in the fourth quarter of 2017, following the revenue rules under ASC 606. As a reminder, the $12.6 million, a record high, included a one-time catch-up fee of $0.9 million due to a customer audit. Quarterly gross margins were 91% on US GAAP basis and 92% on non-GAAP basis.
Our total operating expenses for the quarter were below our guidance at $17.3 million due to lower SG&A costs associated with marketing activities and comp-related provision. OpEx included an aggregate equity-based compensation expense of approximately $2.3 million and $0.2 million for the amortization of acquired intangibles of RivieraWaves. Our total operating expenses for the third quarter, excluding these, were $14.7 million, also below our guidance. US GAAP net income and the diluted EPS for the third quarter were $2.5 and $0.11 respectively. Our non-GAAP net income and diluted EPS for the third quarter of 2018 were $5.2 million and $0.23 respectively. Other related data. Shipped units by CEVA's licensees during the third quarter of 2018 were approximately 263 million, up 19% sequentially and down 8% from Q3 2017 actual shipments reported in the fourth quarter of 2017.
Of the approximately 263 million units shipped, 165 million units or 63% were for handset baseband shipped, reflecting a 24% sequential increase and a 19% decline on a year-over-year basis. Non-baseband volume shipments reached another record of 98 million units. Up 11% sequential and 22% on a year-over-year basis. The music shipments continued to be strong and broke another record with 83 million units shipped in the quarter. For our balance sheet items. As of September 30th, 2018, CEVA's cash equivalent balances, marketable securities, and bank deposits were approximately $167 million. We continued our active buyback program and repurchased about 216,000 shares during the quarter for approximately $6.3 million. During the second quarter, our board of directors approved the expansion of the existing buyback plan, and as of September 30th, we have approximately half a million shares available for repurchase.
Our adjusted ASC 606 DSOs for the third quarter of 2018 continued to be low at 41 days from prior quarter level of 48 days. During the third quarter, we generated $0.3 million of net cash from operations, with depreciation of $0.7 million, and purchase of fixed assets was $0.7 million. At the end of September, our headcount was 329 people, of which 266 were engineers. For the guidance for the rest of the year. On royalties, as we forecasted early, an even slightly better result, we managed to record an impressive 56% sequential growth in the third quarter and believe that the fourth quarter royalty revenue is estimated to be similar or slightly better. On licensing and related revenue, we continue to experience healthy demand for our products, and we'll keep the same pace quarterly target, past quarterly target, as we've demonstrated over the last two years.
Specifically for the fourth quarter of 2018, gross margin is expected to be similar to the third quarter level, with approximately 91% on GAAP and 92% on non-GAAP basis. Overall, OpEx is expected to be lower than the third quarter level and lower than the first two quarters of the year, and is anticipated to be in the range of $16.4 million-$17.4 million. Of the anticipated OpEx for the fourth quarter, $2.4 million is expected to be attributable to equity-based compensation expenses and $0.2 million for the amortization of acquired intangibles. Our non-GAAP OpEx is also expected to be lower than prior quarter levels in the range of $13.8 million-$14.8 million. Net interest income is expected to be approximately $0.8 million. Tax rate for the fourth quarter, approximately 17% on GAAP basis and 11% on non-GAAP basis.
Share count for the fourth quarter is expected to be in a similar level to the third quarter number. Operator, you could now open the floor for the Q&A session, please.
Thank you, sir. To ask a question, you may press star then one on your touch-tone phone. If you are using a speakerphone, we ask that you please pick up your handset before pressing the keys. To withdraw your question, please press star then two. Today's first question comes from Gary Mobley of Benchmark. Please go ahead.
Morning, guys, or good afternoon, I guess. I want to start with a question about the China mobile handset market. What is your sense of your market position in China through your largest licensed relationship, in particular with Spreadtrum?
Gary, when it comes to China in the last two or three quarter, we iterated, we said that there is softness within this specific customer. This is not over, although we saw improvement in this quarter. The target market for this customer is the low-tier in China, in India. Far, the bottom is over, is behind us, we would like to see much better.
Okay. My sense is that you saw a nice contribution or at least some contribution from 5G cellular base station SoC in your royalty contribution. Can you share with us what the unit contribution was, what your non-mobile handset baseband royalty unit rate was in the quarter with that contribution, and what your outlook is for contribution from your two main 5G base station SoC licensees?
About base here, I'll let Yaniv to comment.
We don't see today a meaningful 5G base station royalty. As I said, when it comes to ZTE, they passed a very advanced stage, which will allow them to start commercializing, meaning installing a base station in a more substantial way. Up till now, they were in testing, deployment, and stuff like this. Nokia, as we said, and they are saying, the ReefShark, and this is where our entry point there is key for their operating margins there. So far, they are not installing. It not just depends on the product, it depends, as you know, in many parameters, operator deployment. To me, it looks like more 2019, you start see a few things and to more extent 2020.
Okay. With respect to your license pipeline, can you share with us what has been recently made generally available in your new product development, such as NeuPro? Is that generally available, and if that you're able to recognize some license revenue? I noticed that your accrued revenue, $16.5 million, was up substantially on a quarter-over-quarter basis. Is that an indication of the license revenue backlog?
Hey, Hi Gary. Let me start with the technicalities. No, not necessarily. Two things increased it. One is higher royalties, and because of the new rule, all the $11.6 million that we recorded were all recorded after the quarter, and we did not invoice our customers in September, but in late October and November, and this is why they're accrued. That number, of course, is higher than the first two quarters of the year. On top of that, we have few deals that were very back-end loaded, and we only invoiced in October after the quarter ended. This is just a technical number. It will continue to follow with us on the royalty side. The licensing is just a question if the timing of the invoices, in this case, it was just a bit of a back-end loaded quarter from that aspect.
It has nothing to do with the newer technologies that Gideon mentioned in his prepared remarks. We are developing them, we're offering them, and we have licensed already few agreements around those technologies, and we continue to be quite excited about their opportunities, both on the licensing and later on the royalty side.
Okay. All right. Thank you, guys.
Thank you.
Today's next question comes from Mike Walkley of Canaccord Genuity. Please go ahead.
Great. Thank you. Congratulations on the strong sequential royalty growth. My question is just on the Q4 for kind of flattish royalty growth. Can you kind of walk us through your thought process and the puts and takes, with Nokia expected to have a very strong shipment for new base stations with ReefShark, I would've thought that might have helped, along with maybe the seasonality for the new iPhone. Can you kind of just walk us through puts and takes why you see Q4 flattish with Q3 for royalties?
Yes, sure. In our expectations for the fourth quarter, surprisingly enough, we do not take into consideration any revenues from Nokia yet. Yes, they have a lot of interesting opportunities, pipeline deals closed, and we follow them. We do not know yet to specifically estimate how many chips translate on a quarterly basis, and even at this point on an annual basis. We are waiting for these new royalty reports to kick in. I'm not sure if we answered Gary in the prior question, this also implies for the third quarter. We are still waiting for that ramp up. We believe it is going to come. We don't know the magnitude, and we don't know exactly the timing of these ReefShark chips.
As soon as we get those reports, like the first report, of course, we'll be happy to share that and to add it to our, with our numbers or estimates going forward. Q4 is really built on the existing interesting dynamics that happened this year for us. Some of the positive side, like the well-known U.S. high-end devices that were launched recently with Intel and CEVA powering them. On the other hand, the softness in some of our largest customers in China, not necessarily to the Chinese market, but to the rest of the world and India that they ship into and other parts. They continue to be out there with some improvements, but not to the same levels that we have experienced over the last couple of years. We hope that will again kick in at the later stage.
For now, we have bundled all that. We do add into consideration, I think as Gideon mentioned, ZTE, which is already in production, that is very fast, and we saw initial reports and we are sure now that Q4 will have a base station revenue coming from them. We don't have that same indication yet from the Nokia standpoint. We are very happy that after two quarters of $7.5 million and different explanations around that, we are back to how we started the year with a different allocation of royalties. Second half is probably much stronger than the first half, and we even believe that level could be even stronger than in Q3, but not yet with the full engines and royalties kicking in for us.
That's not happening yet in Q4, although with better estimates for non-baseband, from base stations now back to business or partially back to business, and from the handset side to some degree.
Thank you. Just a quick follow-up question from me. How do you see maybe the annual base station opportunity into 2019 and 2020 from your current licensees? Thank you.
It's hard to answer this kind of question at this stage, because it depends on primarily the deployment or the investment that the operator are going to make in commercializing. A lot of operators are speaking about 5G and signing contracts, but the question is when they will do it and to what extent, how fast it's going to be. What we have today is basically the revenues that we are collecting, and Yaniv mentioned, is 4G. We know better, in my opinion, in 2019, could be first half, we start seeing things. When this ball will start rolling, it will roll, and it could roll fast.
Okay. Thank you.
Thanks.
Our next question today comes from Matthew Ramsey of Cowen. Please go ahead.
Hey, this is Joshua Buchalter on behalf of Matt. Thanks for taking my question, congrats on the results. I guess I'll try to ask the previous question in a slightly different way. If we think about some of the qualitative commentary you've heard on 5G this quarter, it seems like there's a lot of activity picking up. Can you maybe help us understand how you would view the cadence of your customers' products and how they would ship versus some of the things we've seen built so far? Thank you.
Hey, sorry, could you repeat that again?
Hi, I'm sorry. I basically was just asking, given some of the commentary of pick-up and build this recent quarter, could you help us understand the cadence and how your customers' products would ship in relation to those, maybe versus some of the long-haul infrastructure being laid down?
I hope I understood your question, but I believe you refer to 5G, right?
Sorry.
Yeah, that's correct?
Yeah, that's correct.
Our play in 5G, as you know, is in both ends. It's in base station side and in the handset, in the device side. What we see today is the following. In the base station side, they are more advanced. They have the product, they sign contract, and now it's a matter of the pace of commercial deployment because our technology is going into commercial deployment. When you have all those testing and initial deployment, these are not where our products are going. Our products are providing power efficiency and cost efficiency, and people are putting in when they get a commercial deployment. At this stage, we don't get royalties from this pattern of deployment. We believe we're going to see this coming in 2019 based on what we are discussing with customers, based on what we see from public announcements.
On the device side, I think they are behind. We do have customers that use our 5G technology called 5G. Whether it's going to be a meaningful deployment 2019, I saw just recently analysis that they say only 1% of the total handset market will be 5G next year, and will be 20% in 2020. That, I believe, give you colors of what to expect from 5G on the handset side.
Okay, that was helpful, and sorry for the confusion. My follow-up, if we sort of back into the royalty per unit number in your royalty revenue this quarter, it looked like there was a nice sequential step up. Could you maybe talk about some of the drivers of that and how we should think about that going forward? Thank you.
Yeah. That could change, of course, from one quarter to the other, and we continue and explain that we look at the total dollar value as a strong and a key indicator of the healthiness of our royalty. As long as the dollar increases, we're happy with that. With that said, this quarter specifically is a different mix between high-end Phones and newer technology versus lower-end devices with some lower ASPs. What we have witnessed in the beginning of 2018, which will continue throughout the year, is lower volume shipments from one of key customers specifically in China, and this is more lower mid-range type phone. This is a very strong pickup from September this year and the pre-order for that segment, which had newer technologies inside, higher ASPs, and higher content from our point of view.
The mix was favorable this time around, meaning, of course, much higher volume or the highest we ever had offset that as an average, but on the other hand, adds more dollars to the overall dollar content of the royalty. I think we're going to have mixed numbers or calculations around it, and as soon as the 5G stuff and the base station stuff that Gideon talked about kicks in and, as we said, the ball is rolling, not just starting to roll, but is rolling downhill, then those ASPs are much, much higher because we're talking about much more expensive type of chips that will be in, that have a selling price of $100, $155, and that's a much different ASP than what we have been used to in the past.
Thanks very much.
Thank you.
Ladies and gentlemen, as a reminder, if you'd like to ask a question, please press star then one at this time. Today's next question comes from Suji Desilva of ROTH Capital. Please go ahead.
Hi, Gideon. Hi, Yaniv. Can you help us, first of all, in terms of looking ahead to 2019, perhaps the royalty growth, what the opportunities are, how you'd rank smartphone versus wireless infrastructure versus IoT connectivity, video, imaging, voice, just to give a sense of where you think the pockets of growth will be in 2019.
Yeah. Hi, Suji, good morning. When it comes to 2019, of course, we will not quantify it. We still need to do the work. We are not firing in all the cylinders that we can in terms of the product that's coming. Base station, we expect to see a ramp next year. Again, the timing and the pace is something that we will have to see. That's waiting for prime time, I should say. We have the non-handset baseband. You mentioned computer vision, you mentioned there will be AI, Bluetooth, Wi-Fi. The way we see it is one big basket. The good news is they are all different industries, different customers, and we are not biased to one trend or one crisis in the market.
The idea is to collect as many. That's the importance of licensing and the fact that we have 13 agreements means that we have more than 13 new projects starting. Some of them could be in 9 to 12 months in the market. Basically, we're going to see new SKUs coming from existing ones, whether it's going to be consumer, industrial, Bluetooth. All of them should come together there and continue the nice ramp that we see almost every quarter. This will continue at the pace, whether it's a stronger pace than today or what pace, we will have to see, and we'll do the work by the end of the year. The flagship model that we do, we are there.
That's an important contribution for us, and the fact that it's basically one supplier there, that's a dramatic change for us, and we'll have to see because they have different order patterns than others, but that's a dramatic change for us and we hope to capitalize it.
Okay. Specifically on the wireless infrastructure, I recall you had a third customer as well. Just remind me if that's correct, and if so, what's the status of that third customer relative to ZTE and Nokia?
It's a 5G design, Suji, good morning. As we talked about earlier, that's not yet in production or in deployment. That's a little bit of a later on, not even sure 2019, but could be a 2020 type of event for us. Back to the original question about the 2019 royalties, I think what Gideon explained on a product-by-product basis, that we have all the interesting licensing activity that we had over the last year working well for us. I don't think we have done yet the homework for 2019, and we'll do it in the next earnings call and try to give more color of how this all plays out in what quarters and what product lines.
I think we just lined them up and gave each one the explanation, the opportunity, but it's a bit early to quantify the contribution from each one at this point.
Okay, appreciate the color. Thanks, guys.
Thank you.
Our next question comes from David O'Connor of Paribas. Please go ahead.
Yeah. Good morning, guys. Thanks for taking my question. Maybe Gideon firstly, one on China and Spreadtrum in particular. When China rebounds, excuse me, from the soft handset trend just in currently, do you expect customers like Spreadtrum to have a similar market share, or do you think they could be losing share as well, and the dynamics of any rebound could be different to what we've seen in the past? Thanks.
The handset market is very dynamic. One thing I can tell you, Spreadtrum is a very powerful company, and they are in a position to compete with MediaTek and Qualcomm and almost all the tiers there other than, of course, the high-end stuff. They had a customer loss, and they are now, in a way, we see all those new announcements that they are recovering, but we will have to see. Overall, we cannot ignore, the overall handset market, the growth there is limited. The fight is all about gaining share. It's a zero-sum game. As I said, Spreadtrum is extremely determined and very powerful to win a larger share than they have today.
That's clear. Thanks for that. Maybe just to follow up on NeuPro. Can you talk a small bit about the licensing pipeline of NeuPro in particular, and has that changed from the last quarter? What type of customers you're engaged with there, please?
Yeah, that's a good question because when people are, I believe, when people are looking on the AI industry in general, they'll get confused. Many players, so many boxes being built, different unique architecture. The way we see it, there is the first wave of AI customers. Those AI customers or those companies that are vertically integrated, meaning they can develop not just the chip, but also can develop the application that runs the AI application. There are not that many. I'm talking about the edge side and the embedded side, not the cloud. The cloud is completely different, and we don't play in the cloud for now. There are companies in the automotive space, there are companies in the digital camera, there are companies in the drone. You have those companies that know how to develop applications.
In this space, meaning AI vertically integrated people that know and that understand the, but can develop AI application, CEVA position is very strong. People were using our technologies, not just at the context of the NeuPro, which is a hardware-based technology, but also in our DSP. We gain with this customer a lot of experience, especially on the software that we call a CDNN. Just to give you example, what does it mean? One of our competitors announced in a recent technology event, performance metric about his AI engine, and our performance is 3 times better than this competitor because of our software. That's the first wave of customers we are engaging with, and some of them are basically upgrading from the initial DSP base to the NeuPro area.
Now we are now seeing a lot of newcomers, people that see AI as just a block in their chip, which they rely on anybody else from their customer to develop their application. Here's the expectation that 90% of the SOC in the world will have this kind of AI engine. Those companies are a bit slow in their decision-making, but they are catching up, and we see going forward, when it comes to NeuPro development, we see all those companies that see AI as an engine and don't see themselves developing the application. If you ask me how the pipeline is composed, you see all those newcomers coming, and here again, our software becomes critical because without having the software, their customers cannot develop such application.
That's helpful. Thank you.
Ladies and gentlemen, this concludes your question and answer session. I'd like to turn the conference back over to Richard Kingston for any closing remarks.
Great. Thank you for joining us today, everybody, and for your continued interest in and support of CEVA. We will be attending the following upcoming events and invite you to meet us there. The Benchmark Discovery One-on-One Conference in Chicago on November 29th, and Barclays Global TMT Conference in San Francisco on December 5th. Please visit the investor section of our website for further information on these events and other events we will be attending. Thank you and goodbye.
Thank you, sir. Today's conference has now concluded, and we thank you all for attending today's presentation. You may now disconnect your lines, and have a wonderful day.