Quicklogic Corp (QUIK) 2017 Q2 法說會逐字稿

Ladies and gentlemen, good afternoon. At this time, I would like to welcome everyone to QuickLogic Corporation's Second Quarter 2017 Results Conference Call. (Operator Instructions) Today's conference call is being recorded.

At this time, for opening remarks and introduction, I would like to turn the call over to the company's Investor Relations representative, Ms. Kirsten Chapman of LHA. Ms. Chapman, please go ahead.

Thank you, Latif. Welcome, everyone, and thank you for joining us today for QuickLogic's Second Quarter 2017 Results Conference Call. With us today are Brian Faith, President and CEO; and Dr. Sue Cheung, CFO and Vice President of Finance.

Before we begin, I will read a safe short harbor statement. Some of the comments QuickLogic makes today are forward-looking statements that involve risks and uncertainties, including, but not limited to stated expectations relating to revenue from new and mature products, statements pertaining to QuickLogic's future stock performance, design activity and ability to convert new design opportunities into production shipments; timing and market acceptance of its customers' products; our future evaluation systems; broadening our ecosystem partners, expected results and financial expectations for revenue, gross margin, operating expenses, profitability, and cash. These statements should be considered in conjunction with the cautionary warnings that appear in QuickLogic's SEC filings. For additional information, please refer to the company's SEC filings posted on its website. Investors are cautioned that all forward-looking statements in this call involve risks and uncertainties and that future events may differ materially from statements made.

These forward-looking statements are made as of today, the day of the conference call, and management undertakes no obligation to revise or publicly release any revisions of forward-looking statements in light of any new information or future events.

The conference call is open to all and is being webcast live. We will start today's call with the company's strategic update from QuickLogic CEO, Brian Faith, and then CFO, Sue Cheung will provide financial results and guidance. Brian will deliver closing remarks and open the call to questions.

At this time, it is my pleasure to turn the call over to Brian Faith, President and CEO. Please go ahead, sir.

Thank you, Kirsten, and thank you, all for joining our Q2 2017 conference call. We have made significant progress since our last conference call. This includes an increase in the quantity of new engagements, notable improvements in the efficiency of our engagement process and the number of engagements we have converted to design wins. We have also received anticipated production start dates for a number of these design wins and late-stage engagements. With this momentum, we are well-positioned to initiate sustainable growth starting with Q4 2017 and achieve our target operating model in 2018.

However, key designs that we anticipated fueling our targeted 50% growth this year have been shifted 1 to 2 quarters forward, and that has lowered our outlook for Q3 and Q4 2017.

While reducing our outlook for 2017 is clearly disappointing, our growth drivers are intact, and our design win momentum has improved significantly during the last quarter. I think this momentum will continue to build and look forward to sharing with you today why I'm more optimistic than ever in the future of QuickLogic and why I believe so strongly we are well-positioned to deliver sustainable revenue growth, beginning next quarter.

Let's start today with an update on our embedded FPGA IP business. To briefly recap, our IP business model has 2 licensing elements. The first element involves licensing semiconductor foundry partners to manufacture ICs that include our IP. The second element involves licensing semiconductor companies and OEMs that design the ICs incorporating our IP.

In the case of semiconductor companies and OEMs, licenses are negotiated for each specific IC that includes our IP. In addition to the license fee, semiconductor companies and OEMs will also pay a royalty for each IC that is shipped to end customers or, in the case of OEMs, used internally. Due to the long development cycles common for semiconductor devices, I think it's reasonable to model, royalties will begin about 12 to 24 months after a license agreement is signed.

So far, we have licensed 2 semiconductor foundries. Last year, we announced a license agreement with GLOBALFOUNDRIES, covering its 65-nanometer, 40-nanometer and upcoming 22-nanometer FDX nodes. Prior to this agreement, we qualified our eFPGA IP on GLOBALFOUNDRIES 65-nanometer and 40-nanometer nodes. We anticipate the test chip for the new 22-nanometer node will be taped out in Q4 2017 and qualified in Q1 2018.

Earlier this year, we announced a second semiconductor foundry manufacturing license agreement, which we can now say is with SMIC for its 40-nanometer process node. We successfully taped out the test chip on schedule in Q2, and are targeting completion of our test chip qualification with SMIC later this quarter.

In June, we announced that Bernie Rosenthal joined our advisory board. Bernie was a Co-Founder of Tensilica, which was a high-profile semiconductor IP company prior to its acquisition by Cadence in 2013. Bernie was the architect of Tensilica's business model and led negotiations for over $80 million in license contracts with top-tier semiconductor companies and OEMs. Bernie is supporting us as we refine our IP business model and enabling us to leverage his portfolio of high-level industry contacts.

In conjunction with our new eFPGA support center in Taiwan, this is helping us drive a very efficient engagement process with our targeted partners and customers. We have ongoing engagements with a number of leading semiconductor companies, and we believe we will sign license agreements with at least one and possibly several of these companies later this year.

As you may recall from past presentations, license fees are typically paid up front. This means license revenue will flow upon payment to our balance sheet and cash flow statement. However, for our income statement, revenue is amortized ratably across the term of the license and recognized over time. Due to the fact, this amortization schedule can vary considerably from deal to deal, we are not yet in a position to forecast when the license revenue we expect to invoice during Q4 will be recognized on our income statement.

Now let's move to sensor processing. We are seeing some distinct improvements in the outlook and use cases that are very favorable for QuickLogic. A recent report by Daniels Associates predicts the discrete sensor hub market will be roughly 680 million units in 2018, growing to over 1.5 billion units in 2022. Roughly only 1/3 of this volume is in the consumer market, with the remaining volumes spanning multiple markets across automotive, industrial, health care, telecommunications and others.

We have taken steps during the last year that, I believe, will significantly improve our ability to address the breadth of these markets during the next year. In many of these non-smartphone applications, our unique multi-core EOS S3 platform provides customers with a very attractive value proposition that not only minimizes power consumption and enables new features, it also enables designers to reduce the number of ICs needed to support the design goals. And more recently, in some cases, helping our customers shorten their new product development process.

Because of this compelling value proposition, our average selling prices, or ASPs, for EOS S3 is trending about 50% higher than we anticipated it would earlier this year. When coupling higher ASPs with the increased forecast for unit volume and broader market adoption of discrete sensor processing ICs, we believe the dollar value of our served available market is significantly larger than what we envisioned it was as recently as 6 months ago.

Now let's move to wearable, hearable and IoT engagements. As I noted last quarter, the Tier 1 customer that is using our EOS S3 in its new wearable design decided to upgrade one of the sensors last quarter following initial field testing. I can't be more specific about the sensor upgrade other than assuring you it did not affect the use of our EOS S3 in any way and that the primary goals of the vigorous testing are to optimize the design for data accuracy and battery life. With the sensor change completed, we are currently working with the customer's manufacturing and engineering group to prepare for the production launch while additional field testing continues. The continued field testing and extreme focus on optimizing this design for data accuracy and battery life could lead to further delays, so we're not including any revenue from this design in our Q3 guidance. We have built up an inventory of EOS S3 to handle any upsides to this current outlook.

In parallel with the ongoing efforts to bring its first EOS S3 design to production, this customer has recently made wearables and hearables a long-term strategy, and with that, a strategic commitment to include always-on voice capability in future designs. We are working closely with this customer on 2 new high-volume consumer wearable and hearable opportunities that are targeted for mass production during mid-2018.

In February 2017, I noted that the unofficial theme for the Consumer Electronics Show in Las Vegas was "voice is the next interface." Since then, virtually every wearable, hearable and IoT opportunity we have encountered incorporates always-on voice capability. We expect this trend to accelerate as use of cloud-based digital assistance, like Ok Google, Siri, and Alexa become more prevalent, particularly in large markets outside the U.S. Our recent design wins with Qiwo and our collaboration with AISpeech to enable its digital assistant on the new Qiwo hearable designs are just 2 examples of how we are leveraging this trend.

Since our EOS S3 SoC is the only multi-core MCU-based device in the market today that includes hardware-integrated low-power sound detection, or LPSD technology, we can enable always-on voice applications at substantially lower-power consumption than any of our competitors. This advantage and the almost universal focus to include always-on voice in new wearable, hearable and IoT designs benefits us in many ways. It has led to a meaningful increase in the quantity and quality of new engagements. It enables customers initially attracted to EOS S3 for its unique ultra-low power voice capabilities to reduce cost and further lower the power consumption for their new products by integrating other system functions into the multi-core architecture and embedded FPGA resources uniquely native to use EOS S3. It has led a number of customers to select EOS S3 for their targeted variable, hearable and voice-enabled IoT applications more quickly than we have seen previously. And with that, shorten the engagement process.

Due to the recent acceleration of the engagement process, the resulting design win activity, we expect several wearable and hearable designs will move into production during Q4. Our recent announcements regarding the Qiwo and Janyun design wins are 2 examples of this improved efficiency. Qiwo and Janyun are Chinese ODMs, which means they develop products that they subsequently manufacture for OEMs under private label agreements. With this business model, ODMs leverage a single-platform design across multiple OEMs, and the more flexible the platform, the greater the volume potential.

Both companies' CEOs have stated an attention to use EOS S3 SoC solution in current and future designs and acknowledge the highly integrated solution provides not only an ultra-low power and cost-effective solution for their designs, but also that the flexibility of the EOS S3 enables them to address multiple OEM use cases from a single platform. This ability to extend the leverage of a platform is a huge value proposition for ODMs.

There are some very noteworthy aspects to both of these design wins. In the Janyun wearable design, virtually all of the resources of our EOS S3 are leveraged. This includes the use of the integrated ARM Cortex M4-F as the host processor; our patent pending Flexible Fusion Engine, or FFE, for sensor processing; our hardware-integrated version of Sensory's Low Power Sound Detection Technology or LPSD and Sensory's TrulyHandsfree technology for voice recognition; and our embedded FPGA technology as a display driver and GPS interface.

The EOS S3 SoC is the only single-chip solution in the market today that is capable of supporting these features and providing the flexibility required in this design.

Qiwo, which was founded as a joint venture with Qihoo 360 Technology, a large Chinese company that is traded on the New York Stock Exchange, selected our EOS S3 for a new hearable design that leverages AISpeech's digital assistant. AISpeech provides cloud-based voice-enabled AI assistance in the Chinese market, similar to what Amazon does with Alexa in the U.S. market.

In the Qiwo design, our EOS S3 is used as the host processor, and our hardware-integrated LPSD enables ultra-low power, always-listening technology to recognize the trigger word for the AISpeech digital assistant. The EOS S3 is the lowest-power single-chip solution in the market today that can support the needs of this application, and we believe future designs will more fully utilize EOS S3 resources.

Based on what we've been told by Janyun and Qiwo, we anticipate production shipments to support initial OEM orders will start in Q4 2017. And they have also stated their intent to use our EOS S3 in future designs.

On our February conference call, I mentioned several major app companies are in the process of expanding their business models through the development of new hardware products that are designed to leverage their already widely deployed software applications. Our early involvement, with major app companies as ecosystem partners, has given us an inside track in this market and, so far, enabled us to win 2 designs that, we believe, will enter production during Q4. One of these is for a wearable and the other is for a hearable device. We are optimistic that there will be a press release for at least one of these design wins later in Q3.

In addition to these design wins, we are engaged with one of China's largest ODMs for a new hearable design. If we're successful, I believe the design will move into mass production during the first half of 2018 and broaden our channel into large OEMs targeting the rapidly emerging hearable market. We are also beginning to see strong interest from voice-enabled IoT OEMs for smart home products. If we're successful, I believe, these designs will also move into mass production during the first half of 2018.

We are still in the early engagement stages with a tier-1 voice-enabled cloud-based IoT provider I mentioned last quarter. We have a very restrictive NDA with this company, so all I can say at this juncture is that we are making material progress in this engagement.

Let's now turn to the smartphone market. We are getting very close to being able to classify the first of our ongoing smartphone engagements as a design win. If we are successful, we anticipate initiating shipments to support production during Q1 2018. This opportunity has shifted roughly 2 quarters from our original production timeline expectations.

While delays are frustrating, the good news is that the design has expanded from a simple sensor hub function to one that now includes a carrier-driven requirement for always-on voice as well. This means the design will leverage several elements of the EOS S3 silicon platform, including ARM Cortex M4-F MCU, LPSD and FFE. This puts us in the lead position from a competitive standpoint since we are the only multi-core MCU-based solution in the market today that can address all of these in a single chip.

There are other smartphone engagements where OEMs are evaluating our recently announced barge-in feature. Barge-in leverages our hardware-integrated LPSD for ultra-low power always-on listening and our recently added acoustic echo cancellation technology. Barge-in enables a variety of always-on, always-listening use cases. For example, with a barge-in feature, a smartphone can hear trigger words, such as, Ok Google, Alexa, and voice commands like, "Answer Call," when a smartphone is playing music or a ring tone. If we win these designs, we anticipate initiating shipments to support production during the first half of 2018.

In addition to the progress we've made in our IP and sensor processing initiatives, we continue to win new Display Bridge in connectivity designs. These include a new educational tablet design win for our Display Bridge and a new IoT module that is using one of our connectivity solutions.

We expect the tablet will move into production in Q4 and the IoT module will move into production during the first half of 2018. While we are forecasting relatively flat revenue from Display Bridge and Connectivity for Q3, we are expecting a sequential increase in Q4. We believe this will be driven primarily by stronger seasonal demand for tablets and the new promotion from Motorola that is bundling its Moto Insta-Share Pico Projector, which uses our Display Bridge solution with purchases of Moto Z2 phones and Moto Mods.

Before I turn the call over to Sue for her financial report, I want to take a couple of minutes to provide a higher-level view of the trends that are driving customer interest today, and I believe will drive our long-term growth and profitability.

Prior to launching our sensor processing strategy, QuickLogic was operating on the wrong side of Moore's Law. By that, I mean, we were integrated out of designs far too quickly. Our sensor processing strategy puts us on the right side of Moore's Law. Today, we are the integrator. Most of the designs we are winning are driven by the unique integration of our EOS S3 multi-core MCU that includes hardware-integrated version as Sensory Low Power Sound Detector, our proprietary FFE technology, embedded FPGA and another -- a number of other peripheral functions that don't make the headlines but are valued by our customers. I can safely say we are the only semiconductor company in the world today that offers these technologies and capabilities in a single chip. The challenge we faced a year ago is while we had the most advanced sensor processing SoC in the market, we didn't have the tools customers needed to efficiently complete designs. With these now in place, we're seeing engagements move forward more efficiently and with less direct support from QuickLogic engineering.

These tools have also helped customers more fully utilize EOS S3 resources, and that adds to the stickiness of our designs.

When coupled with the leverage we get from our expanding base of ecosystem partners, these tools are enabling us to convert engagements to design wins more quickly. We believe this will not only fuel our growth, but will also enable us to address a broader customer base. A subtle but important benefit of these design trends is that they have the potential to create a positive spiral of events for QuickLogic. As customers develop software to take advantage of the unique resources of our EOS S3 platform, the designs become very sticky. Perpetuating this trend is the fact that customers want to leverage their software investments by using EOS S3 in future designs. These trends also enable customers to shorten design cycles, lower the cost of new product development and bring new products to market more quickly.

In parallel with our efforts to build out the software and tool ecosystem for EOS S3, we launched our eFPGA strategy. This business model is also on the right side of Moore's Law. Due to the tiny geometries used by leading fabricators today, our customers can include a small but valuable amount of embedded FPGA in their ICs for only a few pennies of variable silicon cost.

At the bottom line, I share your frustration with the delayed schedules that cause us to reduce our expectation for 2017. However, our growth drivers are solidly intact, and the overarching trends driving our business and our ability to win meaningful designs has improved.

With that, I believe we are very well-positioned to initiate sustainable new product revenue growth beginning in Q4 2017.

When I look beyond 2017, I believe we are uniquely positioned with the right solutions, ideal market trends and customer engagements to deliver our target operating model of greater than 50% revenue growth at a non-GAAP gross margin range of 45% to 50% in 2018, and as our revenue ramps, report a non-GAAP operating profit range of up to 10%.

Now I will turn the call to Sue.

Thank you, Brian. Good afternoon, and thanks to everyone for joining us today. Please note that we are reporting on non-GAAP results here. You may refer to the press release we issued today for a detailed reconciliation of our GAAP to non-GAAP results and other financial statements. We have also posted an updated financial table on our IR web page that provides current and historical non-GAAP data.

For the second quarter of 2017, total revenue was $3 million within our total revenue guidance range. Our new product revenue was $1.5 million, and the mature product revenue was $1.5 million. New product revenue was at the lower end of the guidance range due to lower-than-anticipated shipments of Display Bridge and connectivity solutions. New product revenue contributed 49% of the total revenue compared to 68% in Q1 2017 and 44% in Q2 2016.

Samsung accounted for 21% of the total revenue during the second quarter compared to 22% during the previous quarter. This reflects the seasonality of the consumer tablet market and expanding customer base for our Display Bridge solutions. Our Q2 2017 gross margin was 46% compared to 44% in Q1 2017, a higher gross margin was driven by an increase in the eFPGA license revenue recognized in the quarter, our product mix and a broader customer base for Display Bridge solutions.

Operating expenses for Q2 2017 totaled $4.6 million, which was flat sequentially and a 17% lower year-over-year. This reflects the cost reductions associated with the strategic realignment that we initiated in the second half of last year. Q2 2017 SG&A expenses were $2.4 million due to our increased sensor processing-related sales and marketing efforts. Our Q2 2017 R&D expenses were $2.2 million, reflecting the shift in R&D resources to our Indian location. The total for other income expense and taxes in Q2 2017 was a charge of $54,000. This resulted in a net loss of approximately $3.3 million or $0.04 per share.

Net of fees associated with the equity offering that closed in April, our Q2 2017 cash usage was $3.9 million. This was within our expectations. We ended the quarter with a cash balance of $22.2 million.

Let's now turn to the third quarter 2017 outlook. Our revenue guidance for Q3 is approximately $3 million, plus or minus 10%. The $3 million in total revenue is expected to be comprised of approximately $1.7 million of new product revenue and $1.3 million of mature product revenue.

Our lower expectation from mature products is due to a normal seasonal lull in Q3 for our European customers.

On a non-GAAP basis, we expect the gross margin to be approximately 45%, plus or minus 3%. As was the case in Q2, we expect our gross margin to benefit from recognized IP license revenue and a favorable mix of customers and products. This will be partially offset by lower mature product revenue and a continued unfavorable absorption of manufacturing overhead.

We're currently forecasting non-GAAP operating expenses at approximately $4.6 million, plus or minus $300,000. We expect our non-GAAP R&D expenses to be approximately $2.3 million, and non-GAAP SG&A expenses to be approximately $2.3 million. We expect our other income expense and taxes will be a charge of approximately $60,000.

At the midpoint of our forecast, our non-GAAP loss is expected to be approximately $3.2 million or $0.04 per share. As was case in prior quarters, the main difference between our GAAP to non-GAAP results is our stock-based compensation expense, which we expect to be approximately $450,000 for the third quarter in Q3. We expect to use between $3 million and $3.5 million in cash. The forecast in the cash usage will be primarily driven by working capital need and capital expenditure associated with our eFPGA development effort.

As in prior quarters, our actual results may vary significantly due to things that are beyond our control, such as schedule, variation from a customer, schedule changes and the projected production start dates could push or pull shipments between Q3 and Q4 2017 and impact our actual results significantly.

With that, let me now turn the call back over to Brian for his closing remarks

Thank you, Sue. With our new software evaluation and development solution in place, we have accelerated and lowered the cost of our engagement process. This has also resulted in more efficient conversion from engagement to design win. We believe this will enable us to drive sustainable revenue growth, beginning in Q4 and realize our growth and profitability targets in 2018.

Also driving this momentum is the fact that a voice interface is rapidly becoming a check-box requirement for new designs, particularly in wearable, hearable and IoT applications. As it stands today, our EOS S3 is the only MCU-based solution in the market that includes a hardware-integrated, low-power sound detection and with that, it consumes far less power than other MCU-based devices when used for voice detection and processing.

Some other customers that were originally attracted to EOS S3 often select it as the host processor and are increasingly utilizing other platform resources, too.

In addition to increasing our number of engagements, and in some cases with very large potential customers, these trends have also led to a meaningful increase in average selling prices.

In summary, we are uniquely positioned with the right solutions, ideal market trends and customer engagements. Our growth drivers are solidly intact, and the overarching trends driving our business and our ability to win meaningful designs have improved for the long term. Between that and the higher forecast we're now seeing from research companies, I'm very optimistic we will begin delivering sustainable growth starting in Q4 2017.

Latif, I would now like to open the call for questions.

(Operator Instructions) Our first question comes from the line of Richard Shannon of Craig-Hallum.

Hi Brian, Sue. Brian, I appreciate all the great detail here on both the sensor hub side and the embedded FPGA side, great to hear all that. Maybe delve into a couple of different topics here. I think a lot of people are probably wondering with a lot of the activity that's starting in the fourth or some of the activity to started in the fourth quarter, is there any way people can think about what kind of revenue level we can -- what you might approach -- what might be a minimum level. It's hard to scale some of these opportunities you're referring to. So I wonder if you can give us a little bit help of what you're thinking there.

Yes, I can do that, Richard. So firstly, these first design wins that we did announce beyond the Tier 1 smartphone OEM during the wearable, these are ODMs and so the revenue they're going to generate is largely driven from the end customer that they get signed up. I know that they're planning to show these products around the October time frame at some shows in Asia. And so depending on how successful they are, these could be hundreds of thousands of units. And with the ASP increases that we're seeing, that could be meaningful in terms of revenue. I think the biggest driver here for Q4 for us is going to be this Tier 1 smartphone OEM that launches the wearable. And when they launch, the slope of that curve is going to be the main driver. So we're comfortable saying that Q4 will be the point at which we're starting that revenue growth. I just -- I'm not comfortable with going so far as actually telling a number on this call just because that would be based on some speculation at this point.

I know you started this topic a bit on last quarter's call. I was wondering if you could expand on your -- or go into your latest thoughts here and that's regarding the kind of category of opportunities. Clearly, smartphones are a very high-volume category and then obviously wearables are as well. You're talking more about hearables in the last several months here. It sounds like some of the exciting designs that you have here is as much in hearables as anything. I wonder if you could give us a sense of how you see the opportunity set for hearables, maybe thinking about what the split of business might be well into next year, wearables versus hearables and phones?

Sure. So firstly, I will make some speculation here that at the next CES, we're going to see a big movement of this immersive user experience from our wrist to our head. And I think that's going to be driven from the fact that we're seeing so much of the activity around hearables in this space. I think just within the last quarter, we're over double digit opportunities now just in hearable. So I think everybody is seeing that there is an opportunity here to leverage these digital AI assistance, move it to the ear. And the unique thing about hearables is that the batteries are even smaller than wearables that are on the wrist. And so that really drives home people have to look for low-power solutions. And if they're wanting to have it be a more immersive experience, they're going to be adding sensors like microphones and MEMS sensors for motion and things like that. So they're trying to cram a lot of functionality into these devices for differentiation, for connection to the cloud and so power is of paramount importance. And the traditional approach as to these is just use a Bluetooth chip, but these Bluetooth chips were not really designed for interacting with these always-on use cases. They were really designed around listening to music and speaking to your device, your phone. Now that people are adding step counters or pedometers and turning the microphones on all the time, they are definitely looking for low-power solutions. And so that's what's really caused people to call us and ask about these types of products. Now your other question I think was related to what do the volume or mix feel like next year. I think, right now, again speculation, but CES is probably going to see a lot of these devices on your head. I think we're seeing more design starts now for hearables than for the wrist one wearables. And I think it is because it does change the user experience pretty significantly for the consumer. So I don't know the exact volume, but I can definitely say that from a design start point of view, we see more activity in hearables than wearables right now. And I wouldn't be surprised if people were underestimating the size of the hearable market next year.

Great. I appreciate that feedback. One question from me on the embedded FPGA side, Brian, I think the phrase you used was you fully believe that you're going to have one to several customers, not foundries but semiconductor companies and OEMs, sign up this year. Any way you can help us understand what several means? And can you remind us what kind of opportunity, revenue opportunity exists when signing up each of these? I know they're recognized over time, but what's the cash inflow that you would expect on each one of those?

Sure. So historically, we've used the word several to mean more than -- like 3 or more. So be consistent with that. And we do have the sales engagement funnel to support those numbers. From a dollar point of view, it depends on the type of deal that we actually do sign with one of these folks. We've talked about different models of engagement where people would use embedded FPGA more as an internal tool. That they don't open up to the outside world or they could actually open up the FPGA and market it as a device from an integrated FPGA to target FPGA designers out in the market. So I think the range for pricing on that would be somewhere between under $0.5 million to a 7-digit number depending on how far out they want to take the technology. And that will vary also by process node. A more advanced process node like 22FDX is going to have a bigger price tag than a more mature node like 65 nanometer.

Okay. And Brian, can you remind us, these time frames you talked about, specifically this year and obviously time is an element to that, are these time frames under your control or mostly under your customers'?

Right now, they're going to be under the customers' control. They're evaluating with our software tools that we've released. They're pretty rigorous about evaluating any new kind of IP. They exercise it with multiple designs to see how big, the power the performance, the cost, and then they make decisions. So it's really driven from their IP group getting comfortable with the IP and it's driven from the product side on when they're actually scheduling their own tape-outs and their own development plans.

Okay. That makes sense. Maybe a couple of quick financial-related questions and I'll jump out of the queue. In your guidance about growth for this year, about 50%, and getting to operating profitability that, obviously, clearly implies you think you've got a revenue funnel that gets you to breakeven. I just wanted to verify that, that's accurate. And can you remind us what your revenue breakeven level per quarter is?

Richard, our breakeven revenue level stays, as we set it before, $8 million to $10 million. Was before it's $9 million to $11 million now we are lower bit because of the IP license revenue that carries a very high margin. It's almost 100% margin. So in a about type of revenue coming in the quarter, that, of course, the revenue levels for breakeven will be lower. Of course, based on the 45% to 50% gross margin.

And then your other question was do we have the funnel to support that, and we do have the funnel to support that.

Our next question comes from Gary Mobley of Benchmark.

I wanted to ask about your comment about higher-than-expected ASP for some of these EOS design wins. I'm presuming the die hasn't changed at all. The higher ASP might be a function of what -- more embedded algorithms which is coming from your partners? And does that mean higher split or higher payment going on the royalty front?

Yes. So first of all, yes, it's the same die. We're not changing the die. So any of these variations that we're talking about by addressing different use cases or segments or customers is all software and how we program the embedded FPGA on the die, same die. So the other part about ASPs, we believe very firmly in value-based pricing, not cost-based pricing. So whereas the die is the same, the value will be different depending on the customer's use case. So if power consumption is absolutely critical and they need every little micro-amp that we can save, our value will be higher and therefore our price will be higher. If it's a fight with another MCU guy and they don't really care about power, then the price will generally be lower. So I think what we're seeing is sort of validation of the fact that a lot of these smaller battery products really value battery life and they value the integration that we can offer, and we can get rewarded for that by having higher ASPs. There is the fact that we can, in certain cases, bundle software and get paid for that and also get a higher gross margin for that. But I think it's really the first reason I articulated as the primary driver for the increased selling prices.

Okay. Just to clarify something on the eFPGA front. You mentioned your test chip was taped out, it's SMIC 40-nanometer and just completed second quarter. So I'm assuming that's the first opportunity to go to customers with respect to licensing, right?

No, so we had actually been engaged already with what we've already been running as production devices with our own chips with 65-nanometer and 40-nanometer GLOBALFOUNDRIES. We have 65-nanometer TSMC already in production by virtue of having our own devices there. And then the 22-nanometer is in development, it's going to tape out in Q4 and we'll have verified production or the test chip validation in Q1 of '18. For the 40-nanometer SMIC, it's taped out, but to be clear it's also back and we're going through the verification now. And we expect to complete that this quarter. So it doesn't inhibit us from actually doing the selling process ahead of having the test chip back. But having the test chip back and qualified gives people the comfort and reduces the risk, in their mind, of something going wrong with the IP at some point. So it's not holding back the sales engagement but it would definitely, I think, accelerate actually once that's actually done. Because then I think you'll see SMIC actually promoting this a little bit more aggressively to their customers after they've also seen it working in boards.

Okay. Based on the customer profile for all these different nodes and processes, at GLOBALFOUNDRIES and SMIC, what would you expect in terms of licensee profile or in use case for the licensees? And as a follow up to that, I think you mentioned $0.01 is additional variable cost for a licensee utilizing your embedded FPGA. Is that to assume that royalties on the tail end of these relationships is $0.02 to $0.03 per unit?

Okay. A few different questions there. I'll answer the last one first. For the royalty, when we're talking about a few cents of variable cost, we're talking about the incremental cost at the Silicon level that they're going to be from manufacturing. The royalties are going to be different than that as a percent of the ASP. And I'm hopeful that royalties will actually be higher than the incremental Silicon cost. And I think that'll be the case. Now getting back to your other question, could you repeat it, actually the first part of your question?

No problem. The types of licensees that you would expect just based on the different foundry processes?

Yes. So basically if you look at where we've historically sold our own devices, the programmable logic was optimized for low-power microcontroller-type use cases. So I think that, that will be the one that's mostly aligned out there with the potential licensee, so microcontroller-based devices. Moving forward, we do see some standard product companies that are not microcontrollers. They're more purpose-built for certain applications. There's interest there and that could be a sort of follow-on wave. And then ultimately, I think you'll start to see us getting into more of the compute intensive applications as a road map item. But for the near term, I think it's going to be microcontroller-based devices. I hesitate to articulate any use cases on the call because anything I say here can be learned by our competitors. And I'd like to keep that close to the vest until we can have some public announcements to customers.

Our next question comes from Suji Desilva of Roth Capital.

The gross margin guidance you gave, the range there, can you talk about what drives the high versus the low end of the range in 3Q? It was a pretty wide range 3%, plus or minus.

Suji, yes. So this is actually conservative. We factor in both the product mix in Q3 and also the potential license revenue that we can recognize in the quarter.

Okay. That helps. And then for the embedded FPGA, Brian or Sue, do you expect the revenues to be relatively steady given the ratable nature of the recognition? Or would it be lumpy where some quarters it might zero out and then come back? Just trying to figure out the next few quarters where the embedded FPGA layer in.

So once they're signed, it will be consistent for that opportunity. And I think like any new business that we're starting, the timing of signing these deals will make it somewhat lumpy in the very beginning depending on how many we're signing. But once they are signed, it will be consistent across those quarters until the term of the agreement is over. So like, for example, for Q3 we're not sure yet because of the potential to sign new within the quarter versus not and like Sue said, that's what's driving the variance in the gross margin.

Okay, great. And then -- go ahead, Sue.

Well, Suji, just on the sign now, it's about the terms of the agreement -- license agreement can range from 6 months to 24 months. So it depends on the customer what type of term we negotiate, and that would determine the recognized revenues on P&L.

And that will be the ratable term, 6 months versus 24 months.

Exactly, yes.

Great. And then on the cost side, the test chips, Brian, are these -- or Sue, are these material to you guys? Are you sharing in that cost with the foundries as you bring these markets -- not market but the test chips, rather?

Yes, so I think that we are getting these subsidized by the foundries, which, I think, speaks to the fact that they really value the potential of having embedded FPGA on their IP line card or at least as an option for their customers to use. So at this point, it's not a significant impact for us. There is work to create the test chip, but the actual cost of manufacturing the test chip is largely subsidized though these agreements that we have.

Right. (inaudible). And last question on the China fab, the semi landscape, you guys seem to be doing well there. Can you talk about the ways you're gaining presence there, and if you could kind of handicap whether your (inaudible) relationships or your SMIC relationships, which one do you think is the more potent opportunity relatively?

Okay. This is pretty -- I'm saying wow because (inaudible) is really for the SoC business. And then SMIC is primarily for the eFPGA business. Being to work with SMIC is definitely a driver for that because as you know, China is very aggressive into getting into all things semiconductor, including programmable logic. And so there's a big advantage on that. And the fact that we're the only eFPGA that's currently in SMIC, I think we're going to get a lot of nice attention from them on that. (inaudible), we do, do joint visits and account calls with them. So we have a very productive relationship with them. But I guess from a top line point of view, the biggest driver, if I had to pick between those 2, would be SMIC and the eFPGA. (inaudible) definitely helps, but a lot of the larger customers are using their own algorithms now. And so it's not going to be a primary driver for us on the SoC side to be sort of captive to (inaudible).

Our next question comes from Rick Neaton of Rivershore Investment.

Can you provide some additional color on what is comprised the inventory increase from the end of Q3 last year until the end of Q2 this year? Is it mostly S3, all S3?

Yes, I can address that question, Rick. For the first half of this year, the majority inventory ramps yields as free inventory. If you talk from last year -- the second half last year and even this quarter, we continue to ramp up inventory for Display Bridge solutions, also we get a feel or forecast from customer who would start ramping up the inventory.

You talked, Brian, you talked about these new markets for discrete sensor hubs, beginning to appear and used the forecast, I think, of 680 million units next year. Do you have -- with only 1/3 of them being in consumer devices, are -- do you have other things in mind to help QuickLogic gain meaningful share in those new markets?

As far as use cases go?

Yes.

Yes, so we are seeing voice actually permeating even beyond the consumer space into the industrial area. And we've had people asking us even more about military stuff now where they're arming soldiers or -- not soldiers necessarily, but emergency medical folks with voice-enabled devices that are battery powered and have to be reliable and need to be hands-free. So a lot of it's just leveraging the same use cases we've already developed for. One of the things, I didn't really go into a lot of detail in the call but I guess it's worth mentioning here, is that from a software point of view, we've really been focusing on building out this open framework. And the value of the open framework is that we don't actually care what people run in terms of software on the device. They can do it themselves. And by having this open framework there, it means that we can actually turn that over to the customer and it's going to be very little support. So we don't actually have to build out the entire stack of use cases for them. But if they value the low power, they value the flexibility, then they can go off and do their own thing. And so I think that those are some of the areas of opportunity we see in these other segments, where we don't have to do as much work as we've done vertically integrating solutions on S3 today.

Okay. With regard to the Chinese ODM design wins that you have, what's the typical number of OEM customers that these ODMs get for a platform like the S3?

From them I've heard anywhere from the range of like 5 to 10 per platform. And then they'll go and spin it and do something else.

Okay. And with regard to your engagements at the Tier 1 smartphone OEM that you said had projected production dates for next year, do you have any more confidence that the current projected production dates for these devices, should you win the design slots, would be any more reliable than the wearable that's having the sensor replaced in it?

I think so because the smartphone engagements, they have more of a rigorous release process that they typically stick to because they don't want to miss a window of opportunity. And there's pretty regular releases for phones in particular. Wearables are a different beast, and they don't necessarily follow the same rigorous release process. They're not tied to a carrier for example, for the most part. And so they have a very asynchronous path to that. So I do have more confidence in the dates associated with those smartphone guys, unlike what we've all experienced recently with this Tier 1 smartphone guy doing the wearable. Let me add one more comment to that, too. Sorry, Rick, one more thing. On the -- this Tier 1 smartphone guy doing the wearable, I think that they're also -- they're just pushing the envelope so far with the functionality that they want to deploy with this wearable. And they really want to get it right. And so if you combine that with the fact that it's not tied to a specific release pattern in the market for phones, I think that's why -- another reason why we're seeing the shift here and why that's different from a traditional smartphone launch.

Last quarter, you talked about a possibility of a 55-nanometer process for licensing at a foundry. Is that opportunity still open?

It is. We're seeing opportunity there. We're also seeing opportunity at other workhorse nodes like 28-nanometer and interest in much more aggressive nodes than that on FinFET processes. So where we're at now, I think we've been fairly open about what we're doing from a development and verification point of view with test chips on current nodes that we've talked about like 22FDX, and now the 40-nanometer SMIC. So as we free up from these developments, we are going to go off and do the next node. And what that node is, I'm not going to communicate on this call yet, but we will be able to talk about that in future calls at some point. But yes, the opportunity absolutely still exists.

Right. One last question about 4Q here. You talked about the slope of the ramp and the sustainability from there forward being a lot dependent upon the Tier 1 smartphone customers bringing this wearable into production. Is there any possibility that you could still come very close to the 50% revenue increase target for the year?

In 2017?

In 2017.

I guess, theoretically, that's possible. I don't know, again, the slip of the ramp if they were a steeper slope, and they do ramp in Q4, then, it is possible. And we have a lot of other things going on in the funnel. So I wouldn't take it completely off the table, in that sense. I'm a pretty binary guy there. I'm not going to exaggerate. But we're just trying to give you some sense of what the drivers are for that, and that some of the big drivers that we had in our assumptions have shifted. So I think it was just in the theme of being very open with you as investors, where the risks are.

Your predecessor used to use the term significant revenue increase when he was talking about a percentage that was say between 15% and 30% year-over-year. Is that still doable in 2017?

Like I said on a previous -- I think somebody asked a similar question, I'm not able to really articulate a percentage on the call. I'd like to, but I think just given the uncertainties and not knowing for sure exactly what their forecasts are, I don't feel comfortable giving you that number just because of that uncertainty. That being said, I do feel like the momentum is there. The engagements are there. We started to announce wins. And we see enough to know that Q4 will be the start and it will be up over Q3. I just -- I can't give you a number on the call as much as I want to. I just don't know what that number is at this point.

At this time, I'd like to turn the call over to Brian Faith for any closing remarks.

So one final note, we're participating in the upcoming events. Sue and I will be meeting investors in New York City and Boston on September 7 and 8. We'll be at the SMIC Technology Symposium in Shanghai on September 13; the Design and Resource IP Conference in Shanghai on September 14; GLOBALFOUNDRIES Technology Conference in Santa Clara on September 20; in Munich on October 18; and in Shanghai November 1; We'll be at the Arm TechCon in San Francisco between October 24 and 26. And our CTO and SVP Engineering, Dr. Tim Saxe will be speaking at the 15th International SoC Conference at UC Irvine October 18 and 19.

Thank you, again, for your participation today. And we look forward to updating you on our progress during our next quarterly conference call on November 8. Thank you, and goodbye.

Ladies and gentlemen, this concludes today's conference. Thank you for your participation, and have a wonderful day.