Greetings. Welcome to the QuickLogic Corporation acquires SensiML conference call. At this time, all participants are in a listen-only mode. A brief question-and-answer session will follow the formal presentation. If anyone today should require operator assistance during the conference, please press star zero from your telephone keypad. As a reminder, this conference is being recorded. I would now like to turn the conference over to Moriah Shilton with Investor Relations. Please go ahead, Moriah.
Thank you, Rob. Welcome everyone. Thank you for joining us this morning for a discussion of QuickLogic's acquisition of SensiML Corporation. With us today are Brian Faith, President and Chief Executive Officer, Dr. Sue Cheung, Chief Financial Officer, and Chris Rogers, CEO of SensiML Corporation. Before we begin, I will read a short safe 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 design activity and its ability to convert new design opportunities into production shipments, timing and market acceptance of its customers' products, statements regarding its future stock performance, schedule changes and projected production start dates that could impact the timing of shipments, statements regarding the expected benefits or costs from any acquisition, and 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 and the SEC's website.
Investors are cautioned that all forward-looking statements in this call involve risks and uncertainties. Future events may differ materially from the statements made. For more details of the risks, uncertainties, and assumptions, please refer to those discussed under the heading Risk Factors in the annual report on Form 10-K for the fiscal year ended December 31st, 2017, the company filed with the SEC on March 9th, 2018. These forward-looking statements are made as of today, the day of the conference call. Management undertakes no obligation to revise or publicly release any revision of the forward-looking statements in light of any new information or future events.
Please note, QuickLogic uses its website, the company blog, QuickLogic HotSpot, its corporate Twitter account, Facebook page, and LinkedIn page as channels of distribution of information about its products, its planned financial and other announcements, its attendance at upcoming investor and industry conferences, and other matters. Such information may be deemed material information, and QuickLogic may use these channels to comply with its disclosure obligations under Regulation FD. This conference call is open to all and is being webcast live. At this time, it is my pleasure to turn the call over to Brian Faith, President and CEO of QuickLogic. Please go ahead, Brian.
Thank you, Moriah, thanks everyone for joining our call this morning. We're excited to be talking about this transaction with you, I hope by the end of the call you'll be as excited as we are about what this means for QuickLogic and what this means for you as investors. As you can see from the slide here on the transaction highlights, we're announcing the acquisition of SensiML. They are a software-as-a-service or SaaS AI company. They are U.S.-based. They're based in Portland, they're a provider of end-to-end software that allows OEMs to develop pattern recognition sensor algorithms using machine learning technology. Chris Rogers, who is also on this call, will be going into much more detail about SensiML and their technology, he can share those details with you in a moment.
The consideration for the transaction was all stock, while we were not required to have this call due to below the threshold of materiality from a company point of view, we wanted to have this call with you because we do feel like this is a very strategic event for QuickLogic in total, I think, again, you'll be pretty pleased to see the outcome of this by the end. The benefits are somewhat or pretty obvious. We do hope that this is gonna have a target positive EBITDA of the business unit for fiscal 2019, we do think that this will significantly increase the served available market that we have as a company. If you think about this, the business that they're bringing is a SaaS software business, that we do not have that today as QuickLogic.
Adding that to our revenue streams is going to increase our served available market. I think more importantly, there's gonna be some cross-leverage between their SensiML software suite, our QuickAI platforms, and QuickLogic eFPGA IP or hardware-based IP. I'll get to that more in a second. Let's go for a little context now, though, on AI processing. If you look at data center and cloud, that's really driven a lot of the revenue recently from some of the high-performance, large, expensive FPGAs that are really optimized for speed. People that fall under this category would be the Altera business unit of Intel and Xilinx. This is really all about performance, all about compute.
One of the reasons why the FPGAs are used in the data center and the cloud is because for a lot of the AI or data computing applications, people recognize that you need to change the algorithms or want to change the algorithms in the future. Programmable logic and reprogrammable logic is a great way of doing parallel computing at scale with the ability to reprogram that in the future. Cloud-based AI is gonna continue to drive that growth of those FPGAs.
Now, interestingly enough, if you go to the entire opposite end of the spectrum to the edge or the endpoint, you still have a desire now to have more compute residing as close as possible to the sensor. For this concept of localized AI, FPGAs are still a very good technology to do those types of accelerators that you need to do the localized AI. The big difference is that, while in the cloud you have infinite power and cost budgets, at the edge and the endpoint, you don't have that. You have very restrictive cost and power budgets. If you think about deploying solutions at scale in that market, the edge and endpoint market, you have to have cost-effective low-power silicon. You also have to have software that allows the masses to take their ideas and implement those, and go to production.
That was one of the very interesting things that we learned about as we partnered with SensiML for the better part of a year now, is that their analytics toolkit is really designed very well to help people accomplish that, to size the algorithms, and be mindful of the resource-constrained applications or processors that are used in the embedded world. That was one interesting point. The second is the fact that they can take advantage of hardware accelerators, and Chris is gonna talk more about that in his slides. The fact that they can make use of embedded FPGA in these platforms really makes sense, and another interesting reason why we really wanted to partner with these guys for the long term.
We feel pretty confident that there's a multi-billion dollar opportunity for companies that can actually deliver this so-called practical end-to-end solution for localized AI. That's really what helped drive this transaction. One of the points I mentioned in the early slides was this notion of cross-leverage of the full stack solution. As we've been talking with customers and partners for the better part of a year now about our QuickAI, it became very clear that a lot of the masses of customers do want a full stack solution. They want a processor, they want reference designs, they want software that allows them to take their ideas and realize it in a system and go to production.
There's very few companies that have the financial resources of Apple and Google and Facebook and those big platform companies, there's actually a lot more volume in those other companies to bring to market. We realized that as we were going out and doing side-by-side selling with Chris and his team at SensiML. One of the interesting things that came out of that is this notion of cross-leverage. You can imagine that when we go out and sell to customers directly with SensiML, we're providing a full stack solution, software to hardware. There's also a lot of customers that Chris was already selling at with SensiML directly. As you'll see from his slides, they've actually ported their solution to other microcontrollers, as well as QuickLogic.
The cross-leverage notion is that as they get more users of the SensiML software, more of these other processor companies may realize the value that can be delivered through having on-chip hardware accelerators like embedded FPGA that can then drive more demand for the hardware IP that we provide as QuickLogic. The more platforms that are out there, or silicon devices that have embedded FPGA, the more platforms that are available for them to optimize their software for. The fact that SensiML software is actually designed to be aware of the platform it runs on, that creates this nice cross-leverage and virtuous cycle between the two. The fact that both of these will now be under the QuickLogic domain, I think is a wonderful thing for QuickLogic, it's a wonderful thing for the market, and it's a wonderful thing for investors.
At this point, I'm going to turn it over to Chris so that he can give you a better sense of SensiML, then I'll come back and wrap it up at the end. Pass the ball to Chris.
Great. Thank you very much, Brian. I think to your list, I would add that this is a wonderful thing, also for SensiML. Myself and the rest of the SensiML team is very excited by the opportunities brought by this announcement today and the fact that we have a very shared vision, for where AI and the overall system can go. We complement very well in terms of software and hardware, this is very exciting for us. First, before I get into the details, a little bit of background on who SensiML is. The team, myself and the other core developers that are part of the team, we have a team that's comprised of data scientists, firmware developers, and software developers, all originated as an intact team out of Intel Corporation. The genesis of this was back in 2012.
Intel was making a foray into heterogeneous core microcontrollers for IoT and wearables. They had a group called the New Devices Group that was very much focused on consumer wearable devices and targeting the developers of those devices with an end-to-end solution that's comprised of hardware and software. I led that software team, and our goal with this was to really democratize the process of creating algorithms for endpoint devices in a way that could make those devices truly intelligent. The reality then and still today was that the intelligent endpoint devices were by and large done by highly resourced teams that had lots of expertise in data science, in firmware development, and in coding so that they could translate the given application into practice and something that could fit within the device using the tools that are available today.
We spent a lot of time taking the expertise within Intel and trying to codify the process for creating intelligent algorithms that fit on resource-constrained and power-optimized devices into a software tool that makes that readily accessible and practical to many users. If you look at AI, historically, most of the computation for AI takes place in the cloud, right? Data centers and cloud-centralized approaches to acquiring big data problem sets and then analyzing those are great for traditional AI workloads. When you start applying AI to IoT applications, in many cases, those applications are real-time applications. The latency and performance characteristics of running that in the cloud just aren't practical for what needs to take place. You've seen in recent years, a trend towards shifting centralized cloud processing to the edge.
To date, a lot of the edge analytics that are taking place are the same deep learning types of approaches run on relatively high-end hardware, but they push more towards the edge of the network itself. The missed opportunity to date so far is the underwater portion of the iceberg here, which is the billions of endpoint devices that can't use AI in the same manner that is being applied to high-end, resource-intensive computing devices today but have a lot to contribute in terms of processing locally and making applications much more scalable.
The advantages of enabling these devices are that, in many cases, you can get the kinds of insight that you're looking for directly on the device itself, thereby eliminating a lot of the network latency involved, lowering the power requirements of the device, which is counterintuitive, because if you're doing the processing on the device, you would think that would consume more power. In fact, a large amount of the power budget is spent, in the case of battery-powered wireless devices, just transmitting lots of data. If you can do the processing locally and just transmit the insights, you not only reduce the net power budget for the endpoint device itself and extend battery life and make possible battery-powered sensors, but now you also can look at other network options that weren't really practical.
Like putting rich sensors of video and audio and high-frequency data on networks like cellular IoT networks that have long range, but relatively modest bit rates. The problem with this is that, while these are all great opportunities in the endpoint space, the challenge is that building AI algorithms that run on these devices is no simple task, as witnessed by the lack of software tools that are available today that make it practical for a developer to go take a data set and develop an algorithm that can run in a power-efficient way on these devices. If you look at the market, I think there's a general recognition, even in the business press, there's been articles recently about the big opportunity for AI is not so much in the cloud these days, it's in the edge.
Here are some reference from Forbes, talking about the next goldmine is in the edge. Another reference here talks about the opportunity on embedded IoT devices for AI approaching $26 billion in five years. There's a general acknowledgment within the market that edge and endpoint will be the growth space for AI. The challenge here is, as I said, that not only the need to be able to create highly optimized code that can run on these resource-constrained and power-budget devices, but also the expertise that's required. This data here shows, if you were to compare and contrast the data science skill sets that are available versus those of general software and application developers, this data comes from the U.S. Bureau of Labor Statistics, and it shows 28,000 data scientists available. Most of which are consumed with traditional cloud-based applications.
You've got just under 200,000 users who have data science skills but aren't data scientists per se. Compare and contrast that against the 1.6 million application and software developers out there that, if they had the capability to build intelligent devices, could take advantage of it. There's a real constraint here in terms of the bottleneck being the access to skilled expertise to do these tools with hand-coding. By contrast, what SensiML does is it embodies the process of that expertise into a standardized workflow and a toolkit where a developer of modest understanding of AI or machine learning can take a data set that they create for their own application, collect it, choose their target endpoint device processor, and then submit those data sets and training metadata into the SensiML tool.
The tool will optimize and generate the firmware that will provide an inferencing algorithm for their particular application. It is a process that not only democratizes the access to many more users, but also greatly accelerates the process. When we talk to developers that were doing this the hand-coded way, they would spend six months in effort building this code and validating the code, optimizing it to fit within their appropriate device. Whereas with the SensiML Analytics Toolkit, we know from the outset what the target processor is because the user selects their desired device. Take the case of the QuickAI processor. We know that device has a DSP, we know it has FPGA and a CPU.
The code that gets developed from the SensiML Analytics Toolkit utilizes those capabilities and knows what the compute resources are, and it won't build a model that can't fit on the device. Yeah, that's a huge time savings in terms of the iterative process that's normally required. In the context of the overall market for edge and endpoint AI, our world view of this is that there's four major sectors here. Starting at the very high end, you've got the autonomous driving ADAS and VR/AR applications. These are the things that you're seeing press from things like the Google TensorFlow processor unit, Intel's Movidius, NVIDIA, these types of things. Moving from there down into the smartphone applications, as sensor hubs or AI co-processors within the phone platform.
We're purposely focused on the underserved space within industrial and consumer IoT devices, and bringing this kind of intelligence and capability to a very much underserved space of microcontrollers that are capable, provided they have sufficient tools. If you look at the corresponding market TAM for each of these sectors, you can see that while the ADAS space is quite fascinating in terms of unit volumes, it's fairly nominal, right? Smartphones is a sizable market, but relatively flat. The growth opportunities in industrial and consumer are predicted to be huge, by the next five years, up to 8 billion units. Granted, not all of these units are valid addressable market. The way we think of it is that you've got roughly 2/3 of that market that uses 32-bit microcontrollers.
You've got the things that are using commodity devices below that, like elevator controllers and these types of things, that don't really even necessarily need that kind of complexity. The majority of this market is moving towards more intelligence and can benefit greatly from providing that at the edge as opposed to centralized. Just some example applications here that just show you the breadth of opportunity that can be addressed. This isn't a vertical solution for any one market. It's a common tool flow that can be used across many different verticals. If we look at our focus is predominantly within either consumer devices or industrial IoT, you get a flavor from this here of the variety of opportunities that we've seen and are engaging on active customer projects.
Things like industrial wearables, where we've created motion and gesture sensing devices for first responders, process automation, a project we've done recently in fleet maintenance, where we can do predictive fault detection for vehicles on chassis issues and wheel and tire issues. Sports and fitness, for sure. We had a lot of history there with Intel on creating intelligent sports and prosumer devices. You look at areas and then smart home and then smart city, there's just lots of opportunity for taking sensor data and doing much more with it, by making it practical to do that kind of insight at the edge, as opposed to having to send lots and lots of data to a centralized location, which just doesn't scale when we're talking about billions of devices. For SensiML's business model, how we make money is in three ways.
Brian mentioned that we're a SaaS company. Predominantly, we make money by providing developers with a toolkit on a subscription basis that lets them take their data and very quickly turn that into developed and validated algorithms that they can run on their device. A typical design here would be in the mid to high tens of thousands of dollars per year for access to that service, which is still a very attractive value proposition relative to the amount of effort and labor and time that it takes them to do it through the traditional means. The next layer on how we monetize this is in licensing of the generated code. At the point that you've got an algorithm that you're happy with from a development license standpoint, when you're ready to commercialize, we monetize the resulting code on a per unit basis as a license.
That, again, is on the mid to high tens of thousands of dollars per year for resulting code on an average design win. The other aspect of this is that on an ongoing basis, then you've got continuous learning, so it's not a one-and-done thing. This is what's really compelling is that at the point that you ship the device, over the entire life cycle of that device, you have the opportunity to provide model updates that continue to add value. It's a way that the downstream customers can monetize their products, by providing services on top of the hardware that they ship.
When you combine all three of these things, we see this is very much a design win business, as a software corollary to QuickLogic's hardware business, with low to mid hundreds of thousands dollars per endpoint times thousands of design wins that we can go after. As I said, we see the complementary visions between QuickLogic and with SensiML as a very exciting thing to show this sort of architecturally where we see the complement is that SensiML provides sort of a common layer for rapidly building the firmware that can take full advantage of the hardware's capabilities. In QuickLogic, that's the multiple different cores, the Flexible Fusion Engine, the FPGA, the neuromorphic memory that's in the QuickAI module.
By leveraging all of those things and exposing them to users in a way that's simple and straightforward for them to take advantage of, it's not just an advantage for SensiML to monetize, but it also is an accelerator for QuickLogic's hardware business. As we look at enabling third parties, the embedded FPGA business is an opportunity for us to take the learnings from the FPGA libraries we've built for QuickAI and make those available and expose them for ArcticPro, and provide that as a means for third-party SoCs to take advantage of this as well. The ongoing support for third-party platforms as a whole, provides us with the breadth as well as the credibility as a true software-agnostic company that our customers have grown to trust and will continue to trust as we'll support a variety of different platforms.
I think we see lots of opportunities here to complement, and the fact that we're now integrated fully within QuickLogic provides us the insight and capability to really take advantage of the latest and greatest hardware capability and expose it to the tool. Okay. At this point, I'll turn it back to Brian.
Thanks, Chris. As I go to this closing slide here, I want to reiterate some of the key points for our investors. I think one thing to note is that over the last year and a half or so, we've been talking a lot publicly about making our software platform that runs on the EOS S3 an open framework, so that we can engage with a lot of the application software companies that can deliver the full solution to the market. I think that we've talked a lot about Sensory and DSP Concepts around the voice space, and SensiML is another example of a company that we were able to work with because of that. The openness of the platform allows us to have them run their software intuitively on top of our platform.
I think that it's important to note that, as Chris mentioned, they already have their toolkit running on other microcontrollers like Nordic and STMicro, which I think is wonderful, because if you look at those companies, they already serve a large part of the microcontroller BLE space that's in the IoT area. We're going to continue to expand that as we move forward. I think an interesting distinction there, again, is this notion of cross-leverage. The more platforms that their software is running on, the more customers that are using it, is that virtuous cycle of cross-leverage to drive more demand for the hardware acceleration capabilities that we have with our eFPGA, as well as being able to run their same software on the QuickAI platforms now and in the future.
As I discussed earlier, we're not going to share financials beyond the point of saying that it's target positive EBITDA for the year for their business unit. Any Q&A questions, please hold off financial related questions for the earnings call in February. We'll be able to discuss in more detail on that. I'll just close by saying that, again, as we've been partnering with SensiML over this last year, it's really clear that we are aligned on this strategic vision of democratizing the technology, making it available to the masses, which is going to have that increased served available market. You can do the math on what Chris's business model slide was and see that that's probably adding multiple hundreds of millions of dollars of available market to QuickLogic now, with this SaaS revenue stream.
A significant increase from just being a device and an IP licensing company, and that's that notion of cross-leverage. As we've gone through to meet with customers, it's very clear their technology works, and customers really like the fact that it's easy to use and very quick to get to a workable model that they can test in the market. I think most importantly, there's a strong cultural fit between us and SensiML. When you're selling next to each other on planes and in hotels, you get to know the person, and I can say that there's a strong cultural fit between QuickLogic and SensiML. Really pleased to have them as part of the team, and we're looking forward to doing some great things together.
I'll close by saying that I think hopefully you can appreciate now, after hearing about the context and more from SensiML, that this really is the practical end-to-end solution that the market is looking for this underserved edge and endpoint space, and looking forward to some great results as a result of this. I'll also close by saying that both Chris and I will be at CES, so if any of the investors or analysts on the call are going to be at CES, we'd be happy to show you some demos and products, and look forward to that. After the call, I guess, the next time we'll be touching base with investors will be the February earnings conference call. Thank you for joining, and we'll open the call for questions at this point.
Thank you. We'll now be conducting a question and answer session. If you'd like to ask a question, please press star one on your telephone keypad, and a confirmation tone will indicate your line is in the question queue. You may press star two if you'd like to remove your question from the queue. For participants that are using speaker equipment, it may be necessary to pick up your handset before pressing the star keys. One moment please while we poll for questions. Thank you. The first question is from the line of Suji Desilva with Roth Capital. Please proceed with your questions.
Hi, Brian. Hi, Sue, welcome, Chris. Congratulations to all on the deal.
Thanks, Suji.
Thank you.
I know you don't want to talk about specifics about SensiML's revenue, but can you talk about maybe the end market breakout, Chris, for your business so we can get a sense of which end market segments you've had success in, if that's something you'd look at?
Yeah. I can say that of the markets, we focus on consumer, industrial IoT, and some in automotive. I would say predominantly the first two. We've had some opportunities in automotive. We've purposely not gone after things like medical devices and some of the other smaller verticals.
Any breakout, Chris, between consumer and industrial, just roughly to understand where you've gotten traction initially?
Yeah, it's been about 2/3 in industrial and about 1/3 in consumer.
That's helpful. Thank you. Then, I know you talked about the landscape in broad strokes. Can you talk about who you guys compete with directly? If you don't think about what you do that way.
There are relatively few tools out there that are in this space. There's one emerging company called Reality AI that's been there. Then there's another forming company called Xnor.ai. When you look at things like TensorFlow or Caffe, those are deep learning tools that really don't apply to endpoints. The other two I mentioned, in the case of Reality AI, they provide a portion of the solution, but they don't go to the level of taking it to optimized for FPGA and the DSP functions and bringing it down to a packaged firmware that we provide the assurance to the user that when they generate the model within the tool, they know it's going to fit on the device. That is a big time savings for them in terms of iterative process versus just getting a theoretical model that may or may not fit.
Okay, helpful as well. Then this question perhaps for Sue on the balance sheet, the recent revolver you announced, what's the expected cash flow impact of bringing SensiML into the fold? Talk about your funding position pre and post this, if you're drawing down the revolver or so forth as backstop? I know it's a stock deal, but any color there would be helpful. If you want to wait till the earnings call for that, you can let me know as well.
Yeah. Actually, from a cash point of view, Suji, we're fine with this additional revolving line that will keep us afloat above working capital, at least. Again, the transaction is a pure stock purchase, really it doesn't have much impact on our cash usage other than add a few engineers.
Could you comment whether the deal will be accretive to the cash flow or not, or if it's too early to have that discussion?
As Brian mentioned that we expect the EBITDA positive by end of the year, you can see that by end of the year, it should be neutral.
Okay, great. All right, I'll pass it along. Thank you.
Thanks, Suji.
The next question is from the line of Gary Mobley with The Benchmark Company. Please proceed with your questions.
Hi, Brian, Sue, and Chris. Thanks for taking my question.
Hey, Gary.
Happy New Year.
Happy New Year.
Happy New Year.
I wanted to clarify on this SaaS model. When you basically license the toolkit under a SaaS model, whatever the arrangement, $10,000 a year plus follow-on licensing, who are you licensing to? The system OEMs using the microcontrollers? Are you licensing to the MCUs as a tool that they can utilize to sell to their customers?
Our customers are the OEMs that are building devices using the hardware. It's the developers that are actually creating products using microcontrollers. Just to clarify on your point, the $10,000, it's actually tens of thousands of dollars, because the value there is significant relative to the effort and labor that's required to do hand coding. There's quite a bit more revenue there.
Sure. I guess $90,000 is better than $10,000 a year. Thanks for the clarification. With respect to that point, and as a follow-on, is it the main reason why you think you can continue to maintain the relationship with other silicon providers, is the fact that you're licensing to the end customers versus licensing specifically with STMicro or Nordic, and that's why under QuickLogic's umbrella, you can remain neutral, or the Switzerland as it relates to algorithms?
Yeah, that's right.
Okay. As a follow-on to that, for Brian, you've been partnered with SensiML for a while. Why not just maintain that partnership and forge a deeper partnership? Why did you feel a need to acquire the company?
Because we felt like. For a few reasons, Gary, and I'll enumerate them. Firstly, we know that with the software revenue that we're talking about here with SensiML, that going back to the slide, they're talking about potentially hundreds of millions of dollars in SaaS revenue. I would rather that be a QuickLogic revenue overall, as opposed to just a third-party company with SensiML running independently. I think that's better for our investors and better for QuickLogic, clearly. The secondly point is that, again, going back to the eFPGA, we see the value of embedded FPGA as this hardware acceleration capability to reduce power and free up MIPS on processors. I think that we're starting to see other companies or institutions like ETH gravitate to that notion as well because of the test chip that we're doing with them.
I know deep down that once we have people like SensiML really targeting and optimizing their tool to take advantage of that will create this cross-leverage where that will help drive more business for the embedded FPGA as a hardware accelerator into these other companies that have processors out on the market. I think having both of this under one company with that shared vision, that's going to help realize that much, much more than just running as independent companies. I'll say the last thing, too, and this is also very important and I think maybe counterintuitive for some people, is that we're not doing this acquisition so that we can shut down SensiML's software business on these other processors. We absolutely want them to continue to support these other platforms that they're already on and expand that.
I think that there's always this danger in the market today where if you have a really good software company, that another company decides to acquire them, then they will not have that same view as us. They will say, "No, I want it only running on my chip," and everything else now is not going to get any support. That would be devastating for a company like us, because we have such good traction in that shared vision, and we don't want that to happen. The net of that is that it makes total sense for this to be under the same roof, knowing that we have the shared vision of having them continue that notion of democratizing the technology while being able to optimize it for platforms that we have, or technologies that we can license to other people.
Got you. Okay. How many employees in total?
We have six employees total.
Okay. Can you at least share with us whether or not you've generated revenue at this point?
Yes, we have.
Okay. I think that's going to do it for me, but congrats on the acquisition. It's seemingly a good fit for QuickLogic.
Thanks, Gary.
Thank you.
Thank you.
Our next question is from the line of Richard Shannon with Craig-Hallum. Please proceed with your question.
Hi, Brian, Sue, and Chris, good morning, and congratulations on what looks like a very exciting deal.
Thank you, Richard.
Thanks, Richard.
Most of my questions have been answered, but I want to follow up on one regarding the competitive environment. Chris, your response to that mentioned a couple of companies, one of which I know sort of. I don't know the other one. Tried to look it up real quick, and it didn't seem to me that either of them have an integrated platform that seems to deal appropriately with a wide range of hardware platforms, and obviously, the optimization problem is difficult across a much broader array. Maybe if you could address the extent to which those other competitive platforms can do that to give us a sense of the competitive dynamics, that'd be great.
Yeah, I think that's exactly it. We strove from the outset to make this an end-to-end solution that could integrate with standard evaluation kits, because as you walk through the design win process, most of the OEMs start with a concept. They're going through a proof of concept phase, and they need something to rapidly prototype. The fact that we can take standard evaluation kits, integrate that with our software from the data collection standpoint, generate models that target the SoC that that evaluation kit is built around, and then provide the firmware as well as a test and validation tool that allows them to have some confidence in the resulting code.
I think one of the pushbacks on machine learning and AI as a whole is that the developers tend to fear black box approaches that they don't know whether they can support or not when problems arise. The validation is an equally important part of this. I need to know how this works, not just assume that it's magic and it does, right? We've made a holistic approach to addressing OEM build needs from years of experience of being in the space of creating devices and reference devices for customers from the Intel days. That's, I think, one of the big highlights that we have relative to the other competitors.
Okay. That's helpful, Chris. Thanks for that. My other question, I guess mostly for probably Brian. If I understood your comments earlier in the call regarding the opportunities to weave in your embedded FPGA IP, one of the opportunities here is not just with the platforms that you've announced in your press release, QuickLogic, the Arm-based ones and Intel ones, but possibly internal SoCs that someone might create to be enabled by your embedded FPGA technology. If I could ask you to get your crystal ball out, Brian, and look out two to three to five years out, how much of the hardware do you think is going to be one of these standardized platforms you list in your PR versus ones that are SoCs and potentially enabled by your embedded FPGA?
Wow, that really is a crystal ball question, Richard. From a revenue point of view, in the next couple of years, clearly, it's going to be from platforms that are available, especially if you look at these markets that we're talking about. There's definitely some things on our roadmap that I think would be key platforms for this to be integrated and optimized for that would be out in that time horizon you're talking about. I also feel pretty strongly that one of the reasons why we were working with ETH over this past year is this notion that once they flip that concept public, that test chip public that we've blogged about, that's going to drive more interest in the FPGA or eFPGA being a hardware accelerator that I think SensiML's tool can easily port to that will also drive follow-on.
I guess in terms of number of platforms, three or five years out, probably more platforms targeting all of the eFPGA and the SensiML software than clearly today, so more than 50%, just because of the fan-out effect that we've been cultivating. Nearer than that, from a revenue point of view, clearly it's going to be our own platforms and, I would say, announced devices that we've already got public. Other things that we have cooking that are not necessarily requiring new silicon to bring more value to the market that are not announced yet.
Okay. Fair enough. Very fascinating.
Just one other point on that, Richard. If you look at the PR today, I think one of the platforms it says in there is the to-be-announced. Stay tuned for that.
Yes, I'm sure we'll see that very soon, so I look forward to that as well. I think that's all my questions, so thank you very much, and congratulations on the deal.
Thanks, Richard.
Thank you.
Thank you. At this time, I will turn the floor back to Brian Faith for closing remarks.
Yeah, I'd just like to close by saying thank you for joining today on short notice. I hope you're as excited as we are about what this deal can mean for QuickLogic and our investors. I'll reiterate that we will be at CES next week, Sue, Chris, and I, as well as our CTO, Timothy Saxe, and we'd be happy to have follow-on meetings with anybody on the call in our suite with our demos and customer products. After that, we will talk to you folks in February. Thank you.
This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation.