Hello, and good afternoon, everybody. Thank you for joining us for the final day of Needham's twenty-sixth Annual Growth Conference. My name is Quinn Bolton. I'm the semiconductor analyst for Needham & Company. It's my pleasure to host this presentation from QuickLogic. QuickLogic is a fabless semiconductor company that develops innovative embedded FPGA IP, discrete FPGAs, and FPGA SoCs for a variety of industrial, aerospace and defense, edge and endpoint AI, consumer, and computing applications. The company's wholly owned subsidiary, SensiML, completes the end-to-end solution portfolio with AI and ML software that accelerates AI at the edge, and endpoint. Joining me from the company today are President and CEO, Brian Faith, and CFO and SVP of Finance, Elias Nader.
Before I hand the presentation over to Brian, I'd like to remind folks watching the webcast, if you do have a question for management, please submit a question at the bottom of your screen through the dialogue box. I will come back to moderate Q&A after the presentation. And with that, Brian, over to you.
Thanks, Quinn. Thanks, everybody, for joining us, and thank you, Needham, for the invitation once again to present to everybody. I'm gonna use our corporate investor deck that you can find on our website as a backdrop for this. As Quinn mentioned, please fire away with questions so that we can have this be a fun and exciting presentation to end the last day of Needham conference here. So first slide, QuickLogic, Riding the Renaissance of Programmable Logic. We specifically call it programmable logic, not FPGAs, the reason being that programmable logic is actually fragmenting into different product categories. For the benefit of everybody, we're just gonna define what those are. FPGAs are field programmable gate arrays. Essentially, they are standard product devices that fit into the programmable logic category of the semiconductor industry.
The great thing about programmable logic is that they are programmable by the user, by the end customer. Why would you need that? You could use that to extend different use cases for your designs that you do as a company using programmable logic. You could extend life cycles for production devices so that you can stay in market longer and not have to redesign expensive systems. Increasingly, now what you see in the press is that you can uniquely adapt programmable logic devices to run different algorithms.
With all the talk and buzz about AI, a lot of different people are coming up with innovative ways to implement AI and machine learning, and what better way to do that in a more cost-effective and time-effective manner than to use programmable standard products to go and try your algorithms out as you see fit? So the sort of market for FPGA standard product devices that Quinn was mentioning earlier, there is a new category of programmable logic that's emerging, and that is an IP for FPGA technology, and that's called embedded FPGA IP. The notion there is that what you could previously only buy as a standard product is now an IP that you could integrate into a custom ASIC that you're doing. And so you get all the benefits of FPGA devices, except you don't have to buy a device.
You can actually just integrate that into the ASIC that you may already be doing. So we play in both of these product categories. And recent market research put the FPGA market in total just over $10 billion in 2022, and forecasted to grow at a good clip, a good CAGR over the next several years. And interestingly, one of the highest growth, if not the highest growth, sub-segment of programmable logic in the same forecast period is aerospace and defense. And we actually play into that space, and I'll go more into that in a bit. And there's also talk among the same report about the emergence of this FPGA IP category, embedded FPGA IP, and how that's also being used in the same market segment.
So excited to talk in more detail, but just wanted to sort of set the stage for everybody on what programmable logic is and the two product categories within that, that QuickLogic plays. So now a little bit more about the company. QuickLogic is a programmable logic company. We started life as a discrete FPGA company, standard products only. Over time, we added capability and products to that, so as Quinn mentioned, doing SoCs that have FPGA capability in the same chip. And increasingly now we're seeing a lot of traction around the embedded FPGA IP business that you see there in the middle. So different product categories, different ways of bringing it to market. It's all essentially connected to what we call Australis, which is our embedded FPGA IP generation capability that is unique to QuickLogic.
On the right-hand side, you do see the AI/ML software category. We will spend some time on that in future slides. And lastly, I'll just say that I think with all the renaissance going on in programmable logic, with all that's going on in AI, QuickLogic is absolutely uniquely positioned for growth. We've demonstrated that over the last few years, and we've forecast that for the future as well. So with that, many of you on the call may be familiar with what I'll call the old QuickLogic. Some of the one-on-ones that we had today were investors that had a perception of QuickLogic maybe 10 years ago.
So I love this slide because it really talks about the last few years in particular, and the foundation we've been building that's enabling all of these contracts that we're landing today and what's taken us to profitability in 2023. It's really about retooling how we've built FPGA and embedded FPGA IP to bring a lot more automation. It's about working with really large ecosystem companies like Google, as an example, to help us scale that in a different way than we've done in the past, and also integrating in this AI/ML capability into the product offering. All of these things have led to the strong revenue growth and again, full non-GAAP profitability in fiscal 2023.
If we look at it in a different way, and we think about this vertical technology stack, what's enabled us to do this with FPGA and now embedded FPGA capability? It really comes from the fact that we have been a very well-established company in the programmable logic space for 30-plus years. We've sold tens of millions of devices, thousands of customers, many different process nodes. The bottom line here is we have we know what we're doing when it comes to programmable logic, and there's probably only companies that you can count on one hand that have been shipping programmable logic devices for the last 30 years to the market. We're taking that know-how, and we're applying that now to this IP generation capability, so that we can scale from an IP perspective, how we've done the devices in the past.
And a lot of what we're doing, again, is based on automation. Some of that automation is based on open source capabilities I'll get into in a second, and that's what's making this incredibly exciting for us to take this to market. And then I would say lastly on this slide, there's a lot of press about chiplets and how chiplets can be the next disruptor in the semiconductor space, because it's well documented, Moore's Law is probably ending, or it, if not ending, it's slowing down. And how do you counter this slowdown in Moore's Law? Well, one way to do that is to take more of a Lego approach to how you build chips, and maybe it's not one big device, you're doing a bunch of devices.
This is all about chiplets, and we're excited to see how we can be part of this whole chiplet movement in disrupting the semiconductor space in that sense. So more on that to come, but the fact that we have all this deep domain experience in programmable logic really lends itself to us enabling these new business models. Speaking of business models, this is how we monetize our technology. So starting from the middle, Auroras is an IP generation tool that we use to create and deliver IP to our customers. I'll get more into what that is in a second. But once we have the IP, we can license and collect royalties on that IP to customers that we license it to. So it's very much a semiconductor IP business model.
IP is licensed upfront for revenue, back-end royalty when devices are shipped with that IP. In the case where customers just want to buy a standard product from us, be it an FPGA, be it a SoC with FPGA, or perhaps a custom FPGA for that customer, if we are taking orders from the customer, and we're working with our supply chain to build that, deliver that to the customer, we take revenue at the time of shipment to the customer, and that flows through what's called our device storefront, so that category on the dark blue on the right. And then increasingly, because people are looking at doing their own ASICs these days and customizing the functionality, optimizing it for them or taking more control of their supply chain, they still need design services to actually do those customizations or the device developments.
And that's what enables the far left part of our revenue category, which is design services, and that's revenue that we recognize during the development process when we're actually doing the work. On the right-hand side is the AI/ML software. That is a cloud-based tool for teaching and learning and creating the AI models from sensor data. Once that model is deployed by the customer and devices, there's a royalty model for that. So when you blend all these nice businesses together, you get a nice blended margin, gross margin that we've been reporting. Share more about that in a second, but they've generally been between the 60s into the high 60s, and more recently into the 70% range. And that's because of this blended model of license, royalty, funded development, and gross margin on devices.
So, that leads to what's driving the near-term revenue increases for QuickLogic. What's been, you know, driving that for the last couple of years? What do we foresee into the next few? Aerospace and defense is definitely, it's a hot sector. There's a lot of investment going on, a lot of innovation going on in that area, and, we've landed a contract. We won a contract in 2022, and this is the largest contract in QuickLogic's history. It's also the first time that QuickLogic has been awarded what I'll call prime contractor role, for a DoD-funded program. And this is to build a strategic radiation-hardened FPGA. The scope of the project is outlined here. It's intended to be 4 years, a budget of $72 million. We've finished the first phase, which is this press release snippet here of $6.9 million.
We're executing on this $50 million option that was awarded to us in the summer of last year. As we continue to execute and the government continues to need this technology, we expect this to go the full 4 length, 4-year period. And that's great for, for not just for the defense community to have this technology, but for QuickLogic to have a growing presence in that same area. I'll make note that the $72 million is just for the development of this device. It is not for shipment of the resulting devices at the end of that 4 years.
Our goal, clearly, as a company, is to execute on the program, establish ourselves within this subsegment of the defense community we've been serving for a long time, so that we are awarded the storefront status or responsibility for the device at the end of that four years. What does that do? That opens up us to be able to sell the device to the defense industrial base for programs of record. And in my personal view, this opens up a $multi-hundred million revenue opportunity, for the company for several years into the future as the Department of Defense continues to build systems that need this technology. So excited to be a part of this, this group, excited to be prime, excited at the, the revenue opportunity, both for the funded development of this and then future device shipments, thereafter.
You see the companies on the bottom. Again, we're prime. They are part of our team in executing on this. SkyWater and Everspin being publicly traded companies. Trusted Semiconductor Solutions is a privately held company, but they do a lot of work for the government in this area of radiation-hardened microelectronics. So we, I think, assembled a very good team. We're executing and looking forward to what this comes out of after the performance period of four years. Now, as exciting as that is, we are not just serving the defense community, there are other opportunities to take our technology into the market. Some of our recent announcements are shown here. I'll actually start on the right-hand side.
So, our IP is portable to any process, any foundry, and one of the commonly used foundries and process technologies that we're seeing, both on the defense side and non-defense side, is GlobalFoundries' 12-nanometer technology called 12LP. It's manufactured onshore, it's in upstate New York, and again, they're getting a lot of design activity across segments there. So recognizing the popularity of that particular node, we've ported to that. We do have customer opportunities for that, and as we start to land more of those deals, we'll be announcing them. But that's a big step up for us in terms of process technology and opportunity. Furthermore, we've also announced a GlobalFoundries, excuse me, 22FDX process technology, that's a different one too, GlobalFoundries.
Also a lot of popularity within the automotive, AI, ML, and communication space for that technology. Also, customer-driven design win. In addition to IP, if you look to the left side line of the chart now, we're talking about chiplets. And remember when I was first talking in the earlier slides about this notion of vertical integration and how are people countering the fact that Moore's Law is slowing down, chiplets being one of them, we view there's a big opportunity here for an FPGA-enabled chiplet. Why would somebody want an FPGA-enabled chiplet? Well, today, the choices for programmable logic are to buy a discrete device or do your own ASIC that has embedded FPGA IP. There's actually an option in between those two that a chiplet could enable.
So whereas, you know, if you're a company doing your own ASIC, you're spending a lot of money invested into that ASIC, you want it to last a long time. One option to have, you know, more capability in that, more longer time in market for you, could be to add embedded FPGA IP to that ASIC. Some customers are not prepared to do that fully for integrating in the ASIC. Integrating a chiplet and selecting when you stack it with your ASIC and when you don't, gives you that choice, gives you that ability to optimize for, specific functionality or specific price points. And so we have announced a collaboration with a company called YorChip to do an FPGA-based chiplet.
Keep your eye open for press releases that, you know, either YorChip does or we do in combination with YorChip, that gives you some flavor about when the timeline would be for this and the types of market segments that we would be, enabling with that. But excited to really see that our technology is definitely grown beyond just devices, and spanning different ways that we can get that into the hands of the much marketing customers. Speaking of markets, I mentioned earlier, you know, large served available market. We view that our served market is about $1 billion of that $10 billion that the market is today, growing into the future. The primary markets for us today are aerospace and defense, and industrial.
We're increasingly getting opportunities for people that are doing ASICs, that want some optimizations for AI/ML workloads, and then security as well, being an area where people tend to do a lot of customization, programmable logic being a great way to do that. We do have a consumer IoT market on this slide. I would say that while we do have revenue in the consumer space and mobile space, that is becoming a smaller percentage as we are growing these other, what I'll say, more lucrative and more longer time in revenue segments, to the mix, and that's the view, it is a good thing. I'll also say on this slide that QuickLogic is the first programmable logic company to not just contribute to fully open-source tool chains, but also integrate some of those into our proprietary workflows.
We've worked with a lot of companies in this area now for probably four years, and this has really helped us scale our engineering team in a mindful way to grow the capability to generate more revenue, but generating more revenue at a faster rate than growing headcount. A lot of that has come through the optimizations that we're seeing with pulling in open-source components to that. I'll go more over that in a bit, but we're proud to say that we are indeed the first programmable logic company that has taken that jump. In terms of ecosystem, in terms of customers, you can see, you know, all the logos here, you can look offline at this slide. Big names.
What I will say again, as it relates to the aerospace and defense industry, for the vast majority of our history as a company, we have sold programmable logic to all of the top five and eight of the top ten DoD prime contractors. And so as we're making proposals for, like, the strategic rad-hard FPGA or other proposals that may come up through the government RFP system, this puts us in a good position to demonstrate that we have been a trusted supplier for decades, and that becomes a lower risk perception in the minds of customers that may want to look at embedded FPGA or FPGA chiplets for the future. So where this all comes to, obviously, is generating revenue.
To generate revenue, you need incoming opportunities to the sales funnel, and we've been growing our funnel consistently every quarter since we started disclosing the size of the funnel about a year and a half or two years ago. Currently, it stands at $162 million. This is as of our last earnings call in November. Sales funnel or pipeline for us are qualified opportunities. These are not just, you know, hopes and prayers. These are actually entities that we're talking to about real programs, programs of record, where technical teams and business teams are discussing technical viability, feasibility, business terms, et cetera. And as in the view of this is around two years, so if an opportunity comes in and the revenue for that would be four years out, we don't put that in the sales funnel.
Once it gets within around two-year window, it gets included in that number, and once it becomes booked business, it goes out of that number, it goes into booked business. So the fact that we've been growing the pipeline on a net basis, quarter on quarter for the last several quarters, is a good thing, 'cause we're not only generating revenue, we're also making sure that that funnel is nice and healthy for future opportunities. So I've spent a lot of time talking about Auroras and what that has done to the company. This really has transformed our ability to design products that customers want, that they're willing to pay for, and develop it in a scalable manner that allows us to generate lots of revenue, close out the designs, and not have to grow the headcount at the same rate.
The old way, that we used to do FPGA design, and I think almost every FPGA company does it this way still today, is you assign a whole engineering team, and you do full custom design, and that is very labor-intensive to get to that level of optimization. And for us, that would take our entire engineering team over a year to do that, and that's just to take our technology and port it from one fab process technology to the next. So as an example, if we wanted to port from TSMC 40 nanometer to GlobalFoundries 22 nanometer, just that port alone takes more than a year. We knew that that was not gonna be a way to scale the IP business. There are so many foundries, so many process technologies, it just wouldn't work if we had to do one team for the entire year.
So we were on a voyage of discovery, if you will, to take all of the know-how, and the trade secrets, and the patents, and the technology that we have, and transform that into a way of doing a much faster design. So fortunately, the U.S. government is also looking at how do they do this? How do they start automating this manual process? And the U.S. government, through DARPA, invested in an automation capability with the University of Utah, and it's called OpenFPGA. And it's all about how do you kinda optimize or automate some of this workflow. So we got connected with that group, and initially thinking, you know, this is good for academic use, good for papers, research, not really commercial. But if we can sort of combine the automation from that with our know-how, then we've probably got something.
It was about two years ago that we sort of unveiled this new capability called Auroras, that in fact does that. So what was an entire design team for a year, is now just a couple of our engineers for about three months for the first time, and then subsequent IPs off that same port become weeks, and hopefully, and as soon it's gonna be days. So think about the level of scale we're getting with that now. We went from one for the whole year for one team. In Q4 of 2023, we've already said publicly we were gonna be doing five process ports in that one quarter. So that's a factor of 20 difference, if you think on an annualized basis.
So this has really sort of unlocked, I think, the capability of our company to serve the demand that's there for this and turn the funnel into revenue. So that's on the programmable logic side, and we'll get into the financial impacts of all those in a few slides, but really, I think off to the races on the FPGA side. On the AI/ML software tools, so it's a wholly-owned subsidiary company called SensiML, S-E-N-S-I, big M, big L for machine learning. They were a team within Intel about 8-10 years ago. They spun out from Intel and they were a private company for a couple of years, and we were a partner company with them, looking at how we could take one of our SoC devices that has an Arm core into the IoT space.
Software is great, workflow is great, customers it was resonating with, and we ended up joining forces with them in 2019. So we announced that acquisition. We're running them as a wholly-owned subsidiary with the intent that we want their software running on everybody's microcontroller, not, not just QuickLogic's devices. And I think that strategy has, has been well-received by the marketplace, certainly by the list of logos you see at the bottom. SensiML is collaborating with each one of those, in some cases, doing reference designs, co-marketing of the capability. In one case, the last bullet there, one of the microcontroller companies has done a private branded version of SensiML within that microcontroller company's toolkit that they're selling to their customers. So you think about the scale now, we're a big microcontroller company, large install base, big sales force.
They are now bundling and taking the SensiML toolkit to their customers. And so that was launched, I think, in Q4 of 2023, and now we're working with them to really help them achieve their goals from the opportunities that they're getting in the funnel, converting to revenue, that would then flow through to SensiML. So that's not an exclusive relationship. There definitely the desire, I think, on our part, for that to be a rinse and repeat with other microcontroller companies, and we're doing our best to support SensiML in that endeavor so that they can truly realize this value of enabling machine learning and edge AI models at where the sensor resides, not forcing everybody to go, sending data back to the data center. Okay, so we talked about programmable logic, we've talked about SensiML AI.
How does this roll through to the numbers? As I've said, we've had really strong growth these last few years. You can see in 2020, when we really started to sort of unveil the new way of doing our IP design with Auroras, and you see the growth coming from the green line there, the eFPGA-related products. You see the mature kind of kicking along there at the bottom. The gray area there is EOS sensor processing. As Quinn mentioned early on, we have devices that have their SoCs, so small microcontrollers with FPGA. Last year, 2023, we're forecasting that to be down from the previous year, excuse me, largely because of the smartphone inventory digestion that's going on in the worldwide market for Android phones.
So we've already said publicly, we think that Q3 of 2023 was sort of the trough for the EOS S3 sensor processing revenue, and 2024 will be up over 2023 based on the bookings that we were seeing. And also projecting for 2024, we said publicly on our November call, we saw probably another 30% growth here in 2024 as well.
So the big message within the company and, and externally is just closing these opportunities in the funnel that we have, generating more revenue on that green line, and then, getting to a point where in a couple of years, when some of these devices are done, both from our customers and ourselves from the defense side, then we can start to see that step function increase, in revenue, not just the 30% growth. But step function increases because of devices going to production and us generating gross margin on shipments or, or royalties on, customer shipments. Elias and I love this slide because I think it really speaks to the efficiency and the automation, with Auroras. So normally, you know, you'd see 30% revenue growth year-over-year, you'd expect operating expenses to be doing something similar to that.
We have in fact been keeping operating expenses relatively flat, and the reason being is because we're automating a lot of the process. And so, it takes a little while to get to that point, but once you're there, like we achieved, and you see the 2023 forecast here, and in fact, what we did in 2023, like Q3 actuals, achieving profitability. And that's really speaking to A, the value of the products, B, the revenue growth, and C, being able to maintain reasonable operating expenses to get to that profitability point. So that really is at least the summary slide. This says 2023 breakout year because we haven't fully reported 2023 yet. We'll do that on our February call. But really, that was a big expansion year for us.
All that bricklaying and foundation laying with Auroras and retooling how we take to the products to market being done in prior years, 2023 really being about grow the funnel, close us to revenue, and execute. We've been doing that on the Rad-Hard contract. We've been doing it with other contracts that we've been announcing in press releases, SensiML, doing the private label launch with the microcontroller company, and then getting to that point in Q3 of last year, where we hit target gross margins and profitability. So again, the fiscal 2023 call will be in February. That's where we finish off the year with everything else, but the guidance that we gave in November was profitability in Q4, the second half, and for all of 2023. So that's the plan, and we don't plan to stop growing in 2024.
It's continuing up and to the right. So with that, I will pause and turn it back to you, Quinn.
Perfect. Well, thank you for that, Brian. I really appreciate it. I've got several questions, and I'll moderate questions coming in from the webcast as well. The first question, Brian, maybe it's more of a clarification, but you shared your revenue mix from 2020 to 2023. Most of the revenue now coming from the eFPGA segment. Is that all licensed revenue? Is there product in there? Maybe just to clarify, what's in that eFPGA revenue bucket?
So in that bucket are three areas. There are IP license, royalty, and services associated with IP. There is no product revenue in the sense of devices.
Got it. Okay.
We haven't necessarily broken out how those three play into that, but they're all related to either the creation of or the license or the royalty of the IP part of the business.
Got it. So it is a sort of true IP, you know, traditional IP kind of bucket.
Correct.
Got it. The second question is, you talked about a $72 million contract to develop the radiation hardened FPGA technology. How much of that $72 million is left to be awarded, and when does the four-year period sort of wrap up? You know, when, what, when does this development period, you know, when is that expected to conclude?
Yeah. So what's been awarded to QuickLogic so far is the $6.9 million plus the $15 million, so $22 million. We can't go into contract specifics, so we haven't said when exactly the $15 million ends in time, other than early 2024, which we expect would result in the next part of the contract tranche. So essentially, 72 minus 22, so about $50 million left in the total contract, and that's over four years from the start. The start was August of 2022. The press release says the eighth of September, but there's a gap between contract signing and when press releases get approved. So think about, like, end of 2026, roughly, is that end of the four-year performance period. And again, that's just for the contract of $72 million.
It's not for device shipments that we could do at the end of that, if we're awarded the storefront.
Got it. Some of your partners, you know, that you're working with, SkyWater, Everspin, I think, have talked about, you know, perhaps devices, you know, kind of by the end of this year, early 2025. Is that just devices that then, you know, kind of think of them as sort of almost prototype level devices, not necessarily the high-volume production?
Correct. Yeah, what I've been given permission to say on that topic is that this four-year performance period is comprised of a test chip and a final chip. So yeah, we haven't said exactly when, but you can imagine that the test chip is probably in the middle of the time period, and the final chip's at the end. And so you could imagine sometime in 2024, early 2025, there should be something to, to test.
Got it. Okay. Great, and then the $162 million pipeline for IP, can you give us any sense how does that split by end market application? Is it largely AMD? Is it largely AI related? You know, what can you tell us about that opportunity set?
Yeah, so the funnel today, more than half of that is aerospace and defense. There is so much going on in that sector right now, and the fact that we have not just the contract that we have, we have other opportunities in that same segment, almost by virtue of us being such a visible performer now in that space, that's growing that sales funnel part of it for us. So there's no stepping on the brake on that. We're all in to grow that as much as we can. In addition to that, I would say the next biggest segment is in the industrial, not necessarily industrial IoT, but just industrial as a macro category. We do have AI, ML opportunities. They're interesting in how people are looking at applying FPGA technology to optimize certain workloads.
It's probably less than 25% of our funnel, though, and frankly, I'm okay with that because I think, you know, everybody and their mom is doing something in AI and ML, and we need to be very cautious about sort of oversizing or getting overhyped on how much that's in our funnel, compared to industries that we know are, you know, have very defined use cases and don't have so many players trying to do the same thing, frankly.
Yeah, yeah, I was gonna say, it sounds like the aerospace and defense opportunities you're going after. You know, nothing's a sure thing, but that certainly seems like those are some pretty attractive opportunities. And you're right, it does seem like everybody in the semi industry seems to be trying to position themselves for AI, so, you know, could be, could be pretty crowded going forward.
Yeah.
Another question just on the Auroras, the software tool. Is that really just... I mean, I shouldn't say really, 'cause it sounds like it's bringing you significant advantages, but, is that really sort of - should we think about it as, as a porting tool? Does it-
Mm-hmm.
... like, you have to develop the IP using sort of the traditional, you know, hardware engineers, you do the layout, but once you have an IP block, then with Auroras, you can port it, say, from UMC to TSMC or UMC to GlobalFoundries, you can go, different nodes. Is that really kind of the benefits of that tool?
Yeah, the benefit is that you can do the actual port in a more automated way, almost like an ASIC design flow. And so it's a combination of a lot of things: know-how, workflow automation software, and also just industry typical EDA tools from Cadence, Synopsys, and Siemens or Mentor Graphics. It still requires the same type of engineer to understand how to build FPGA fabrics, how to stitch them together, and how to guide actually the tools to do that actual creation. But the real benefit is the fact that we're compressing that, to me, more like an ASIC design flow. And even further, once we've done a process port, the first time Auroras absolutely creates the derivative cores that we would license to customers for different sizes and densities and whatnot, much, much faster.
Like, push button, optimize, do the verification for a couple of weeks, and then get it out to the customer.
Got it. Yeah, this will be a technology question. You know, in this radiation-hardened FPGA-
Mm-hmm
... you've partnered with Everspin, which is an MRAM-based technology. What's unique or special about MRAM as you think about, you know, applying it in a, you know, radiation-hardened application? Are there advantages to MRAM? You know, I imagine there are versus SRAM. But, you know, kind of maybe spend a minute of the choice of the MRAM technology for that rad-hard solution.
I'm gathering my thoughts on how I'm gonna answer this without going beyond what I'm allowed to go to.
Okay.
So within FPGA technology, FPGAs are programmable, and so how you store your design on that chip or configure the chip is stored in memory elements. And you're right, SRAM is one technology that's actually the most mainstream technology that every FPGA vendor uses. But SRAM is not immune to radiation, and it's also, it's also not the best for certain applications that the Department of Defense needs. It's certainly not non-volatile, so when the power is taken away from the device, you lose the configuration, you have to reboot it. So for a lot of reasons, including what drove some of our mature products into the defense space, are that it's non-volatile, so as soon as the part, it's power, it's working. And then it's well known that MRAM technology is actually very resistant to radiation effects.
From a Department of Defense perspective, just like the press release reads, there are lots of applications that encounter radiation. They could be things like satellite communication. It could be things that are more strategic defense in nature. As the Department of Defense and these prime contractors building all these systems are looking at how can they bring and continue to use FPGA technology, they can only really use FPGA technology that, frankly, is very different from what people do at the commercial level. Which means looking at different configuration memories that meet all those specific operating environment needs. MRAM has been standing out as a technology that I think the DOD trusts in a lot of those applications, but there's never been an FPGA that integrates that.
And so this whole project, as you can see at the high level, starts to bring some of those capabilities together for the first time, in an FPGA technology. And I'll kinda have to stop there, 'cause I think that's probably all I should really say about it.
Okay. I was gonna ask about, you know, perhaps why not NOR? But NOR, NOR's got some scaling challenges, right? It just can't get much below 40 nanometer. Is that perhaps why NOR might not have been, you know, at least on the non-volatile side, maybe why NOR was not looked at?
Flash has a couple challenges with radiation, but there's a scaling problem with flash, just universally. I think there's also challenges with respect to radiation hardness with flash technology as well.
As well, okay.
Yeah.
Okay, perfect. I appreciate you bearing with my questions. I know there's only so much you can say.
Yeah.
You know, you guys, you know, obviously have, have a lot of irons in the fire. You, you've got, you know, this contract, you've got that $162 million opportunity to, to, to go out and execute on. You know, you've kinda said you, you wanna grow 30% this year. I'm sure those are all, you know, priorities for, for calendar 2024, but, but anything else you would, you would highlight as kind of your biggest areas of focus or, you know, biggest priorities for the company in 2024?
You know, for 2024, I'll go to this slide to answer that question. So if you think about 2021, 2023, that was all about foundation laying, changing how we design products, changing how we bring them to market, and making sure that the dog is eating the dog food with revenue. And getting into 2023, reporting profitability for the first time in Q3, and then guiding, obviously, the same for Q4, I think that really sets the stage for... The priorities in Q4, obviously, continue the revenue growth, but continue it in a way that we can remain profitable. And that means, yes, we're gonna have more contracts coming in to get to that 30% growth for 2024, that are gonna cause us to have to hire a few more people in the company to achieve that.
It's not gonna be going crazy on OpEx, it's gonna be very mindful investments for its very specific hires, so that we can grow the revenue again, at 30%, but then doing it while we remain profitable. And so now the whole notion is really around scalability and mindful scalability for 2024, and that's, you know, been communicated throughout the company, in every department, really, so that we can achieve that. There's lots of companies that can grow by hiring a ton of people, we're not of that mind. We think we can do the 30% by being very prudent with our hiring and execution.
Got it. So, scalability, but also profitable growth?
Exactly.
Perfect. All right, well, looks like we're at the end of our time, for this session. So Brian, Elias, thank you very much for joining us at the Needham Growth Conference. We really appreciate your participation.
Thank you, Quinn. Appreciate it, invited interaction.
Thank you, Quinn. Thank you, everyone. Thank you, everybody.