We'll move right through to the first slide and get you going.
Okay. There we go. Peraso is a fabless semiconductor company. I just read today that the top seven of the top 25 companies in the world are semiconductor companies, so we're in good company. Basically we focus on wireless technology, and our specific focus is on 60 gigahertz millimeter wave, which is very high frequency. We are a global leader, really the number one player in the market. We've sold about 2.5 million chips worldwide. We're up to about 72 patents. This is a bit out of date, but we put a lot of effort into our intellectual property efforts. We have secured an expanding design win across a diverse set of markets that I'll really talk about today.
The kind of the bread and butter market that we're focused on is what we call Fixed Wireless Access, which is kinda high-speed internet, and that's really become a mature market for us, and frankly, we really now are the dominant player. I would say we have over 90% market share, and that's the idea of just providing you know companies with high-speed internet, consumers with high-speed internet access. You know, kinda looking to the future, we're seeing some nice traction in what we call Edge AI. This is the idea that you know companies need have things like either robotics or robo-taxis or even you know delivery systems that require you know real-time processing of information, thus they need a high-speed interface. Our technology is good for that.
I would call that more of a future opportunity. What I'll spend a lot of time talking about today is what we call tactical communications. That's really what we see the strong growth opportunity over the next couple years here, and this is the idea of secure communications in the military, and particularly for drones. You may have seen we did a press release about on March 6th with our efforts in the drone space. That really is a subset of our overall efforts in the military space, but I'll kind of drill down into the details. That is a quick summary of the markets, the primary markets that we focus on. Quick engineering lesson.
You know, basically, we really exist in very high frequency space. The key takeaway from this slide is that we have a key technology called beamforming, and you can see in these diagrams, in these photographs, either in the fixed wireless space or in the you know, in the drone space. As opposed to standard wireless, which you know, kind of permeates in all directions, we really focus that energy into a very specific direction, and it can be steered, so it can be dynamically steered in real time. The idea there is that, so, you know, really, you know, a key benefit in the fixed wireless market is the idea that we, you know, that we can operate in very dense environments.
You can have many systems in a very tight space. That concept is gonna permeate throughout the presentation. As it turns out, in the military, a key aspect is the security and the stealth. We call it stealthy capability on the battlefield, and that is, it's basically like finding a needle in a haystack. What we've found from Ukraine and from the war in Gaza now in the Middle East is that standard communications are compromised because the enemy can track them down. With our technology, we can have secure communications, and it's very difficult to jam as well. This has become a very important aspect of modern military communications, and a real benefit from our company.
If you look at, so looking at, you know, the growth, I mean, broadly, our millimeter wave chipset market is predicted to be at about a 25% growth rate. That's really just kinda overall, whether it's Fixed Wireless, again, which is becoming a bit mature, or things like robo-taxis or military. But then if you look at the very specific part of the military aspect, it's growing at really a rate of about 42%. Of course, that's because of this ability to provide stealth capability in a military environment. We see even a better growth rate in the military side of things. If you look at how, you know, really our, the first pillar of our business, we call it Fixed Wireless Access.
Our lead customer there is a company called Ubiquiti, and Ubiquiti now has become about a $50 billion company, so a very, very large company. They brought our silicon to market. Again, just to clarify the value chain here, we are a semiconductor supplier, so we sell to OEMs, and then the OEMs sell to service providers. That's kind of the sequence of events. Basically, Ubiquiti is a very prominent player in this space, and you can see from this diagram, last year at a prominent trade show, the entire booth was really products based on Peraso silicon. Real strong validation of our technology, as they called it bulletproof. Now, their basic market was more, I would say, rural and particularly in America. For us, it was a good pipe cleaning exercise.
I mean, it really got our technology mature, got our production efforts and test efforts mature as well. That was a very strong first customer for our company. What we found over the last few years is that we really, I'd say, evolved into a broader marketplace, which is what we call dense urban. The idea here is, again, because of that beamforming capability, we can really address broader markets than just what we call rural markets. We've got deployments in Los Angeles with a partner called WeLink, in South Africa with a partner called Occasion, even in places like Kenya with partners like POA. Really, you know, this technology is what we call DUNE. It's referred to as dense urban networking environment. It's a fully integrated hardware and software solution optimized for very, very dense environments.
Again, that, you know, going back to my original technology slide where we talked about our beamforming capability, that's really what allows us to go into these very, very dense environments. You know, as it turns out, there's about 2.5 billion people worldwide that have actually poor or no internet access. Our technology is really an ideal, you know, vehicle for getting really the world population significantly online using for internet technology. We feel that's a very strong growth market for us. Again, it's quite a mature technology. Basically, we're just kind of riding the wave now in terms of our customer engagement and really the natural growth rate of the market.
As I mentioned earlier, we over the last, I would say, three years, we've really kind of not shifted our focus, but really expanded our focus, I would say, to what we call tactical communications. Again, traditional wireless is susceptible to enemy detection. We are an inherently stealthy protocol. Basically, because we have a narrow beam and we have beam steering and also oxygen attenuation, it's very, very difficult for the enemy to track us down. The other benefit we have is that we work at the 60 gigahertz frequency. It is unlicensed on a global basis. That means, you know, that's actually quite important because if the U.S. or IDF or whatever military goes into foreign countries, they don't interfere with their traditional wireless networks that are licensed, actually.
Basically they can go to these places and really not be that disruptive. We're really seeing a variety of applications for operating base command posts, surveillance vehicle to vehicle communication, vehicle to personnel communication. An announcement we made on March 6 was really another application which we call friendly fire avoidance or IFF. Identification of friend or foe, as it turns out, and certainly this was started in the Middle East with the IDF and the kind of the problem that they've been trying to solve is 30% of their casualties are from friendly fire. Really an amazing kind of number. Believe it or not, I think like yesterday in UAE that the UAE shot down three friendly planes, basically. Right. Friendly fire is a real issue in the world.
Certainly, you know, you can imagine. The concept here broadly when we started out was for infantry. The idea was if you have a soldier before he takes a shot, you know, basically he sends a signal out and then, you know, a receiver is on his colleague's helmet. Basically, if he sends a signal back, you know, the soldier knows not to take that shot. That's the fundamental concept is, you know, am I a friend or is he a foe? That's the fundamental technology that we put in place. Now we've kind of taken that concept and applied that to drones. I think there might be even more of an application for drone communications.
If we kind of look at specifically what we're doing on the drone front. Basically, you know, again, you can see that it's kind of a two way street on drones, like on one way, you know, basically what we're seeing and where the you know, the value proposition is if you have a swarm of drones in the sky, some of them are friendly, some of them are not friendly. We just got a stat. And again, I'll, you know, have to verify this. Basically, we've been told by IDF that 60% of the drones they shoot down are actually friendly drones. This is even a more substantial problem now on the drone side of things. It's almost impossible to identify whether that drone is friend or foe.
Don't forget, like basically, yes, you can solve from a technical perspective, you can solve this problem with traditional wireless. But the problem is, of course, it can be detected and jammed. That's really a problem. With our technology, because it's stealthy and very difficult to detect in the battlefield and to jam the battlefield, it's an ideal way to address this problem. Of course, the flip side is that the drone also has. There's been documented cases where it takes out targets it's not supposed to take out, i.e., friendly targets. Basically for drones, it's really a two-way street to see whether the drone itself is friend or foe and whether the drone should be taken out, a friend or foe on the ground.
We really feel drones, you know, certainly, you know, in the Middle East, you know, now we have discussions in Ukraine with the U.S. military, the Canadian military and even some places in Asia, you know, are kind of embracing this technology. We really see this over the next couple of years. You know, broadly, IFF technology is an important technology. Then very specifically, we feel the drone aspect of it is very important market for us over the next 12 months to 18 months. Then just, you know, I think I spoke at the beginning about, you know, future markets. Certainly a market that we are engaged in is what we call Edge AI.
Again, the concept here, and I'll be cognizant of time, the concept here is that basically, you know, we have applications we're seeing at the edge of the network. Basically let's take robotaxis for example, and we did a press release on this a few months ago. The idea is you have, let's say, a self-driving car driving around San Francisco, collecting video data, collecting you know, radar data, collecting Lidar data. There's a lot of information that it collects. Now, what do they do with this information? Well, this information has to be processed at the data center. The problem is there's so much information that it's very difficult to get off the car quickly.
When they come back to the charging terminals, they have about one hour to download a terabyte of data, and that's a lot of data, even wireless, with a wire. Our technology can handle that, but the idea is that that data, as the car goes back to get recharged, we download the data, but not only do you download the data, once it's downloaded, of course, it's processed at the data center, then the models are re-uploaded back into the car. All of that information can be actually used to help navigate those cars, you know, back into their network. Basically,
There's actually other examples in factory automation, for example, where you know robots are the same thing, like processing massive amounts of data, but as opposed to doing that on the robots, getting downloaded to the data centers and re-uploaded. All of that needs very high bandwidth capability. Those are new opportunities that we're just starting to see now over this last quarter or so. I you know broadly would frame it as you know we have Fixed Wireless Access, which is you know 5-6 years old now and becoming a mature business, and we dominate that market. Over the last 3 years, we've really started making inroads on the military with using our stealth capabilities.
Now over the last quarter or two, we're seeing these applications on the Edge AI, where these customers need very high data rates to offload the information, process it, and re-upload the models to optimize performance. I would say those are really the three, you know, primary pillars of our business. I'm going to skip to this slide. Basically, you know, I think one thing to keep in mind about our company is that, you know, all these products that you see on the screen, and frankly, you know, in the order of, say, you know, at least 60, are actually shipping. That's really meaningful because basically, you know, as I mentioned earlier, we're a semiconductor company. Our customers are our manufacturers.
When something goes into production, it means a lot. It means that they've had to, you know, engineer the product, they've had to test it, they've had to prototype it, they have to put a production line, they have to put test capability, they have to get FCC approval. There's a whole process. It takes, you know, a good 12 months-18 months for these customers to get into production. We've got , 57-60, I would say now, customers in production with real products and, you know, really are shipping. You can see some of the logos there of some of our key customers. Basically, you know, Peraso is a company that's been shipping for, you know, several years now.
We are in production, and almost more importantly, our customers are in production as well. Going to my concluding slide, you know, basically we're a company that actually we have a very sophisticated, advanced technology enabling really, you know, existing and future applications. I mean, whoever heard of stealth technology in the battlefield? You know, spectrum congestion is a real problem, and bandwidth limitations of traditional wireless are just not working, and so we're an ideal technology to kind of come into play. Our stealth capabilities in the military space are vital for some of these applications like drone friendly fire avoidance. That's certainly an important market for us.
You know, we've got a robust pipeline of customers in a diverse set of application space. That's really kind of a quick snapshot of who we are. From that, I'll let Jim take it from there. Jim? Jim, I don't think we can hear you. I can't hear you. Okay. Well, now I hear. Jim?
Am I there?
Now I can hear you.
Now you can hear me?
Yes.
Not sure what happened there. Apologies.
No problem.
Okay. I can see myself, so obviously you can hear me. All right. Well, back to the slides here. Well, it shows our revenue went down from $14.5 million roughly in 2024 to $12.2 million in 2025. You know, the story behind that is that in 2024, approximately 90% of our revenue was from our legacy memory business, and 10%, or roughly $1.4 milliom-$1.5 million was from our millimeter wave business, which is the future of the company. As you heard from Ron's presentation, our focus. In 2025, we flipped it to about an 80% split. We did about $9.4 million of millimeter wave and the rest memory.
Looking ahead to 2026, current analyst estimates to us have us doing anywhere from $12.7 million to over $15 million. You know, obviously we're targeting 50% growth. You know, right now with a lot of uncertainty in the markets, we're seeing companies, you know, definitely don't have the visibility we'd like in our backlog, but we expect that to kind of firm up in the second half of the year. You know, from a margin story, the memory products that dominated revenue in 2024 have very high gross margins, high 60%. The millimeter wave, our target is around 50% gross margin, 50% or higher. When you look in 2025, 58%, fourth quarter 52%, you know, we're pleased to be achieving our targets there.
You know, from an operating loss perspective, over the last couple of years, we've squeezed out all the cost out of the business. I'll just flip to the next kind of income statement view. You know, at a high level, our non-GAAP OpEx, all these numbers are non-GAAP, was running, you know, just under $3 million a quarter. When you look at us to get to break even, we need about $6 million of revenue on a quarterly basis at a 50% gross margin. You know, we are averaging kind of about $3 million here for Q3 and Q4. You know, we have announced and pre-guided, we do see a dip in Q1. We had a shipment pushout that impacted those numbers.
Instead of shipping by March 31, we're gonna ship in the first half of April. You know, but as I said, we're holding on to our forecasts and expect to see meaningful growth in 2026 over 2025. Next slide's just our non-GAAP to GAAP reconciliation. Basically, primarily stock-based comp in 2025 and some severance charges. The amortization of intangibles was all done at the end of 2024, so going forward, that will not be a reconciling item. Looking just the gross profit only impacted 2024. Twenty twenty four gross margins, w e had some amortization of intangibles, not a factor in 2025 and going forward.
Just quickly on our balance sheet, as of December 31st, we had about $2.9 million in cash, $6 million in total assets, $1.5 million in liabilities, and stockholders' equity just over $4.5 million, and no debt. TSO as of December 31st was, you know, about 10.7 million shares. And then we obviously have, you know, done financings over the last couple of years. We do have some warrant coverage and then shares from our equity plan, so fully diluted about 19 million shares as of December 31st. As we're tight on time, I'm going to cut it there, and we'll open it up for Q&A. Thank you all for listening today.
Yes. Thank you very much.
Hey, operator, are there any questions?
Hello. Okay. Seems like there is a question from the investors. Okay. From Sardar, there is a question from...
Oh, so this is the you recently secured strategic engagements that question? Okay. Okay, we got a couple of questions here. We've got a question. You recently secured strategic engagements in the defense sector and for AI-driven robotaxis, given the supply chain challenges with your Asian vendor in Q1. Oh, right. Yeah, that was a one and done. That was actually. That supply chain issue has been resolved. Like, it's not at all in any way a long-term issue with our Asian suppliers. We don't see that in the future. I think in terms of you know the defense sector, stay tuned with that one. We'll see sooner than later.
That really dovetails with what I said in my presentation. I mean, I think, you know, again, Fixed Wireless is here and now. You're gonna start to see those defense announcements coming over the next couple of quarters. I'd say the robotaxis, if there's—you know, that will be as soon as fourth quarter or Q1 of 2027. Furthermore, actually to answer the second part of your question, we for the Asian supplier situation, even though we have resolved the issue. I would actually almost say chalk it up to a bit of a misunderstanding, but that's been resolved. We have brought on a second supplier to mitigate that situation in the future.
You know, that obviously stressed a pain point for us, so we've made sure to kind of address that for the future. Although we don't see that particular supplier having that issue again, we don't really wanna take that chance for the future. I hope that answers your question. Okay. Sorry. For non-scientific...
I think it's number two here, the question's regarding the current-
Okay.
...you know, price pressure near the $1.20 level. I'll take a first crack at that.
Yeah.
Ron, and you can chime in. You know, I think the current price pressure was more driven by our earnings call, you know, a week ago Monday, and our announcement on the guidance for, you know, Q1. The market, you know, reacted on that. You know, obviously we were disappointed that we had the vendor issue and supply chain issue and had to push the order from March into April. I think there was definitely some reaction to that. You know, to kind of dovetail on Ron's response to the previous question, you know, between the, you know, robotaxis and some of these other applications, you know, they'll start with, you know, non-recurring engineering, and some licensing.
Mm-hmm.
Initially and then transition to product sales. From management's perspective, those non-recurring engineering and any paid licensing, you know, is truly an offset of R&D expense and drops kind of straight to the bottom line, since for most of those we don't have added incremental costs, you know, unlike a product sale where you have COGS. So we're looking to secure initially NRE to work on these, since in many cases our product isn't set up to be plug and play for these applications, particularly the defense side. We've been working with IgniteNet, our customer there, to to make modifications for the specific application, i.e. reducing power consumption, things like that, uh, given the, you know, the nature of the application. As far as, you know, there also is some financing overhang.
Right now without additional financing or meaningful NRE, you know, at a roughly $3 million cash level, we don't have, you know, adequate funding to achieve, you know, revenue breakeven based on the visibility we have and our revenue forecasts, you know, et cetera. You know, we're targeting to be there kind of in the first half of 2027. You know, we would need additional, you know, NRE and, you know, in the shorter term to provide us with enough runway, particularly given, you know, the visibility on the forecast, et cetera, from where we sit right now.
Actually, there's one more question, Jim, which is the fixed wireless part of the long-term strategy. You know that, you know, no, absolutely. That, that, you know, that business is growing, frankly. It, you know, we're investing very little in it. I would call it almost a cash cow. For us, it's, there's, you know, there's no reason to. You know, when you say part of our long-term strategy, I mean, it's just part of our long-term cash management. You know, we, you know, from a R&D perspective, you know, it's done. But we continue to grow sales and grow customers, just because, you know, the product is maturing.
We, you know, we really want to obviously exploit that technology for its cash or that market for its cash contribution to the company.
I think, Ron, when we started out, you know, that was 100% of sales.
Yeah.
As we've worked over the last you know year and a half or so to really focus on diversifying the customer base, we're bringing that percentage down. I think today we're-
Yeah.
we're 80/20, 75/25.
Yeah, I'd say.
Um-
75/25 if you include NRE.
Yep.
Yeah.
Absolutely. That's been a key focus is to diversify and expand beyond that. We cover a fair amount of the market by being a supplier to Ubiquiti-
Yeah.
...to incorporates our chips into their products.
Definitely.
Definitely it remains part of the long-term strategy. In addition, our DUNE product is basically-
Mm-hmm.
...a Fixed Wireless Access product-
Mm-hmm.
...just a broader solution-
Yeah.
... for Dense Urban Networking Environments, so, you know, we still expect that to, you know, to be meaningful.
That's true. Mm-hmm.
I don't see that dropping below a meaningful percentage of our revenue you know, in 2026. I think, you know-
Oh, sure.
...maybe 2027 and beyond.
Yeah. Yeah. Not in the short term, that's for sure. All right. Well, that gets us to the top of the hour it looks like. I encourage anyone to have a one-on-one with us tomorrow. We're available all day, so hopefully we can speak some more, you know, get into some further detail. I appreciate everyone's time today though. We appreciate everyone's time. Thank you very much.
Thank you. Have a good day.