Good morning, everyone. I appreciate everyone coming to our conference. My name is Blake Roberts, a Partner at Centri out of the New York office. I'm happy to introduce Ron Glibbery, Chief Executive Officer and Co-founder of Peraso. Appreciate you guys being here today. Thank you.
All right. Thank you everyone for attending this morning. As Blake mentioned, my name is Ron Glibbery. I'm the CEO of Peraso, and I would like to thank you for coming to our presentation today. Peraso is a semiconductor company. We do wireless chips in high frequency. We call it 60 GHz mm wave. We're the global leaders in this technology. Basically, we've really outperformed our competitors, if you will, and really are the dominant player in this space. We've sold over 2 million chips, and we put a lot of effort into our patent process. We've got over 60 patents issued and pending. We continue to actively secure and expand design wins across a diverse set of markets, which I'll tell you about. Really our first pillar is what we call fixed wireless access. This is high-speed access, internet access.
This is actually getting a lot of press these days because basically, people talk about fiber, but as it turns out, fiber is very expensive and it takes a long time. It's really fiber-like speeds, kind of in the multi-gigabit range. That's really our anchor market, if you will. That market, we really are the dominant player, and there's really nobody left to really go after. We've done a very good job of kind of blanketing that market. The market that we're focused on now is what we call tactical communications. We did a press release this morning, for example, specifically for drones. That's a very hot topic. Our claim to fame there, and I'll get into more detail, is that we can't be detected in a battlefield environment. We're stealthy.
Militaries around the world are recognizing this capability. The idea there is, and you can see on the drones, the idea is to control these drones using our wireless technology, but it, again, can't be detected with traditional. If you've seen the videos out of Russia where they're using fiber optic cable to kind of control their drones, that's the way they're solving the jamming problem now. We avoid the jamming with our technology. Lastly, we're actually starting to see some nice traction on what we call high bandwidth AI, and that's a little more in the future. The idea there is, robotics like Optimus Robotics, robotaxis, like even video on the edge of the internet actually needs very high data rates. One characteristic of our technology is very high data rates and secure data rates. That's a third market.
These are really the main pillars of our markets. Just a quick engineering lesson. Basically what we do is we operate in these high frequencies. What we do really well is we beamform. This is why we're stealthy. Basically the idea is that because we have a very narrow beam, it can't be detected on the battlefield, and it can't be jammed. Again, anti-detection and anti-jamming. Also in the fixed wireless space, what that allows us to do is to actually have very, very high-density networks, very high-speed networks as well. We do this because we've actually got this capability to create these very narrow beams. That's, I guess, a characteristic to keep in mind throughout the presentation. That's kind of the technology behind what we're doing.
In terms of growth, millimeter wave is really the thesis of the company originally was, we're going to run out of spectrum at these lower frequencies, and that indeed is what happened. Actually, I just read a report this morning that, even in America, for 5G, it's taken a long time, but now they are running out of spectrum at the 3 GHz band, and so they're starting to move over to our frequency bands, if you will, called millimeter wave. The growth for the broad market is about 25%, annualized. Specifically for defense, we're putting a lot of effort into that market specifically because of the stealth capability of our chips. We started off, our lead customer was a company called Ubiquiti. It's a OEM. In terms of the signal chain, so basically we are a semiconductor company, so we sell to OEMs.
Really the gorilla in the room of OEMs in this marketplace is a company called Ubiquiti. We've got eight or 10 products that they're selling in the marketplace. They're really kind of the main guy. Their really focus was what we call rural America, where the big carriers couldn't really address those market opportunities because it was low density. That's how we got started, but they're a very strong company, so it allowed us to pipe clean our process, our technology, and our operations. Because of the beamforming that I talked about, basically what that's allowed us to do is to go to what we call urban markets, so much more volume, much higher density markets, so we can get multiple access points into those marketplaces.
Because of our beamforming capability, it allows us to get many networks in a small, dense area. Now we've really rolled our products out with partners, of course, in Los Angeles. We've got a partner called WeLink there, that we've started to roll out about two years ago. South Africa, so in the townships of South Africa, as it turns out, those are probably the most high dense urban places in the world. We've got partners there, that we're shipping into and even places like Kenya. We call this technology DUNE for Dense Urban Network Environment. Basically, again, the idea there is that we've got very strict beamforming, very high data rates, again, competing with fiber in terms of our data rates. A characteristic that's becoming more popular is low latency.
For gaming, for example, when you're doing gaming, you can't have a lot of latency, of course, in that transaction. We have a low latency technology. We started shipping into this market about two years ago. For tactical communications, basically, again, the idea there is that we have these very narrow beams that, A, we call it low probability of detection, but it's essentially zero. I would call it zero probability of detection because you can imagine you've got a beam the size of my finger in 2 sq mi. How are you going to find that beam? It's almost impossible for the enemy to find those beams. Once they do, it's more or less impossible to jam it. Again, we're starting to see real traction in this marketplace because of those characteristics.
We've got a variety of applications, but basically, the net there is that what we're really seeing is just those simple communications links on the battlefield using our technology so that they have secure communications and they can't be jammed. We did a press release this morning, so if you go to our website, you can see that. Basically, it's with this company. Well, our lead customer is actually in the States with the US military, but that's really more prototypes at this point. The first press release we did was in March, and basically, the idea there was, in this situation, this is a solution for the IDF in Israel, and it's basically the idea there was that, as it turns out, 30% of casualties in the IDF were from friendly fire. Okay?
Well, it's obviously a situation that has to get fixed because a lot of the casualties are from their own soldiers. If you look at this situation, what they would do is they would put one of our radios on the rifle and put a receiver on the helmet, and then basically, before you take the shot, you send a signal out, make sure that it's not your guy, and then you can engage. If it is your guy, obviously, you're not going to engage. We've sold about a thousand systems to the IDF so far. The technology per se is called Identification Friend or Foe. That's really the first volume production that we have in this marketplace. Again, what the IDF was looking for, obviously, the main thing is for safety and a communications protocol. Obviously, it's got to, again, stealthy.
You don't want people knowing where you are. Just to put that in perspective, the idea there is, in the Ukraine war, there was a group of Russian soldiers who actually were using their cell phones, and the Ukrainians immediately triangulated to their position and sent ordnance in that direction. It's a real problem. Where we're really seeing the, I would call it, the fallout of that is that if you look carefully in Ukraine or even Israel, the drones part of the marketplace historically has been controlled by traditional wireless technology that can be detected in the marketplace. Now what you're finding, for example, even as we speak in the Iran war, the Shahed drones that Iran is using actually doesn't use any wireless. They just actually hope for the best. They've got some guidance system.
The Russians are actually using fiber optic cable. We're starting to do tests now with various militaries, including the IDF, using our chips to replace existing wireless in those circumstances. We think that's very important. I would say, in summary, we've been very successful in the high-speed internet market, but starting in 2026, this is the market that we're really going after. We've been engaged with these guys for a couple of years now, so it's not like we're starting from scratch. Basically, the two scenarios are in the drone situation. If you have a swarm of drones, for example, the military wants to know who's friendly. It's the same concept. Before you take a shot and you shoot down that drone, you have to make sure that it's not your own drone.
Alternatively, on the drone, you actually want to send a beam down to Earth, and if it's a friendly entity, you don't want to shoot him either. It kind of works both ways. That's really where our focus is. Our press release this morning actually spoke to our ability to go into production, and so we've started that. Basically, that was important for us because it basically shows we actually took our original products that were designated for high-speed internet, and so they needed two things. One was make it much smaller so it could actually fit on a drone, and the other was power consumption. You can imagine all of these applications are very dependent on batteries. You're in the battlefield, the drones are battery operated, but the soldiers have battery packs as well, and they need to last a long time.
We spent about a year and a half actually modifying our main products so that, A, they were much smaller, and B, they're much less power hungry, like one-tenth the power. That was a significant milestone for our engineers. The last thing I'll talk about in terms of applications this morning is what we call Edge AI. The idea here is that we've all heard about AI, of course, and the issue with AI in general is just the high speed it needs to the memory in the data centers. Basically, the issue with that is that you just need very, very high data rates. That's fine in the data centers, and they're using fiber to kind of address that.
The problem is a lot of the AI applications, particularly with regards to video, are actually at the edge of the Internet. If you take one of our main, I guess it's still early, but we're working with a robotics company for manufacturing. The idea is that the entire factory will be robotics-driven, and they want each of those robots to have access to very, very high data rate wireless. The problem with traditional wireless, and believe me, they've tried, and when I say traditional wireless, what I mean by that is Wi-Fi. The problem with Wi-Fi is why it works so well in here is that it lights this room up like a light bulb. Okay, it goes everywhere, and that's great.
When you get into a factory environment, if you have a dozen of those, they cancel each other out, and so you have a very limited bandwidth. Again, I go back to our beamforming capability, and that allows us to get into very high-density environments. In the example of the robotic control and manufacturing, we can have many access points existing in that same room. We get into that high-density concept again. We're really seeing some nice traction here, robotaxis as well. We've done a press release on that. The idea there is that robotaxis actually gather terabytes of information throughout the day, and they want to get that back to the data center, process the model, rejig their models, and send them back to the cars. The problem is, in an hour, they've got an hour to recharge.
They need to get a terabyte of data off the car in that time. Basically, it's a very, very high data rate application. Again, the problem is when you get back to the staging center where you've got, let's say, 100 robotaxis, basically, again, you'll have congestion. We avoid that congestion, and that's really the hallmark of our technology. We actually did a press release with a company in California called Virewirx. We're working with a very large client now who's a large robotaxi manufacturer. You can kind of guess who that is. Basically, it looks very promising in terms of that application and moving forward with that application, because they've tried all wireless technologies, and again, traditional wireless has this kind of light bulb effect, if you will, and they interfere with each other.
In our situation, with our technology, we can get to very high-density applications. Really, the main thing here is, again, on video delivery, again, very high data rates at the edge. We're actually working with a large customer there. They've got like a half a million delivery drivers, and basically, they want to get video, again, back to the data center. The only way they can get it off their headsets fast enough is using our technology. We think video is another good application for us as well. I would say the three pillars of our business today that we want to focus on are high-speed Internet, military, and the robotaxi kind of robotic applications. I think the other thing to keep in mind is, so again, I would stress we're a semiconductor company, so we sell to OEMs.
We've probably got in the order of 30 customers shipping. We have almost 60. I think this is up to 60 products in the market now. Basically, it's actually kind of interesting to see because when these customers go into production, you can see that there's a lot of effort they have to go through to get those products in production. We're a real company in terms of just really shipping these parts. A lot of it is military, so it's really complex operations to kind of get these guys to market. I would also stress that we're quite a mature company. We've actually been around since 2008. This technology, the other thing I would stress is that our moat is significant. It's very, very difficult, almost impossible for anyone to compete now, because we've done such a great job of capturing these market opportunities.
The eight minutes is including questions?
Yeah.
Okay.
Yeah.
I've got eight minutes. All right. We're a global pioneer in this technology. It's called millimeter wave. We're really focused on current and future applications. The original thesis, as I mentioned, is that there was spectrum limitations on the bandwidth, and that's the problem we're solving. We see it every day now in terms of our customer base. The problem they have is that existing technologies can't handle the density that's required for these applications. It really is playing into our favor. Of course, the idea that we're stealth, when we started this company 15 years ago, nobody was saying, well, you know what? Guess what? Nobody can jam you in on the battlefield, so that's going to be a great application for you. It's just amazing how these applications have kind of evolved over the years. Yeah.
There's a very high barrier to entry. I didn't go into technology at all, but this ability to make these beams from a technology perspective is very difficult to do well, and we continue to evolve that. We think, over 2026, the military application, 2026-2027, is going to be really the growth opportunity, but also Edge AI. We have a robust pipeline of diverse customers in all these applications. Really that's the main focus of our technology over the next two years. That's a good summary of who we are and what we do. Our CFO ran into logistical problems this morning. Our revenue in 2024 was $14.6, and we went to $12.2 in fiscal 2025. The reason is because we had a legacy business.
We came into this market, we merged with a company that was already on the Nasdaq, and so we lost about $2 million of revenue there. We ran last year about $12.2, and Q4 was at $2.9. We really are targeting, I think, 50% growth or I would call 30%-50% growth. That's really what we're looking at in terms of our growth opportunities. Fiscal 2024, we lost $5.1 million. In 2025, we lost $4.3 million, and that was because even though our revenue was down, we actually put a lot of effort into cutting expenses. That's a really important thing to keep in mind about our company as well is, a semiconductor company is a very expensive, capital-intensive thing to run. We've actually trimmed all that until we get profitable.
That's another thing to keep in mind about our company is that right now we're focused on eliminating the cash burn and get the cash-generating. That's really the plan over the next couple of years. This is our income statements for non-GAAP income statements for Q4. We're doing Q1 sometime in May. Yeah, you can see our revenue was about $2.8 million. Our margins are at over 50% as per standard semiconductor margins, if you will. Our operating expense was $2.7 million. That's our core operating expense is about $2.7 million per quarter. Pretty reasonable and we think a real path to get profitability. Hop into that. Yeah. From a cash perspective, it's kind of interesting. In fact, this morning we announced a new ATM. Basically we've been focused on using our ATM, and it's been a good way to shore up our cash.
We don't have a lot of cash, but we actually have a continuous, nice stream of cash. That's been working well for us. We haven't really done a huge financing for quite some time now, right, because we've been able to raise money on the ATM. That's about it, so thank you very much for joining us today. Do you have any questions, comments, concerns? Nothing? Going once.
I'll ask one if I might.
Okay.
A question for you, just from a military standpoint.
Mm-hmm
I'm just thinking of it as a guy who hauled all that gear around, and just thinking about adding extra weight on my weapon.
Yeah
On my Kevlar, how much weight are you adding, because you're talking battery packs, things like that for the IFF that you'd be putting on each of the soldiers?
Well, that's a great question. We don't make the end product, of course. Our partner does. They're called InTACT in Israel. Our chips themselves would weigh maybe 4 oz, right? I think the primary weight would be the battery, right? The battery that they're using would be actually, frankly, on a soldier, would be, I'd say, a quarter of the size of the battery on your phone. That's the metric. This is maybe 5,000 milliamp hours. I think the IDF wants a quarter of that, and they want it to last a month. I don't know. I would say, how much does a battery weigh? It's a good question. Maybe eight ounces total that would add to your helmet and your rifle. Yeah, less than 16 oz. I'd say about 8 oz -10 oz. It's a good question.
I could circle back with our customer and see what the total weight is. Very good question from an expert. Hi.
Do you project when you'll be cash flow positive? I can probably calculate it myself, but are you stating that publicly?
Obliquely, we're hoping this year, but early into 2027, so that kind of timeframe, yeah. Hi.
Why would your customers pick you over the larger players in this space?
Well, actually our large customer was Qualcomm, and two things happened. One, Qualcomm's product had several problems. They were winning. We were splitting the market, but now we've taken over every single customer because their product didn't work and they didn't revise it. Actually, there were several competitors, but as I said at the start of the presentation, we've actually beat them out. Well, a lot of them are small. Qualcomm's obviously not small. Basically, they didn't execute, and then they didn't want to spend the money to catch us. They actually use our chips in some situations, actually. Yeah, the main reason is because there's no other competitors in the marketplace that have this. That's the main reason. Anything else? All right. Thank you very much for coming and seeing our presentation today. I appreciate it.